----------------------------------------------------------------------- ; Copyright (C) 1999-2000 ; Associated Universities, Inc. Washington DC, USA. ; ; This program is free software; you can redistribute it and/or ; modify it under the terms of the GNU General Public License as ; published by the Free Software Foundation; either version 2 of ; the License, or (at your option) any later version. ; ; This program is distributed in the hope that it will be useful, ; but WITHOUT ANY WARRANTY; without even the implied warranty of ; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ; GNU General Public License for more details. ; ; You should have received a copy of the GNU General Public ; License along with this program; if not, write to the Free ; Software Foundation, Inc., 675 Massachusetts Ave, Cambridge, ; MA 02139, USA. ; ; Correspondence concerning AIPS should be addressed as follows: ; Internet email: aipsmail@nrao.edu. ; Postal address: AIPS Project Office ; National Radio Astronomy Observatory ; 520 Edgemont Road ; Charlottesville, VA 22903-2475 USA ----------------------------------------------------------------------- This file must be used to record any changes made to the Master version of AIPS kept in Charlottesville. An entry is: line 1: Date Program name Your name line 2: What was done (may span multiple lines) line n-1: Where has this change been moved (e.g. 15JUL94, nowhere) line n: Blank The details of line 1 are important since programs are used to find specified strings and do conversions on them. It MUST be typed as follows: 1 blank, 4 digits, a period, two blanks, then the Month, day, comma, year, some space, a title, some more space, your name. *********************************************************************** *********************************************************************** Changes while 31DEC00 (31DEC99) was TST *********************************************************************** *********************************************************************** 10387. September 23, 1999 POSSM Chris Moved the statement incrementing the number of a plot on a page in MULTPL so that it is only executed if the plot is not blank. This avoids problems setting the coordinates and opening the plot file if the first plot on a page is blank and should also save some paper. Moved to 15OCT99 on this date. 10388. September 27, 1999 $QPGNOT/UVCON Leonia UVCON has not had default for the number of frequency channels. I added: IF (NFRCHA .LE. 0) NFRCHA = 1. The help file has been updated. Moved nowhere. 10389. October 6, 1999 $APGNOT/SPLAT Leonia SPLAT has been created wrong file if one source is selected in the option ASSEMBL. SU table was not created and random parameter SOURCE was REMOVED. It happened because UVGET changes ILOCSU to -1 if only one source was selected. I circumvented the problem changing the list of selected sources to blanks if only one source is selected. The minor following modification have been fulfilled also: SOLINT changed from beeing measured in seconds to minutes. The message reminding about NX table creating by INDXR has been added. Moved nowhere. 10390. October 8, 1999 $APGNOT/MATHS Leonia The default for CPARM(2,3) has not been determined explicitely for optype POLY. So they are not zero in all machines. I fixed it. Moved nowhere. 10391. October 8, 1999 $APGNOT/SPLAT Leonia As I found the problem of one source to be assembled has not been solved in all cases by the way described at 10389. So I have simplified the solution creating a single source file if one source is selected even if assembling (APARM(7) =0) was requested. The relevant message has been added. Moved nowhere. 10392. October 9, 1999 $HLPFIL/MCUBE.HLP Leonia IN3SEQ has had the restriction (-10,10). Craig Walker needs remove this restiction. I did it. Moved nowhere. 10393. October 18, 1999 VBGLU Chris Fixed an error in retrieving the catalogue number for the output file when updating the history records that caused the header of one of the input files to be corrupted when there were more than 2 input files. Moved to 15OCT99 on this date. 10394. October 19, 1999 FIXWT Chris Revised FIXWT to use the real parts of the visibilities to estimate noise levels rather than the amplitudes. This should give better estimates of the noise at low SNRs. It also seems to reduce the occurrence of isolated points with discrepent weights. Moved to 15OCT99 on this date. 10395. October 20, 1999 $YPGNOT/CONFI Leonia The two new options have been added. 1. The possibility of optimising inside of a sector in the sky limited by two circles and two radius. 2. The new inititial configuration and relevant constraint in the shape of two circles with inner one tangent inside to the outer one. The last option is added in view of serios topography constraint in the ALMA site at Chilie. Moved nowhere. 10396. October 20, 1999 $YPGNOT/VPLOT Leonia Parameter MAXBT has been increased from 100000 to 1000000, to allow plotting of more data. Moved nowhere. 10397. October 21, 1999 FRPLT Chris Removed an assignment to the unused and uninitialized common- block variable UVSCAL that included a division by its uninitialized value. Moved to 15OCT99 on this date. 10398. October 25, 1999 UVMAP Eric Believe it or not, there are still people using UVMAP/APCLN. They discovered that the xshift parameter is in backwards for the output header, messing up self-cals... Moved to 15OCT99 October 26. 10399. October 28, 1999 PLOTR Eric Found a bug when plotting symbols. The plot paraform has an internal scaling which was not applied to the symbols coordinates before calling the general symbol plotter. The failure affected only certain ranges in the data. Moved to 15OCT99 this date. 10400. October 28, 1999 ACCOR Leonia ACCOR has not worked if NX table was absent. The routine UVTIME has gotten zero for LREC because UVPGET (which estimates LREC) was called after UVTIME. I moved CALL UVPGET before CALL UVTIME. Copyleft fixed 29Oct (Eric). Moved nowhere 10401. October 29, 1999 TICINC Leonia/Eric Small plots sometimes end up with only one tick mark which does not say much about the scaling. Fixed a .LT. to .LE. which should require more ticks in general. Moved to 31DEC99 this date. 10402. November 1, 1999 TVBLNK Eric TVBLINK did not work right on XAS with 8-bit pseudo color display. It tried to do split screen which it can do, but it cannot have 2 different LUTs. Fixed code to recognize the less capable situation and to do 1 channel at a time. Moved to 15OCT99 this date. 10403. November 1, 1999 SPLAT Leonia All tables have been copied with subarray=0. I fixed it adding relevant cards in subroutine COPTAB. Moved nowhere. 10404. November 1, 1999 CLFIX, CL2HF, ASTROMET.HLP Ketan/Dave Gordon Dave Gordon's latest batch of changes to the astrometry programs include: modifications to handle different CL table epochs at the reference and remote stations, calibrating by subarray, a new algorithm for the Haystack reference frame conversion, and some other minor things. He reports that he has run through the data path end-to-end with no problems. Moved to 15OCT99 this date. 10405. November 8, 1999 FITLD Chris for Eric Fixed a faulty FORMAT reference that caused FITLD to complain about running in batch when it should have complained about reading too many files that were not FITS files. Moved nowhere. 10406. November 8, 1999 CALIB Leonia CALIB using ANTUSE with CPARM(2)>0, has printed information about the average gain as: "The average gain over 'all' antennas ...." although only antennas selected by ANTUSE participate in the averaging. I changed it providing additionnaly the list of antennas participated in the averaging, which is selected by ANTUSE. Moved nowhere. 10407. November 8, 1999 UVMTH, DIFUV Eric UVMTH did not defend itself adequately when applying one data set to another. It used the NVIS from data set 1 to read data set 2. (Common variables strike again.) I fixed it to handle compressed data, to use dynamic memory for the summing array (which made it go much faster), and to handle the two data sets independently. Now, if the user at least sees to it that baseline 1-2 in the one is about the same as in the other.... DIFUV also lacked defenses. Fixed NVIS and made it handle some differences between the two data files. This one really requires them to be very similar. Moved nowhere. 10408. November 9, 1999 SPFLG Eric Changed the SWITCH ALL-BL FLAG option from current vs all baselines to a 4-position switch with all baselines to each 1 of the antennas included as choices. Changed selection of data to flag, status display, SPFLG.HLP, HLPSPFLG.HLP. Moved nowhere. 10409. November 10, 1999 Editing Eric Based on Ketan's latest round of suggestions, changed: EDITR Add REASON adverb and pass it on. EDITR HLP: Add REASON adverb. Add amplitude-range setting option, NEXT CORRELATOR, EXPERT=TRUE affects EDITA Add REASON adverb and pass it on. EDITA HLP: Add REASON adverb. Add amplitude-range setting option, NEXT CORRELATOR, EXPERT=TRUE affects SNEDT HLP: Add amplitude-range setting option, NEXT CORRELATOR, EXPERT=TRUE affects SCMAP HLP: Add amplitude-range setting option, NEXT CORRELATOR, QEDIUTIL 1. To read in REASON, to set it to TASK/TIME/DATE if blank when written to FC table. 2. If these tasks start with EXPERT set to true: a) The ALL-pol and ALL-IF flags are set true, else they are false b) The display ranges are set to 0:max for amplitude, -180:180 for phase, -1000:1000 ns for delay, -1000:1000 mHz for rate These (1 & 2) can be changed interactively. 3. Add self-scale max but keep min at 0 option for amplitude plotting. 4. Add NEXT CORRELAT operation which increments polarization and IF: R/1, L/1, R/2, L/2, etc to both TV and expert menus. 5. Changed to hold off computing running differences until they are needed. Also changed HLPEDICL.HLP, HLPEDISN.HLP, HLPEDITY.HLP, and HLPEDIUV.HLP, to add NEXT CORRELATOR and to add info on setting the amplitude display range. Moved nowhere. 10410. November 11, 1999 Editing Eric Corrected QEDIUTIL to compute difference data only when needed and only at plot time itself. If failed to do it sometimes when it was needed and it is best to move it to the EDPLOT routine itself which is called from all over. Moved nowhere. 10411. November 11, 1999 DOUVCOMP Eric Corrected FITLD.HLP to state what the task already does: DOUVCOMP <= 0 is uncompressed data, > 0 compressed. The "default" value in POPSDAT.HLP is already 1. Corrected an oversight in AU2A in which this adverb was set to 0 by DEFAULT. Moved nowhere. 10412. November 11, 1999 CALIB Leonia CALIB has not written history about the new adverb ANTUSE. I modified CALIB to add the relevant lines in the history file. Moved nowhere. 10413. November 12, 1999 TBIN Eric TBIN did not test enough things so that it would flag a row at the very last access tio that row. Added another test and improved the error handling. Added a bit to clarify things in the help file as well. Moved nowhere. 10414. November 12, 1999 IMAGR Eric The TELL call was just too late to detect the case in which the user drops NITER to force it to quit and so it looped for the next field and then decided that the infinite loop (overlap=2) condition was present. I have moved the code forward so it can quit abruptly if requested. Moved nowhere. 10415. November 12, 1999 OTFIN Eric Added the OFF scan number to the displays and upgraded the help to advise users what widths are better. Moved nowhere. 10416. November 12, 1999 BSTST Eric Changed it to offer the option of a Clean deconvolution as well as the EKH one. Improved the labeling of RMs's etc Moved nowhere. 10417. November 12, 1999 BSCLN Eric New task to Clean beam-switched images. Moved nowhere. 10418. November 13, 1999 BSMOD Eric Changed the task so that it does not require an input data set. It now takes observatory latitude, longitude, radius from center of earth, telescope diameter, REFDATE, OBJECT in addition to the old adverbs. Frequency is set by beam width and diameter. The 12M is the default for the observatory. Moved nowhere. 10419. November 15, 1999 New adverbs Eric Created new adverbs to help replace the overused *COUNT ones. Changed POPSDAT.HLP and NEWPARMS.001 to create them. Created new help files for BPRINT, EPRINT, NPRINT, NPLOTS. Changed NCOUNT to ORDER in BSGRD and BSCOR. Changed NCOUNT to NPLOTS in CONFI, ELINT, POSSM, SNPLT, VPLOT. Changed *COUNT to *PRINT in PRTAN, PRTSD, PRTUV, UVPRT, HFPRT, and PRTCC. Changed BCOUNT and ECOUNT but not NCOUNT to *PRINT in PRTAB. Changed PRTSD, UVPRT and PRTUV to allow 30000 NPRINT on 0 for CRT and to limit NPRINT to 2000 for printer. Also changed PRTUV to do extra decimals on the time when possible. Moved nowhere. 10420. November 15, 1999 APCAL Ketan APCAL would refuse to do opacity corrections if source fluxes were not found in the SU table. Now, APCAL will go ahead assuming a source flux of zero - a reasonable default - with a stern rejoinder to the user that they may have wanted to run SETJY. Also, APCAL would refuse to do opacity corrections if a text file containing weather information was not supplied. Internally, the only use of this file was to find the maximum recorded ground temperature for each station. The WX table is now supplied by VLBA calibration transfer and APCAL needs to be made to read that table. For now, the behaviour of APCAL has been changed not to crash with opacity corrections are requested and no weather text file is supplied. The new behaviour is to assume a default maximum ground temperature of 20 C at each station, _if_ the INFILE is left blank [with a warning of course...]. Moved nowhere. 10421. November 15, 1999 KRING Ketan/Craig W. Craig was lamenting that for any reasonably complicated model the model division stage of fringe-fitting was very slow. The model division code in KRING (and in FRING) is hard-wired to request only DFTs. Based upon Craig's assertion that using a gridded FFT results in no noticeable degradation of the final solutions, the code has now been changed to let the model-division routines choose the fastest method rather than insist upon using DFTs. Craig reports a tremendous improvement in speed. Moved nowhere. 10422. November 15, 1999 OMFIT Ketan At user-request, added a new mode to allow the user to strip the amplitude information off of a UV data set. The new switch is set using OUTVER=7 and sets the amplitude of each incoming visibility with unity in the output UV data set. Moved nowhere. 10423. November 16, 1999 GRBOXS Eric The call to DRBOXS to erase the current boxes can return negative IERR now without cause for alarm. Unfortunately, in one of the 2 calls IERR.NE.0 was used. Moved nowhere. 10424. November 16, 1999 IMAGR Eric Changed UVWAIT - the Linux compiler cares how the round-off computation is done. A = B*C ; I = A+0.5 gets a different results than I = B*C + 0.5. The Q routine was doing the latter, UVWAIT was doing the former. This is going to get us. Moved nowhere. 10425. November 17, 1999 EDITR, EDITA Eric Changed QEDIUTIL, EDITR, EDITA, SNEDT and their help files to add the ANTUSE adverb. This is use to specify which antennas appear in the initial display on the TV. Moved nowhere. 10426. November 17, 1999 TICINC Eric Added code to figure out the actual number of tick marks in a plot and if it is <= 1, to decrease the interval to get more even if that is more than would be desirable. Made a similar change to the equivalent routine (EDTINC) inside QEDIUTIL. Moved nowhere. 10427. November 17, 1999 FITLD Chris Fixed an error in the VLBA saturation correction that was introduced in Version 10.5 and that could lead to errors of the order of 10% in auto-correlation amplitudes. Incremented the revision number to 10.5.1. Moved nowhere yet: will need to be released as a patch to 15OCT99. 10428. November 19, 1999 SPLAT Leonia SPLAT has given wrong reference frequency if BCHAN .NE. 0 for APARM(1) = 3. I have fixed it. Beside of that I improved the help file to clarify relation of APARM(1) and CHANNEL and CHANSEL. Moved nowhere. 10429. November 21, 1999 VPLOT Leonia VPLOT has not worked for the hude data if APARM(3)>0. The parameter MAXSOL should be increased until MAXBT. I have fixed it. Another small improvement has been done: If APARM(3)>0 and APARM(4)=0 and BIF=EIF, then MIN is equal to MAX, because MIN/MAX is found comparing the value in different IFs. So if user forget to put APARM(4)>0, then there will not be vertical line in the ploting of points at all. To help user I added the following card: IF (DOAVG .AND. BIF.EQ.EIF) APARM(4) = 1 Moved nowhere. 10430. November 21, 1999 CROWDED Eric Created new adverb CROWDED and added it to EDITA, EDITR, and SNEDT (and of course POPSDAT and NEWPARMS). Also new help file. Changed QEDIUTIL to read the keyword and, when true, to allow multiple polarizations and/or multiple IFs to be displayed and edited at the same time in EDITA, EDITR and SNEDT. Corrected a bug which occurred when the display was crowded (> 1 sample at the same TV x pixel) and FLAG TIME was used. Not all samples on that vertical line were deleted. Reworded HLPEDITCL, HLPEDISN, HLPEDITY, HLPEDIUV as well Moved nowhere. 10431. November 22, 1999 VPLOT Leonia Recentely (#10429) I modified VPLOT fixing APARM(4) =1, if APARM(3)>0 and BIF.EQ.EIF. Craig Walker considers that it is a bad idea changing input parameters inside of a task. So I removed the line: IF (DOAVG .AND. BIF.EQ.EIF) APARM(4) =1 and added the note: Note: APARM(3)=1 but SOLINT=0, APARM(4)=0 and BIF=EIF so MIN=MAX. Moved nowhere. 10432. November 22, 1999 FLGIT Eric Fixed writing of selected sources into HI file. Moved nowhere. 10433. November 22, 1999 JMFIT Eric OFFSET was correctly scaled and used with the image data. It was not rescaled for reporting to the user. Moved nowhere. 10434. November 22, 1999 DBCON Eric DBCON forgot that the number of output correlators can exceed the number of input in 1 of the 2 data sets. It also did not zero the right array to help cover that. Fixed. Moved nowhere: should go to patch. 10435. November 23, 1999 CALIB Eric Added more comments in the help file about what has to be set to do Clean models on a multi-source file. Moved nowhere. 10436. November 24, 1999 UJOIN Eric Changed handling of tables so that the IF selection will be made in them. Moved nowhere. 10437. November 24, 1999 CALIB, BPASS ERic Changed both to try to read Craig Walker's mind. The task will quit if a. Mutiple sources selected and both IN2NAME and IN2CLASS non-blank b. Single-source file with SMODEL=0 and one of IN2NAME and IN2CLASS blank c. Single-source file with SMODEL>0 and neither IN2NAME nor IN2CLASS blank Note that BOTH IN2NAME and IN2CLASS must be specified, SMODEL(1) is ignored for multi-source files, and must be > 0 for single-source point modeling. Moved nowhere. 10438. November 26, 1999 $APGNOT/UVHOL.FOR Leonia UVHOL has been able to keep the given number of samples at the end of each pointing, dropping the relevant samples at the beginning. Durga Bagri needs to drop samples at both beginning at the end of each pointing. The new option has been added to satisfy this request. HELP file is updated. Moved nowhere. 10439. November 30, 1999 File naming Eric Due to user complaints I have changed the extended hex naming of fields to regular names. This will have a bunch of affects - the worst being that NEW cannot read the new names. xCLNnn -> xCLnnn, xMAPnn -> xIMnnn, xBEMnn -> xBMnnn and the new numbers are 001 - 512 rather than blank, 1-511. Changed AU5C Recognize either old or new names (by the 4th character) SETGDS Recognize either old or new names. LPCAL Recognize either old or new names. GPHAS Recognize either old or new names. IMCREA Create xIMnnn or xBMnnn class images. Corrected error - LFIELD was not initialized if creating an image here. WFCLN Change output class names. MX Change output class names. APCLN Changed default input and output classes. SDCLN Changed default input and output classes. SCMAP Changed output class names. VLABP Changed default input class. VLACLOAD Changed *CLN to *CL001 and *BEAM (wrong) to *BM001. VLALLOAD Change IMAP to IIM001 etc DDTLOAD Changed IMAP to IIM001, ICLN to ICL001, etc MAPIT Changed class names HYB Changed class names, corrected IBEAM to IBM001 COMAP Changed class names, corrected IBEAM to IBM001 FLATN Changed to recognize old or new names. Changed comments in help files for: WFCLN, MX, APCLN, SDCLN, TABED, ALLDEST, SCMAP, VLABP, VLAC, FLATN, IMAGR. Moved nowhere. 10440. November 30, 1999 ALLOKAY Eric Added adverb ALLOKAY to IMAGR to try to allow quicker restarts. ALLOKAY=1 => beam images are already computed correctly. ALLOKAY=2 => beams okay and also the UV work file is a correct weighted and residual data set for the components now in the Clean images. The latter requires BCOMP to be set to include all components and all of the name parameters for the work file to be set and the OUTVER to be set. Also the OUTSEQ must be set. Changed: IMAGR Help file - add adverb and all sorts of requirements. IMAGR Add adverb; test ALLOKAY=2 to see if work files exists (sets it to 0 if not); create, fill, weight the unsubtracted data file if ALLOKAY=2 when filtering CCs; QUVUTIL To write the SUM of the weights to beam images as a header keyword and to use that as a default sum of weights when imaging. QCLEAN To create filtering file only when ALLOKAY low enough, to avoid making beam images and subtracting initial components when instructed. Moved nowhere. 10441. December 1, 1999 FILLM Eric For some reason the weights in the CL table were set to 0. Changed to 1.0. It did not seem to hurt anything, but SNPLT would refuse to plot the 1,0's. Moved nowhere. 10442. December 1, 1999 Naming Eric Changed SETGDS and LPCAL - on old names they were changing the last 2 characters to blank. WE have tolerated anything in those chars and I guess we still have to. In the new names the last 3 chars are 001 and there is no reason to not force this - new names are deduced from finding a number in the 4th character position and it would be crazy for someone to have such a name while being outside of the new name system. MOved nowhere. 10443. December 3, 1999 FDEFAULT Eric Added -Wimplicit and -Wall (unused and uninitialized) to the Linux compiler options. Moved nowhere. 10444. December 3, 1999 TBIN Eric Removed unused variables detected by testing the new options. The g77 unused detector is not fooled by DATA statements while the Sun compiler is. Moved nowhere. 10445. December 6, 1999 CALREF in APLNOT Chris Rewrote CALREF. - CALREF no longer applies a spurious delay correction. - Smoothing times now conform to the header comments. - The correct interpolation table are now used if the smoothing times are small. - CALREF now reports an error if the old and new reference antennae can not be related to one another. - CALREF will no longer fail if there is only one solution linking the old and new reference antennae. - Delays are now interpolated using the rate as the first derivative of delay with respect to time. - Faulty logic that attempted to avoid overflows while computing ATAN2 has been replaced. - Excess arguments (including a large work array) have been removed. Calls to CALREF have been updated in FRING, KRING, CALIB, and CALADJ. Moved nowhere. 10446. December 6, 1999 FRING and KRING Chris Corrected offsets into the column pointer arrays in calls to FNDCOL and CALREF for the second polarization. The erroneous offsets caused CALREF to fail to change the reference antenna for the second polarization. Moved nowhere. 10447. December 6, 1999 OTFBS Eric Changed DSDD.INC amd SDDPTS to comments everything and to set the apperture efficiency pointer. Changed OTFBS to scale the data to something like sky temperature (multiply by eta_l * eta_fss / eta_a). Moved nowhere. 10448. December 7, 1999 JMFIT Eric Fixed default handling of GMAX to match the help file and IMFIT. It was supposed to allow GMAX < 0 and the use DMIN if it is greater in abs value than DMAX. Moved nowhere. 10449. December 9, 1999 ZLASCL and ZLPCL2 Pat Change the lpstat syntax to support CUPS. Does not affect lpr or lp or lpq commands. If the CUPS_SERVER variable is defined, these scripts now assume lpstat is the CUPS version, not the vendor supplied one. Moved nowhere. 10450. December 10, 1999 ZFI2, ZMI2 Eric The error message complaining about being outside the size of the file needs to work with long-long variables. This requires (!) a different C format identifier (%lld rather than %d). Since sizeof is not allowed in #if statements, this format must be compiled into everything - it is seemingly legal everywhere I could check. It does not work on alphas but will not be invoked there. Moved nowhere. 10451. December 10, 1999 QCLEAN Eric Dies if ALLOKAY was not declared - which it is not in SCMAP and WFCLN. Made it optional. Moved nowhere. 10452. December 10, 1999 IMAGR Eric Fixed help file: STOKES = 'RR' and 'LL' not 'R' and 'L'. Moved nowhere. 10453. December 10, 1999 SCIMG Eric New task: A multi-field version of SCMAP. Moved nowhere 10454. December 13, 1999 LWPLA.HLP Leonia HELP file for LWPLA says that if LPEN= 0, then LPEN=3. It is exactly what LWPLA should do. But the range of LPEN given in LWPLA.HLP does not allow LPEN to be less then 1. I modifyed LWPLA.HLP to exclude this contradiction. Moved nowhere 10455. December 14, 1999 VPLOT Leonia When VPLOT is used for editing, a user can desire to edit only one frequecy channel but to flag all of them. It can happen in a spectral project when one channel is used for referencing and for self calibration. The new parameter APARM(6) has been added to satisfy such a requirement. Help file is updated. Moved nowhere 10456. December 14, 1999 SPLAT Leonia SPLAT creates a single source file when multi source file is orded but only one source is selected (see #10391). The relevant message has been printed. Unfortunately the same message has been appeared even if single source file had been ordered. I have excluded the warning appearence in the last case. Moved nowhere 10457. December 14, 1999 IMAGR Eric QUVUTIL (OUVIMG) had a problem in which the gridding routine assumed that the input UV data CATBLK was in common. My recent change to allow ALLOKAY destroyed this assumption in all cases requiring disk-based gridding. The bug was already there when the first field did not require disk-based gridding but a later field did need it. Moved to patch area this date. 10458. December 20, 1999 ZFRE2 in APLSOL Chris Changed struct statvfs to statvfs_t. Solaris 8 EA only changes the second form for large file access. Moved nowhere. 10459. December 20, 1999 OTFBS Eric Changed OTFBS back to scale the data by the counts to temperature scale and the 2 efficiencies (multiply by eta_l * eta_fss) and to write a TY table with TSYS as recorded and TANT as TSYS scaled by the efficiencies (multiply by eta_l * eta_fss / eta_a). Moved nowhere. 10460. December 20, 1999 UVCOP Eric Fixed a long-standing error. When copying single-source multi-array files to a single array it was supposed to subtract the infamous 5 days per subarray. Instead of subtracting 5*(SUBA-1) it subtracted 5*SUBA producing times around -5 days. Moved nowhere. 10461. December 20, 1999 TI2HA Eric The OUTNAME and OUTCLASS adverbs were ignored. Added DOARRAY adverb to give permission to subtract 5*(SUBarray#-1) from times before computing the hour angle. Moved nowhere. 10462. December 20, 1999 Calls to CALREF Chris Modified routines calling CALREF to dynamically allocate sufficient workspace to handle the table being modified. The workspace allocation is conservative in that it guarantees that the workspace will be sufficient. It is unlikely that this will cause problems in allocating the workspace since tables will rarely require more than 1 megabyte of dynamic storage. Also modified the calling routines to issue a warning and continue after receiving IRET=1 (no connection between old and new reference) from CALREF since this is often triggered by the use of different reference antennae in different subarrays. Modified: CALADJ in APLNOT CLCAL in APGNOT (call sequence for CALADJ) CALIB in QPGNOT FRING in QPGNOT KRING in QPGNOT Moved nowhere. 10463. December 22, 1999 FILLM Chris Now writes AC and BD datastreams with different bandwidths in the same observing band to the same output file in continuum mode. This avoids some problems that arise when FILLM tries to switch from one to two output files after the frequency setup changes. Moved nowhere. 10464. December 27, 1999 VPLOT Leonia If SOLINT>0, VPLOT can plot right plots only if X is equal to time and Y is equal amp, phas, weght, rms, real, or imag. I have added the relevant warning if X or Y are not equal to to the above values. Moved nowhere. 10465. December 29, 1999 IMAGR Eric Added a multi-resolution Clean to IMAGR. For m resolutions, it uses m*NFIELD fields. The field parameters and (optionally) Clean box parameters for fields 1-NFIELD are copied to the corresponding fields NFIELD+1 to m*NFIELD. The resolutions are specified with the parameter WGAUSS and NGAUSS are used. There is no requirement toi use a point source, it must be among those requested. FITBM Changed call sequence to add field number and to use that in the message reporting the result. BMSHP Fix call sequence to FITBM (field 0) MX Fix call sequence to FITBM (field 0) APCLN Fix call sequence to FITBM (field 0) SDCLN Fix call sequence to FITBM (field 0) FIELDF Added call argument = # fields for BCOMP. SCIMG Fixed call seq to FIELDF. QUVUTIL Changed taper in OUNFWT to use WTTAPER which is desirable since it is done only once. Put taper back in OUVIMG using UVTAPER. Calling program will choose which to do. CLEAN Changed to get COMPDIAM and set the number of CC columns accordingly. Changed to check NITER before creating CC files. We can now make CC files version N without making N-1 so this should be ok. QCLEAN Beams become function of field number which they should have already been. Add MR Clean widths, controls. Improve beam messages. Fix beam histogram call to be for only current field (another error that previously did not matter). Write CC files with Gaussians and read them. Etc IMAGR Manage the new field numbers, compute the tapers needed to give the widths, pass the taper to the weighting or imaging parts, etc. POPSDAT Create new adverbs IMAGRPRM(20) to replace CPARM(10) making space for future developments and MGAUSS, PGAUSS, and WGAUSS all (10) for 1-D Gaussian parameters. NEWPARMS Added MGAUSS, PGAUSS, WGAUSS, IMAGRPRM. IMAGR HELP: Changed CPARM to IMAGRPRM which gives room to grow new options. Added NGAUSS and WGAUSS to allow multi-resolution Cleans and changed wording for this new option. QTAPER Pseudo Q routine to do taper for a row at a time QGRDCC Added tapering option to call sequence and code. GRDCRM Changed call sequence to return Gaussian widths if found and to all Gaussians as components. It tests for there being only one size Gaussian per file. CCSGRD Changed calls to QGRDCC and GRDCRM to pass on and use Gaussian components. Changed it to call QTAPER and set up its parameters. Moved nowhere. 10466. December 30, 1999 MAPPR Eric Added PROC MAPPR a simplified access to IMAGR to POPSDAT.HLP and created the necessary help file. Moved nowhere. 10467. December 31, 1999 FLATN Eric Corrected typo - the compilers let me WRITE into a non-character variable! Moved nowhere. 10468. January 4, 2000 GRBOXS Eric Fixed the initial cursor on operation - it could request really illegal positions depending on what boxes it started with and what subimage was visible. It now tries to find some legal box and failing that the TV center. Moved nowhere. 10469. January 4, 2000 VPLOT Leonia Added abbility of plotting the following new values for X and Y in the case of time averaging (SOLINT > 0): X: U, V, UV-dist; Only UT has existed early. Y: U, V, W, UV-dist. Moved nowhere. 10470. January 4, 2000 CALADJ Chris The buffer size passed to TABSRT from CALADJ had become coupled to the number of rows in the table to be sorted causing problems for tables with less than about 100 rows. Removed this coupling. Moved nowhere. 10471. January 4, 2000 Doc files Eric Added 31DEC99, 15OCT99 sections to WHATSNEW.HLP. Changed LSAPROPO.HLP since deletions do not make it and MSGSERVER et al. are now fully spelled out. Update ABOUT files: ZZADVERB.HLP ZZANALYSIS.HLP ZZAP.HLP ZZCALIBRAT.HLP ZZEDITING.HLP ZZHARDCOPY.HLP ZZIMAGING.HLP ZZINTERACT.HLP ZZOOP.HLP ZZPLOT.HLP ZZPOLARIZA.HLP ZZPROCEDUR.HLP ZZSINGLEDI.HLP ZZTABLE.HLP ZZTASK.HLP Moved nowhere. 10472. January 5, 2000 FILLM Eric Changed from a fixed max antenna number (28) to a parameter and set that to 29. Introduced a name for 29 as the Pie Town VLBA/VLA antenna. Moved nowhere. 10473. January 10, 2000 UVFIX Leonia Different scaling at the N-S and W-E directions has been implemented. The difference is seen at the output of JPREC routine but can not be explained by defferential abberation. The difference is of order 10^-5 and gives an error of 6msec at the edge of the field 10 argmin. The effect is found by Miller group. Moved nowhere. 10474. January 10, 2000 MWFLT Eric Corrected 2 dimension statements in the code that does 2-dimensional median filtering. Certain counters needed to be max_win**2 in size. The 'ROW' filtering did not handle defaults properly and did a divide by zero. Moved nowhere (except U MN). 10475. January 14, 2000 $YPGNOT/FGPLT.FOR Leonia The new task FGPLT plots the flagged time intervals from the given FG table. The standard parameters (TIMERANG, STOKES, FREQID, ANTENNAS, BASELINE, BIF, EIF,..) are used to select the data. Each horizontal direction corresponds to one of the selected baseline/IF. Each new baseline is labeled. The new IF is not. The more IFs is plotted the less is the space. So the vertical size of the plot is constant for the same number of baselines. If the page vertical size is not enough to plot all selected baselines/IFs the plot continues on the next page. Timerange can be splited on NCOUNT equal time intervals to have better resolution. So NCOUNT plots will be plotted. Each plot plots the consecutive time interval of the length: TIMERANG/NCOUNT. Moved nowhere. 10476. January 16, 2000 FITLD Eric Fixed a string used in an error message. It was declared *8 and DATAed to things 9 or 10 long. Moved nowhere. 10477. January 16, 2000 Number Clean Boxes Eric Defining Clean boxes as simple Fortran arrays (4,mxnbox,maxfld) makes each array take 2 Megabytes of memory and yet limits a user to 256 boxes even if he images only 1 field. Changed to do (4,mxnbfl) with mxnbfl=64K. PUVD.INC Changed MXNBOX to 2048 - max boxes in any one field, added MXNBFL = 256 * 512. GRSLIC Remove unused ref to PCLN.INC WINDF Make WIN be (4,NFIELD,*), compute the LIMIT allowed (min (2048, MXNBFL/NFIELD)). WRIBOX Dropped MXNBOX from call seqence, rearranged axes of WIN. TVUTIL OTVBOX: change order of window, DIM(1) = 4*NFIELD, DIM(2) = LIMIT assumed. QCLEAN Changed handling of WIN and its dimensions. IMAGR Put Win in its own common, reducing memory from 2 to 1 set. rearranged axis order in WIN (makes copy to resolutions messier). SCMAP Simplified it some - the multi-field preparations in it are not needed since that was put in a new task. The window handling is smaller and simpler than the others since there is only 1 field. SCIMG Put WIN in its own common to reduce from 3 to 1 copies in core. Changed axis order, storage declaration, SCHECK call seq to pass MFIELD. WFCLN Changed dimensionality of WIN, altered axis order in WINDF subroutine. SNSMOO Was confused when the refant was 0 => data flagged and so quit rather than handling as flagged. IMAGR Help: changed limit on number of Clean boxes. FILEBOX Help: changed limit to 2048 etc Moved nowhere. 10478. January 16, 2000 FRING, KRING Eric Added CMETHOD to the inputs of both tasks. The DFT method is so slow that the minor inaccuracy of the gridded method is tolerable in many cases. Let's let the user choose. Moved nowhere. 10479. January 19, 2000 AIPS.HLP,AIPS.l Pat Provided more specific example of how to use the Secure Shell (ssh) for remote starting of TV's: set AIPSREMOTE to "ssh -n" and it mostly works. Need to work on some confusing "waiting on clients to finish" messages... A week or so ago I also updated these files to document the debug options (aips debug[=[debug-program]:aips]). Moved nowhere. 10480. January 19, 2000 CookBook Eric Update to 310DEC99 and change psfig to resizebox and includegraphics for LaTeX2e. LaCook1 Changed the line counts, parenthetical remarks, edition number, and other minor things. LaCook2 Update the man page entirely, add information about true color TVs LaCook3 Changed terminal width information, file class names, added TGINDEX and VPUT/VGET/VGINDEX, added info to UVCOP about applying a flag table, mentioned XHELP, added TAPES and FITAB mentions, updated WWW references for FITS, mentioned scp and gzip. LaCook4 Changed RESTORE 0 to DEFAULT, NCOUNT to NPRINT, minor rewording reflecting CROWDED in EDITR, added CLIPM reference, added ANTUSE and CROWDED in description of EDITA, changed to new EDITA plot and add references to NEXT CORRELATOR etc., NCOUNT -> NPLOTS, added references to FITAB, changed IMAGR to use DEFAULT and explicitly show NGAUSS=0. LaCook5 64 -> 512 fields, list overlap/SDI/multi-res as reasons for IMAGR, added use of MAPPR, RESTORE 0 -> DEFAULT, 8192 -> 16384, class names, add info about using simple shifts these days, changed FILLM weighting to disparage using Tsys from tape, CPARM -> IMAGRPRM, add info re deletion of workfile, added ALLOKAY ref, added ref to SETFC and FLATN, added OBOXFILE, add SELECT NEW FIELD, FORCE A FIELD, THIS IS FIELD n, added section on new algorithms (filter, SDI, Multi-res), expand on NOISE in VTESS, added CROWDED, NEXT CORRELATOR, FLAG QUICKLY to EDITR discussion. Mention SCIMG along with SCMAP. LaCook6 Add footnote showing how lower case may be entered in AIPS strings, added PRINTER verb description and info re responding to xterm width and height, *COUNT -> *PRINT, drop ASCAL things, add INVER to LWPLA and info re multiple plots to one file, drop VBPLT ref, mentioned SNPLT enhancements, add SQ enhancements,. mention TVROAM, upgraded discussion of XAS to mention 24-bit versions and verbs TVHUEINT and TV3COLOR, added PLOTR. Remove most references to film recorders. LaCook7 Typo fix and figures. LaCook8 Add mentions of FLGIT, CLIPM, UVMLN, added cross-references in the IMAGR section, modernized some wording, reference full-color workstations. Remove references to film recorders. LaCook9 Minor rev on Solaris large files, add warning re FITAB tables, NCOUNT -> NPLOTS, SNPLT new OPCODEs, SCIMG as well as SCMAP, CC filtering in IMAGR. LaCook10 DOCRT=1, add BSMOD, BSCLN, BSTST, NCOUNT -> ORDER, add discussion to BSAVG LaCook11 Add FITAB mention. LaCook12 Fix up TV connectivity discussion - guest accounts are no big deal now, disk catalogs do not exist, tv=local is good. Cahnged TAPQL to EXTAB, mentioned FITAB and gzip and scp. LaCook13 Replaced all sections with current ones. LaCookA Update figures from psfig. Removed most of $AIPSPUBL/*.NDX, $AIPSPUBL/*.TEX. They are obsolete. Moved nowhere. 10481. January 20, 2000 CALIB Leonia CALIB has used wrong solution interval when the data folow more often than 1sec and SOLINT is selected less than 1sec also to obtain the solution for each time point. I have fixed the problem reducing the magic "1sec" to 0.25sec. Moved nowhere. 10482. January 21, 2000 CALIB Leonia Having thought more about 1sec problem I found that something more should be done: 1.When there is no NX table the LASTIM has been calculated as: LASTIM = CURTIM + SOLINT - 1sec Such a calculation should prevent any solution if SOLINT < 1sec. 2.My previous change 1sec to 0.25s for existing NX can not be enough if the dump time less than 0.25sec. For example VLBA correlator gives dump time ~0.13sec So I have changed 1sec to 0.1sec for both existing and not existing NX table. Actually I'd prefer to remove this 1sec at all. But I am afraid to do it because I do not know all arguments for it. Moved nowhere. 10483. January 21, 2000 BPASS Leonia There has been a misprint which caused printing the same "corr" even for the case of existing many polarizatios. I have fixed it. Moved nowhere. 10484. January 24, 2000 GRBOXS in YSUB Chris Adapted GRBOXS to round to integer using NINT rather than by adding 0.5 and truncating. Although the FORTRAN standard defines NINT(X) to be INT(X + 0.5), NINT(X) is often different from INT(X + 0.5) in critical cases. For example NINT(1.5) = 1 and INT(1.5 + 0.5) = 2. In some cases the tendency of INT(X + 0.5) to round up critical cases was leading to clean boxes "walking" across the field when when repeatedly read from a file and then saved. Moved nowhere. 10485. January 24, 2000 $YPGNOT/FGPLT Leonia Following the recomendations of Eric, Chris, and other people I have made the following modifications of FGPLT: 1. Increased the spacing between the baselines. 2. Plot all baselines and IFs including unflagged ones. 3. Add the space between the baselines if more than one IF is plotted. Moved nowhere. 10486. January 28, 2000 IMAGR Eric When I changed the order of the WINdows array (1/16) I forgot that CLBOX is in a different order. Changed it to put CLBOX values where they belong in WIN. Moved nowhere. 10487. January 29, 2000 UVFIX Eric Changed $APLNOT/LPSEC to use a string that is more Y2K friendly: YYYY-MMM-DD. Changed call to LPSEC in UVFIX and the format. Moved nowhere. 10488. January 29, 2000 CALIB Eric Changed MFIELD to 0 for point model computations to avoid some messages and to avoid some references to CC files. Changed messages to include LIMFLX (FLUX adverb). Clarified FLUX adverb in help file. Moved nowhere. 10489. January 29, 2000 CMPCFL Eric Routine to compress the DFIL.INC commons used heavily by the OOP code caused scratch file and cataloged file entry numbers to change. Many tasks and subroutines assume that they know the index of a file in this common. Changed the subroutine to trim the ends of the lists only. Added a SAVE statement to UVMSUB for the SCRGRD and SCRWRK scratch files. This caused some line cases to fail in IMAGR. Moved nowhere. 10490. January 29, 2000 WHATSNEW Eric Updated for January. Moved nowhere. 10491. January 30, 2000 TIMDAT Eric The short form date is still allowed and used in EXTLIST. The code dealing with it was strange for year. It did a MOD and then tested for <= 0 and > 100 when only 0 was possible and that is not really a bad date. Fixed. Also fixed one call to TIMDAT for REASON in QEDIUTIL. AIPS, IBLED, SPFLG, TVFLG, and the QEDIUTIL ones are all the routines that use this short form from TIMDAT. Moved nowhere. 10492. January 31, 2000 JULDAY Eric Tiring of all the people who go around reading precursor comments and not code, I have fixed the precursor comments to match the Y2K compliant code. Moved nowhere. 10493. January 31, 2000 FITAB Eric Changed comments about FORMAT which really does not apply to UV data. Moved nowhere. 10494. February 1, 2000 IMAGR Eric With blank OUTNAME and IN2NAME parameters, the program's tests for validity of ALLOKAY > 0 did not work properly. Added lots of tests to force the name parameters to be set to allow ALLOKAY. Also fixed help file to explain these limitations. Moved nowhere. 10495. Fenruary 2, 2000 UVFIX Leonia Recentely (see 10473) I implemented different scaling at the N-S and W-E directions. Later I found that found assymetry in the expansion/contraction due to differential abberation is the result of bad test. So I removed the different scaling in U and V directions. Moved nowhere. 10496. February 2, 2000 FITLD Eric For some reason it put weight 0 for CL table entries it created. Then it added real info from various tables without ever messing with the weights! Sigh... Change to put in 1's. Moved nowhere. 10497. February 2, 2000 Rest of CookBook Eric Finish up appendix Z, then the TOC and Index can be done. LaCookZ Change things around dropping IBMs, gripe database, film recorder. LaCookI Updated. LaCook0 Updated. Moved nowhere. 10498. February 4, 2000 Imaging ERic Changed estimation of how long DFT and gridded modeling will take. Changed parameters in QMTYP including changing from linear dependence on size of grid to dependence on the area of the grid. Changed SETGDS to set FLDSZ (if <= 0) so that grid size will be known when CALIB et al make the decision on method. In testing found bugs: IMAGR Changed to suppress message when ALLOKAY or SUBARRAY > 2 doing a test open on the UV work file. Fixed dimension of two parameters which were writing over core potentially when read later. QCLEAN The filtering routine CLFILT wrote outside the APCORE memory area for large images with components near the edges of swaths! Fixed it to convolve only those rows actually needed. Moved QCLEAN fix to patch area. 10499. February 6, 2000 QEDIUTIL Eric The flagging routines were not handling crowded displays with and without ALLPOL and ALLIF properly. Changed a bunch of things including fixing UNDO and REDO flags which did not loop as needed. Limited looping to eliminate flags in the FC table for no data. Made EDFLAG loop suitably for displayed data also. Moved nowhere. 10500. February 8, 2000 FCLEVEL.SH in SYSUNIX Chris FCLEVEL.SH needs to use special processing for Linux. Changed the test to trigger whenever the value of ARCH begins with LINUX rather than requiring an exact match. This may avoid some problems encountered by people trying to port AIPS to LinuxPPC. Moved nowhere. 10501. February 8, 2000 UVCOP, UVFIX Eric In UVCOP, the subarray computation used 1.01 to round up which was not enough for some of the vis recs. Changed to 1.1. In UVFIX, the phase shift vs frequency used channel 1 as the reference channel rather than the correct one. Moved nowhere. 10502. February 12, 2000 Slices Eric The Tektronix emulation is not working well for cursor reading. So changed to do a bunch of functions also on the TV. Changed: AU9B Correct handling TK catalog AU9C Made it loop and forgive errors better POPSDAT Changed TVSLICE to SLICETV, added verbs TVSLICE, TVASLICE, TVMODEL, TVAMODEL, TVRESID, TVARESID, TVGUESS, TVAGUESS, TVSET, TV1SET NEWPARMS Run file - added most but cannot rename TVSLICE, TVSLIN Inits the SL plotting TVVEC Draws a line segment TVLAB Labels SL plots on TV TVSLPL Plots slices, models, residuals, gueses TVSLAC Reads cursor, gets coordinates Changed AIPS, DAIP, AIPSB, AIPSC to know about AU6E. Created new help files. 10503. February 13, 2000 SLFIT Eric Found a nsaty bug - it was using the SL file name from the SL file header even when RECAT had moved (and renamed) the SL file. It seems that this is the only task to fall into this trap (I did a grep for (17) all over). Added an attempt to display the results with errors in internal units and in consistant (and labeled) external units. Moved nowhere. 10504. February 21, 2000 SETFC Eric Found two major bugs: the coordinates for multi-source files were put in the wrong variables and so were not used in setting the field centers. 0,0 was used instead. And they were not handled properly as regards signs. RAs were negative but the sign on 0 hours was lost and the sign on -00 degrees dec would also be lost. Fixed to use > 0 RAs always and to handle the DEC signs correctly. The SUN position was not computed correctly due to the use of a wrong and uninitialized variable pointed out by the DEC compiler. Moved nowhere. 10505. February 21, 2000 EQUINOX Eric EQUINOX has replaced EPOCH in the (proposed) wcs. Added this keyword to DFUV, FVUV, DFIT, and VFIT includes. Moved nowhere. 10506. February 21, 2000 IMAGR Eric IMAGR has found the beam position angle with the wrong sign since 29 December. (CELLSG is all positive in QUVUTIL and has sign in QCLEAN!) Sigh... Fixed QUVUTIL OUBFIT. Messages need to appear. Moved nowhere. 10507. February 21, 2000 UVFIX Leonia It was found recently UVFIX gives a small error in UV calculation which gives an error in the image if it's size very large (~10'). Some people considered that the differential abberation should give unsymmetry contraction in the image while UVFIX has calculated correction for the differential abberation as symmetrical. I have carried out the mathematical analysis of the problem which shows that contraction along the direction of motion is negligible and therefore the differential abberation is symmetrical and the UVFIX error should be in another area. It was found that UVFIX has used wrong picture plane to project baselines. UVFIX has projected baselines on the picture plane corresponded to the non abberated direction of the source. It is figured out now that the abberated direction should be chosen for the projection although the correction of rotation due to abberation was not necessary in the first case. So the picture plane to project baselines has been changed from non abberated to abberated direction. The required correction of the rotation is added. Moved nowhere. 10508. February 22, 2000 SETSLICE Eric SLICETV was a bad name since it blocks saying GO SLICE. Changed POPSDAT, NEWPARMS and the help to rename it SETSLICE. Moved nowhere. 10509. February 22, 2000 POSSM Eric The task used CATBLK even after calls to UVGET('INIT'..). This caused the frequencies and velocities to be wrong for plots 2-NPLOTS on each page whenever BCHAN > 1. Moved nowhere. 10510. February 23, 2000 IMAGR Eric Another bug in the change to fitting each beam separately. In QCLEAN, it did not move the beam parameters of the only beam (when DO3DIM false) to the beam parameters of the other images. As a result the images were restored with width 0 for fields 2-N. Moved nowhere. 10511. February 23, 2000 UVLOD, FITLD Eric I broke these for regular UV loading with the EQUINOX change. The parsers pretend to be list and symbol driven, but actually use absolute subscripts into that list. No longer - made it test strings more and parse out which PSCALn it is and so forth. Moved nowhere. 10512. February 24, 2000 TVFLG, SPFLG Eric Found that the FC table common was not in the right order and that the record size in the FC table open routine was too small - all for a very long time (since REASON) was added. This interfered with the action of LIST FLAGS (and UNFLAG) at least in TVFLG under Linux. However, under Solaris they worked fine and I suspect most other things worked okay anyway. Moved nowhere. 10513. February 25, 2000 Compilation under Solaris/IA32 Chris Added support for the Portland Group compilers to SYSSOL86:CCOPTS.SH, SYSSOL86:LDOPTS.SH, and SYSUNIX:FDEFAULTS.SH. Also added a version of ZTPWAD.FOR to APLSOL86 that will recognize the deviant end-of-file code from pgf77 as well as the normal negative values. 10514. February 25, 2000 CONFRM Eric Lots of places in AIPS.EXE can ask for confirmation with this subroutine. It was not free format but required the user to type in column 1. Fixed that with CHTRIM. Moved nowhere. 10515. February 25, 2000 SETFC Eric SetFC used a linear formula for the field centers which really does not work well especially near the pole. Rick proposes a -SIN projection like formula which is fine up to angles of one radian. That should be enough for the "primary beam" fly's eye. Added comment on that to help file. Moved nowhere. 10516. February 25, 2000 POSSM Eric Fixed a bad error message. Removed an excessive checking of IF numbers which used the UVGET output header rather than the input header in the Main after the POSSIN had already set them. Moved nowhere. 10517. February 29, 2000 $YPGNOT/CONFI Leonia Recentely the new mask of topography of the ALMA's site was produced. This mask has better resolution (4x4 times) and larger area (10 times). CONFI has not been able to handle such a big size file. So I have modified CONFI to use the new topography mask. Additionaly I changed the default of many input parameters to match them with the current vision of ALMA. The help file has been updated. Moved nowhere. 10518. February 29, 2000 $SYSUNIX/START_AIPS Pat Corrected an inconsistency in the use of /tmp and the $TMPDIR variable that was causing the data area listing to show up twice in the AIPS startup. This only affected users who have TMPDIR set in their environments (like me!). Moved to AIPS_ROOT in CV this date (both kochab and elsewhere). 10519. February 29, 2000 DBCON Eric 1. Changed code and help to remove the claim that it did frequency independent phase shifts unless DOPOS was 2.0. It did the right phase shift but would refuse if the two files were not in the same sort order and DOPOS>1. 2. Changed it to examine the FQ tables and renumber the FQs if needed in output UV data, and output FQ, CL, and FG tables. Moved nowhere. 10520. March 2, 2000 EDIT* Eric Changed QEDIUTIL 1. to start ALL-IF and ALL-POL flags as EXPERT .AND. .NOT CROWDED. Changed EDITRm EDITA, SNEDT help files to reflect this. 2. to require point flagging to be close to the point (< 2 pixels); 5 was allowed previously. 3. to loop for more points at the same pixel in crowded displays in POINT and QUICKLY modes. 4. to issue instructions when needed but not too often. The 3rd point was confusing: a user would flag a point and it would appear marked bad. But then on a replot it would return. This is because only 1 correlator was flagged when more than one appeared in the same TV pixel. The flagging marked the pixel in the other color, but the replot marked also the remaining good points. Now all correlators should be deleted and perhaps even two time's worth (point 2 above). Moved nowhere. 10521. March 6, 2000 SNPLT Leonia SNPLT has not worked if the weight is equal to zero for the whole CL table. The weight is compared with CUTOFF to select the data. But CUTOFF could not be negative because the negative input has been converted to zero by the card: CUTOFF = MAX (0.0, CUTOFF) To get passing data with zero weigt allowing negative CUTOFF I have commented this card. This card was in contradiction with the statement in HELP file: CUTOFF....Table weights <= CUTOFF will not be plotted, <0=>plot all ^^^^^^^^^^^^ Moved nowhere. 10522. March 7, 2000 DBCON Eric DBCON did not scale the u,v,w's for any frequency difference between the two data sets! This is okay, so long as they are not multi-channel. The error message about frequencies not matching was wrong. It came out when it should not rather than when it should. Moved nowhere. 10523. March 7, 2000 GRDSET Eric The subroutine that sets the SC file sizes used in CC modeling did not use the correct estimates of sizes. Added a second size estimate and some rounding up and then changed to use the max of the 2 size estimates. One of the files is (2*NY,NX/2+1) in size and that can be larger than NX,NY even with some fiddles already in the routine. We can afford files that are too large and cannot affford ones that are too small. Moved nowhere. 10524. March 7, 2000 SNPLT Eric This task should not plot bad/flagged data. I do not know why it ever has - the help file rather than the code needed changing in Leonia's recent "fix". Moved nowhere. 10525. March 10, 2000 TVFLG Eric Made it smarter about asking for IF and channel numbers - if there are only two switch them; ask only if > 2. Moved nowhere. 10526. March 14, 2000 TVFLG Eric In working on SPFLG, found a bug relating to the storage of Stokes in the FC file. It caused e problem with later REDO operations. Moved nowhere. 10527. March 14, 2000 SPFLG Eric Changed it to plot all frequencies and all IFs on the horizontal axis. Axes 3 and 4 are for baselines number and stokes only - no need to switch IFs. This makes for some interesting changes in how certain flagging operations are displayed and then preformed. Among changes: 1. BIF, EIF, BCHAN, ECHAN are now keywords in the SPFLGR header for cataloged work files. 2. Changed displays for flagging time and time-range to take into account the ALLIF flag. 3. ALLIF is ignored on area flags if the area covers > 1 IF. Otherwise it is applied. 4. An error recording the Stokes displayed when flagging caused some REDOs of CLIPs to fail. 5. An error applying the various DO-baseline combinations could cause the data not to be flaggedin the work file. 6. An error using a constant for an in/out variable was found. Help file reworded as well. Moved nowhere. 10528. March 14, 2000 UJOIN Eric Corrected a bug in the application of DOWEIGHT 2. It used only the flags applying to the highest included IF instead of all IFs. Now it correctly flags a whole spectrum if any channels (in that Stokes) are flagged. Moved nowhere. 10529. March 16, 2000 DBCON Eric For 2 single channel data sets the u,v,w are already in wavelengths at whatever freq they were actually observed. One does not want to change the u,v,w therefore - making the fix of March 7 partly wrong. Backed out the extra mutiplication. Moved nowhere. 10530. March 20, 2000 IRING Eric Changed: 1. The X axis increment should be used with correct sign 2. The angle check for allowed angles for summing had position angle on the X axis not to the North. 3. The angle for the major axis of the ring was similarly measured without regard for North. The help file wrongly stated that CPARM(6) was used for plotting. Also added remarks about East from North. Moved nowhere. 10531. March 21, 2000 TABED Eric Corrected examples in TABED explain section. The file formats of SN and FG tables have changed since this was written. Moved nowhere. 10532. March 21, 2000 DBCON Eric The recent change to handle FQ numbers properly put a road block into the attempt to combine data with different bandwidths as if they were the same. DBCON now - rightly - makes them into 2 FQs. Added FQTOL to adverbs to allow the task to ignore FQ when told to (-1) or to use FQTOL (kHz) to decide whether FQs are the same. Moved nowhere. 10533. March 21, 2000 HLPTVFLG, HLPSPFLG Eric Upgraded the on-line help files for changes made previously. Moved nowhere. 10534. March 22, 2000 UJOIN Eric Changed the meaning of DOWEIGHT to make the flag a whole spectrum if any flagged the default. Added a display of which input channels contribute to which output and a test on frequencies (warning only) to make sure that things agree ok. Changed help file for DOWEIGHT, to explain CHANSEL and OPCODE=DIFF better. Added an explain section by Michael Rupen. Moved nowhere. 10535. March 23, 2000 IMAGR help Eric Modified some wordings based on understanding difficulties encountered by some users. Moved nowhere. 10536. March 24, 2000 $YPGNOT/CONFI Leonia CONFI has been able to optimize an array configuration having a group of antennas unmovable. This option has been worked only for two circle constraint and only without topography constraint. The number and list of the fixed antennas have been taken from IN2FILE. I have generelized this option. Now the user can determine the list of the fixed antennas as first DPARM(2) lines at the INFILE. The posible initial model as a circle with progressive spacing is eliminated. The help file has been modified in particular to clarify the using of IN2FILE as topography mask and as a list of the fixed antennas. Moved nowhere. 10537. March 25, 2000 FILLM Ketan/Eric Changed the option that allowed the weights to be set by the system temperatures to use the nominal sensitivities instead. The former were well calibrated only for some frequency at some time in the past and hence not correct in the resulting weights. The latter should be correct since they are what was also applied to the visibilities. Also help fileto describe and recommend this option. In time, I will switch it to use this by default. Moved nowhere. 10538. March 25, 2000 DOCALIB = 2 Eric Changed all the following tasks to allow DOCALIB=2 => to calibrate the weights, DOCALIB = 1 to NOT calibrate the weights. Previously this choice was up to the programmer and DOWTCL was true only in FRING, KRING, VPLOT, and a few other VLB tasks. Also all help files. APGNOT/PRTSD SDCAL SELSD QPGNOT/SDIMG QPGOOP/SDGRD BSGRD APGNOT/ACFIT BLAVG CLIPM DECOR FLGIT LISTR SHOUV APGNOT/SPCAL SPLAT SPLIT SWPOL UV2TB UVFIT UVFND APGNOT/UVHOL UVIMG UVMLN UVPRM UVPRT QPGNOT/BLCAL BPASS CALIB CPASS FRING HORUS KRING QPGNOT/LPCAL PCAL YPGNOT/CLPLT FRMAP FRPLT IBLED POSSM SPFLG TVFLG YPGNOT/UVPLT VPLOT APGOOP/MAPBM RFI QPGOOP/BLING EDITR GPHAS IMAGR OMFIT SCIMG SCMAP Other changes: SPCAL allow calibrate single-source files, Also BLCAL, LPCAL, PCAL The help filesfor these claimed that they would always calibrate multi-source files. What was true was that they would never calibrate single-source files. SWPOL DOCAL was true only if DOCALIB > 0.5 despite claims of the help file. BLCAL Always applied calibration to multi-source files even if DOCAL was false. Moved nowhere. 10539. March 28, 2000 $QPGNOT/UVCON.FOR, HLP Leonia Modified to calculate the visibilities noise and weight depending on the antennas elevation. The noise temperature of the atmosphere at zenith is added as a parameter. Moved nowhere. 10540. March 28, 2000 Menu title Eric Added to TVMENU the option to display a title line for the menu which could be used as a status line for example. ' ' means do not display or make space for a title. Changed TVDEVICE (TVDMEN) to pass this parameter through. Changed QCLEAN to display the field number, shift, cleaned flux and iterations. Also changed SCIMG to display field number and shift and fixed bug with DOCAL handling. Changed TVHUI, TVRGB, QEDIUTIL, PLAYR, SCMAP to not use this option. Fixed DOCAL bug in SCMAP as well. Moved nowhere. 10541. March 28, 2000 AU8A Eric On 9/2/96, ROTATE was added to the inputs of PCNTR. EXTLIST was never corrected for this - fixed pointers and added display of rotate. Moved nowhere. 10542. April 5, 2000 ZDELA2 in APLSGI Chris Copied the APLUNIX version of ZDELA2, which uses nanosleep(), to APLSGI to over-ride the APLBELL version, which uses the BSD interval timer. This might help with a recently-observed problem where AIPS processes occasionally abort after receiving an alarm signal. Only relinked AIPPGM programs for now. Should probably do this for Solaris and Linux too. Moved nowhere. 10543. April 6, 2000 CLIPM Eric It did not get PFLAGS right when there is only LL data. Moved nowhere. 10544. April 6, 2000 calibration package Eric The cal package tries to go on when there is no NX table - a normal situation with single-source files and trouble with multi-source files. The common parameter CURSOU which identifies the source in the current scan is zero always when there is no NX table. Changed: UVGET to issue a warning on no NX with multi-source files DATFLG to allow CURSOU=0 to mean a flag record for any specific source applies to the current data DATGET to allow CURSOU=0 to allow any source even when source names are specified Moved nowhere. 10545. April 6, 2000 FGSEL Eric FGSEL did not correctly delete records that fell outside the selected range of IFs and channels. Affected UVCOP, FLGIT, SPLAT, UJOIN. Moved nowhere. 10546. April 6, 2000 EDIT class Eric Changed EDITR, EDITA, SNEDT to call the Class function with KILL on actual error and ABOR on user requested abort. Changed QEDIUTIL to make KILL keep the FC table while ABOR deletes it. Corrected EDITUV - it did not reset the SHOW ALSO choice under all circumstances that needed it reset. Moved nowhere. 10547. April 6, 2000 PUTTHEAD Eric Changed this verb to allow adding keywords to a table header (so long as there is space). This takes the adverb KEYTYPE in addition to those already used. Changed help accordingly. Moved nowhere. 10548. April 7, 2000 CCEDT Eric The memory sent in to the editing subroutine is now dynamically allocated. Therefore, it is not zeroed in advance and must be zeroed. The task depended on it being 0 initially which is not right even for normal Fortran memory. Moved some heavy work to be done only when needed and added tests for failure of the algorithm. Previously, on failure it wrote all over itself and really went into the weeds. Moved nowhere. 10549. April 10, 2000 FILLM, PRTUV Eric Changed format of weight scaling message in PRTUV to allow wider range. Changed scaling of weights - I think the bandwidth in MHz not Hz is appropriate. Moved nowhere. 10550. April 16, 2000 ICHANSEL Eric The adverb CHANSEL(3,10) is used mostly to define ranges of channels to average into a single continuum channel or to be used for a baseline fit. These uses allow for different ranges for different IFs, but there is no way in the current usage to specify that. Therefore, added new adverb ICHANSEL(4,20) to be like CHANSEL but use ICHANSEL(4,i) to be the IF with 0 => all IFs. Changed: POPSDAT Added new adverb NEWPARMS Added new adverb ICHANSEL New adverb help file WANTCH Changed to loop to 20 and to ignore ICHANSELs that are 0 (not all 20 are used usually - they must be zeroed in advance). AVGCHN CHANSEL is now IF dependent on 3rd argument, lets CHWT be scalar rather than a large array. AVGCIF CHANSEL is now IF dependent on 3rd argument, lets CHWT be scalar rather than a large array. AVSPC Average and count of channels as well as CHANSEL are now IF dependent. Changed entire pick up of CHANSEL, which was rather confused, to quit when the 1st ICHANSEL(2,i) is <= 0 and to put then in the correct IF(s). Changed finding of average channel to use WANTCH to avoid overlapped windows and the like. If an IF got no window from ICHANSEL, then it takes the center 75%. Fixed AVOPTION handling - anything other than AVIF and SUBS should be normal spectral averaging, not just blank. It was inconsistent on this point. Fixed it to correct frequencies in FQ table when the average channel and/or count of channels in an IF differs from that in the first IF. BPASS Changed dimension and pickup methods for CHANSEL, fixed history display, default handled at pickup now. CPASS Changed dimension and pickup methods for CHANSEL, fixed history display, default handled at pickup now. PHSRF Changed dimension and pickup methods for CHANSEL, added history display. SPLAT Restructured to drop REMOVED random parameters. Change computation of average channel to be IF-dependent and use that to correct FQ output table. Removed some odd stuff with compressed output, rearranged the buffer usage to simplify I/O and averaging, Handle the new ICHANSEL adverb. SPLIT Pick up new ICHANSEL, change call sequences to averaging routines, use average channel out to correct the output FQ table entries. UVLSF Pick up new ICHANSEL, use mask by MAXCIF to avoid something really gross. UVLSD As UVLSF, plus add adverb ORDER. UVLIN Changed NBOXES and BOX, as well as the INFILE option, to support specification of IF. Changed the DOOUTP option to PRTLEV and generalized it to handle any array rather than just the VLA (!). UVMLN Changed NBOXES and BOX to ICHANSEL and added ORDER adverb. FLGIT Changed NBOXES and BOX to ICHANSEL. and all the corresponding help files. Moved nowhere. 10551. April 17, 2000 Update ABOUT files Eric I ran the various test tools and corrected helo files LSAPROPO and ZZADVERB, ZZANALYSIS, ZZCALIBRAT, ZZEDITING, ZZGENERAL, ZZINTERACT, ZZMODELING, ZZONED, ZZPLOT, ZZSPECTRAL, ZZTASK, ZZTV, ZZUV, ZZVERB. Moved nowhere. 10552. April 18, 2000 FILLM Eric In the SOLAR mode subroutine, there were 2 arrays used to get the right VLA IFs together with the right AIPS polarizations and IFs. These arrays were wrong - nominal sensitivity for RL was applied to LL, etc. This error was carried to the MCWAIT routine that applied the Tsys and now nominal sensitivity to weights. 1,1,3,3, and 5,7,5,7 bacem 1,3,1,3 and 5,7,7,5. Moved nowhere. 10553. April 19, 2000 CookBook Eric Updated WHATSNEW.HLP and CookBook chapters 4, 6, 7, 8, 9, 10, and 13 plus the consequent changes to the Index and table of contents. TVSLICE et al, DBCON, SPFLG, UJOIN, FILLM and DOCAL=2, ICHANSEL were the causes of the changes. Moved nowhere. 10554. April 19, 2000 SETFC Eric The NVSS files we made were for Epoch 2000, not too useful for people who have old data or who still use 1950. Wrote a program to convert the 2000 components to 1950 (from EPOSW code) and created files $AIPSTARS/xxxMJY.NV50 with xxx = 1000, 300, 100, and 50. Revised SETFC help and fortran to select the default file based on the header epoch. Moved nowhere. 10555. April 20, 2000 CATLST Eric The buffers used to hold file names for sorting were limited to 512 entries. Changed this to 3*512 = 1536. Moved nowhere. 10556. April 20, 2000 NEWPARMS Eric Although POPSDAT is the important place for new adverbs, NEWPARMS is still used and new adverbs should be placed there as well. ANTUSE was omitted. Moved nowhere. 10557. April 25, 2000 SNPLT, UVPLT Eric Changed SNPLT to also allow parallactic angle as an XAXIS. Corrected Stokes labeling and X axis labeling in SUM mode. Changed UVPLT to plot data weights, elevation, hour angle, and parallactic angle. Corrected UV buffer size, changed fixed scale plot test (not 0 means that 0 cannot be used as a plot limit!) to max > min => fixed scale, added SOUFIL call so that source names will be right on fixed scale plots. 10558, April 26, 2000 UVPLT eric I must not have tested one of my changes, although I swear I did. The call to SOUFIL can only work if UVIIT has set things up. Changed it to call and use SCAL even when fixed scale, just not read all the way through the file. This inits other things not previously initialized (except by assuming they are 0 to start) and checks the name parameters etc early in the process. Moved to AOC this date. 10559. April 27, 2000 CL table weights Eric Weights in cal tables were being ignored in part. This may explain some difficulties with SNEDT which assumed otherwise. BPASET BP averaging used weights without checking for negative values. Set w = max (0, w), then average. BPGET BP fetching did not allow for the case where the two weights were both <= 0. Added handling of that case - data will be flagged, before the 1st BP was used even if it was bad. CSLGET Pickup of SN and CL table data ignored the wight columns entirely. Changed it to blank CALTAB entries when the weight is <= 0. MK3IN Put values for phases in the CL table and left all weights = 0. Changed it to start with all weights 1. M3TAR Same as MK3IN. Moved nowhere, 10560. April 27, 2000 DBCON Eric DBCON attempted to copy CL and FG tables with translation of FQ and source numbers and times. It failed if only one of the two files had a file of this type. It also did not fix the subarray and did not use the correct time offset. Added a TABCOP to fix the missing file problem and fixed the calling subroutine to send in correct time offset and a new subarray offset and fixed DBCAPP to do the subarray offset. Moved nowhere. 10561. April 27, 2000 DBCON Eric DBCON also had an error when combining LL data with an rr/ll data set and writing the result out in compressed form. The output (compressed) LREC was used to zero the accumulation buffer not the LREC prior to compression. This does not matter unless some input correlators are not present so that values are not filled in. Moved nowhere. 10562. April 29, 2000 $YPGNOT/SNPLT.FOR Leonia SNPLT has labeled different IF's plots starting with 1, not with BIF as it should be. It is fixed. Moved nowhere. 10563. May 1, 2000 APCLN, SDCLN Eric These two centered their TV display on the first Clean box rather than the center of the image. Changed them to center on the image. Moved nowhere. 10564. May 2, 2000 VPLOT Leonia When min/max bars is requested, VPLOT and one of the input data points is zero and flagged (re=im=wt=0), the bar has extended down to zero. It is fixed. Moved nowhere. 10565. May 2, 2000 FLGIT Eric Corrected a mistake and a minor issue: a. The flags on deviation from n * rms (APARM(4) and APARM(5)) did their tests on amplitude squared but the APARM values did not get squared. Thus the 5 sigma test was real a sqrt(5) sigma test. Too many channels were flagged. b. The VPOL test was done at the RR polarization. It should be done after the excess flux test on the LL polarization and only if RR is not flagged. Changed the test to check for flagged and to do it on the second polarization looking back rather than the reverse. Moved nowhere. 10566. May 3, 2000 Plotting Eric Discovered that TICINC does not return an error code when it cannot find a plot interval - so things go on to plot a messed up set of parameters. Changed TICINC To return -1 error code when an interval cannot be found and to try harder at both ends of the range. QEDIUTIL To return error -1 when EDTINC fails, but to avoid dying in the end. TKTICS To trap TICINC error < 0 to skip the ticks but not die (i.e. return IERR 0). CTICS To trap TICINC error < 0 to skip the ticks but not die (i.e. return IERR 0). Use CORDDS not COORDD. ITICS To trap TICINC error < 0 to skip the ticks but not die (i.e. return IERR 0). Use CORDDS not COORDD. CORDDS Does COORDD but allowing 3 digits for day number. SNPLT To handle missing antenna file by allowing up to 90 antennas and taking IAT as the default time system. Moved nowhere. 10567. May 8, 2000 KRING Chris Added a trap for the case where there is no data from which to derive a solution. Previously, this case would lead to an abort after a division by zero. Moved nowhere. 10568. May 9, 2000 FRMAP Leonia Today is the victory day, day of end of the world war II in Europe. Congratulation!!!!! FRMAP has not been able to print the table if PRTLEV.NE.0 and number of lines crossing the given cell more than 60. I added the relevant warning. Moved nowhere. 10569. May 9, 2000 BPLOT Eric Corrected a guaranteed zero divide when plotting the phase of the reference antenna. Moved nowhere. 10570. May 9, 2000 UVPLT Eric If all antenna files were missing it would plot nothing since it loops over subarray number. Fixed to do at least 1. Moved nowhere. 10571. May 10, 2000 UVFIX leonia/Eric The change to J2000 coordinates had an error causing the task to precess the first subarray but not the others. There was a longer standing problem of using the time of the first sample for things even if that sample was not in the current subarray. Corrected both. Moved nowhere. 10572. May 10, 2000 Y2K Eric Checked in $RUNSYS/Y2KLOAD.001 and Y2KEXEC.001 which are new versions of DDT suitable for modern computers. Also the help files Y2K.HLP and Y2KSACE.HLP to support them. This is still under development but I need the procs to be distributed. Moved nowhere. 10573. May 12, 2000 FILLM Chris Added checks for frqeuency changes. Without these checks, data was sometimes directed to incorrect files when there were setups using different AC and BD IF frequency settings and setups using the same AC and BD IF settings for the same correlator mode in an experiment. Also revised the main loop over files to prevent FILLM terminating early if there is no selected data in the first 2 files examined. This modification has the unfortunate consequence that many error messages will be generated on hitting the end of tape whan NCOUNT is large since the MCREAD subroutine does not distinguish between read errors and end of file conditions. Moved nowhere. 10574. May 16, 2000 FILLM Eric Chris's recent changes exposed a gotcha when handling the default value for NCOUNT (1.0e10 - a bad integer). Also changed it to honor FBLANK when put in the nominal sensitivity, avoiding use of this special data and flagging any data affected. In the one test so far, the data affected were either not present or fully flagged anyway. Moved nowhere. 10575. May 23, 2000 interferometer model table Chris Added routines IMINIT and TABIM to APLNOT. These initialize and read/write/close interferometer model (IM) tables using argument lists instead of common blocks in INCS:DIMV.INC. Moved nowhere. 10576. May 23, 2000 TABLEUTIL in APLOOP Chris Added new table access routines. OCQINI - initialize CQ table OTABCQ - access CQ table OGCINI - initialize gain curve table OTABGC - access gain curve table OIMINT - initialize interferometer model table (without INCS:DIMV.INC) OTABIM - access interferometer table OTABMC - access CALC model components table Moved nowhere. 10575. May 23, 2000 FRING Leonia FRING has given a wrong message about nonuniform frequencies because the tolerance (1.E-7) has been given in Hz. I have changed it calculating the tolence in Hz as a product of 1.E-7*DF; DF is the found minimum spacing between the frequencies. Inside of FRNMOD I added (as it has been done in CALIB) printing of the CC components actually used in the model dividing. Moved nowhere. 10576. May 24, 2000 FXPOL Chris FXPOL has been completely re-written to be able to deal with situations with odd bands LCP and even bands RCP (and stranger cases) as well as the usual VLBA case where odd bands are RCP and even bands LCP. Moved nowhere. 10577. May 26, 2000 $YPGNOT/CONFI Leonia Fitting to topography, CONFI has shown position of 1000 not fitted elements if at least one element appeared outside of the topography mask. I have modified the task terminating it at this case with the relevant message. Moved nowhere. 10578. May 26, 2000 IBLED Eric Corrected examination of antenna numbers: the delete all baselines with one antenna did not work when applying the flags directly to the data. Moved nowhere. 10579. May 26, 2000 KNTR Eric Found 3 bugs: a. The determination of the end point of a line segment uses an interpolation scheme without (previously) checking the result for validity. Values from 0 to 0.5 could be in nasty cases 200 - causing a line to go off to the edge of the plot. A better algorithm is probably called for, but i simply substituted some tests for validity and provided alternative computations of the interpolations. b. Long lines with the pen down were drawn as a lot of short segments, wasting disk space, computation time, etc. Savings of 10-40% in the plot file are realized. c. A format was inadequate for the range of things that might be plotted. Moved nowhere. 10580. May 27, 2000 EXTLIST Eric Changed AU8A for the addition of NY - nuber of Y panes - which moved a bunch of adverbs to 1 higher addresses in the buffer. Moved nowhere. 10581. June 1, 2000 Linux/PPC support Chris Added support for Linux/PPC running on a PowerMac G3 or G4. Added a SYSLINPPC area and an APLPPLIN. Changed FDEFAULT.SH, LIBR, STARTPMON, ZLASCL, and ZLPCL2 in SYSUNIX. Changed INSTEP1 in INSUNIX. Changed Makefile for XAS. Most of the porting work was done by Craig Lewis. Moved nowhere. 10582. June 2, 2000 CDVER.SH/CSH Pat Modified these to reflect local changes (UPDATE location changed slightly; no more direct automounter maps; /home/aips1/...) Moved nowhere. 10583. June 2, 2000 DBCON Eric Changed the test on coordinate rotation so that an angle of 1e-6 and 0 would be regarded as close enough to each other (!) Cleaned up the code some too. Moved nowhere. 10584. June 2, 2000 QUVUTIL Eric QUVUTIL skipped the call to UVWAIT on natural weighting unless there was some powere (other than 1) of the weight taken. It did not check for a UVTAPER (called keyword WTTAPER at this stage). So UVWTFN='NA' or ROBUT=7 were not tapered in IMAGR and elsewhere. Moved nowhere. 10585. June 6, 2000 TABLEUTIL Chris Added wrappers for phase cal (PC) table access routines. Moved nowhere. 10586. June 6, 2000 FXPOL Chris Added conversion routines for phase cal (PC) tables that were unintentionally omiited from the initial version. Moved nowhere. 10587. June 7, 2000 XAS Eric Changed init to detect the number of bits per pixel as well as the depth. Some XWindows use 24 and most 32 for 3x8 bits of color. Rewrote scrdo24 to handle both 24 in 24 and 24 in 32 - before it only did the latter. In testing, discovered an error that caused it not to repaint all pixels in zoom mode with split screen (used only by TVROAM). A long standing error in the 24-bit and 8-bit codes could cause aborts for addressing exceptions by going on too long in zoomed rows. Moved to AOC Linux machines as soon as bits of it worked. 10588. June 8, 2000 POSSM Eric POSSM plotted R bandpass solutions unless you told it 'LL' and plotted the average of R and L on blank (but not I). But it labelled the plot with whatever the user said for STOKES. Changed it to be friendlier about the user's STOKES, to label it meaningfully, and to warn about the averaging in the help file. Moved nowhere. 10589. June 11, 2000 FACES Eric New task: It makes images and CC files for a multi-field "image" using the NVSS as the list of sources. Moved nowhere. 10590. June 12, 2000 VLBAUTIL Chris Added a run file containing a preliminary set of procedures for dealing with VLBA data. The initial set of procedures is: VLBALOAD - run FITLD with default options VLBAMCAL - remove redundant calibration records VLBASUBS - set up subarrays if necessary VLBAFPOL - fix polarization labels if possible or recommend settings for FXPOL VLBAFIX - fix common VLBA problems (VLBAMCAL, VLBASUBS, and VLBAFPOL in sequence). These are primarily aimed at novice users. Moved nowhere. 10591. June 13, 2000 DOCALIB Eric Upgraded the help file for this adverb to tell about DOCAL=2 for weights. Moved nowhere. 10592. June 13, 2000 EDITA Eric The Tsys - plot was labeled as being in Kelvins but was actually 100 * (Tsys/-1). Changed by dropping the 100, changing the label to Ratio rather than K and the option name from TSYS - to TSTS/-1. CHANGED QEDIUTIL, EDITA.HLP, HLPEDITY.HLP. Moved nowhere. 10593. June 14, 2000 $APLNOT/BESSJ0 Leonia The new function BESSJ0 calculates the values of the Bessel function of zero order. The algorithm is taken from: http://gams.nist.gov... after significant simplification. Moved nowhere. 10594. June 14, 2000 ISPEC Eric Changed to have ZINC < 0 => a request to plot the Z axis in the opposite order (thereby allowing velocity to increase to the right in typical cases). Moved nowhere. 10595. June 15, 2000 BLSUM Eric Changed BLSUM to offer the option of writing a text file for use with PLOTR to plot the spectra from up to 5 blotch regions selected in BLSUM. Changed it to plot flux for Clean images and average brightness otherwise. The business of plotting the ratio of the current channel to the blotch image makes little sense to me (now anyway). Moved nowhere. 10596. June 19, 2000 ZVTPO3, ZVTPGC Pat A brief (and incomplete) security audit of TPMON found a couple of strcpy calls in these routines that were potential buffer overflows. As TPMON acts as a service accepting connections from potentially anywhere, such unsafe strcpy could potentially yield a shell to a "hostile". Replaced these with strncpy calls which prevent buffer overflows. Moved nowhere. 10597. June 20, 2000 CALIB, FRING Eric Both of these rounded the scan time range outward by a bit and both comments were wrong about the size of the rounding. CALIB expanded by 4.32 sec in each direction! With 3 sec between scans this can lead to very bad decisions about what to include and not include in phase switching experiments. FRING expanded by 1 second. Change both to expand by 0.1 sec - the accuracy of a time at 10 days. Moved nowhere. 10598. June 20, 2000 OHGEO, HGEOM Eric Changed the handling of the reference pixel in both cases. In HGEOM, allow the user APARM(3 and 4), and in OHGEO AXREF and AX2REF, to specify the desired ref pixel. Default is the 2nd image ref pixel without regard for output image size. Corrected OHGEO which had very bad methods for setting the ref pixel and which got most of its header from the 2nd image incl frequency, stokes, and more. HGEOM had a wierd break from a sensible ref pixel if IMSIZE was large enough to a center of field when it was not. Moved nowhere. 10599. June 21, 2000 QEDIUTIL Eric The history file was not written correctly due to the use of the wrong variable name. The column labels were repeated rather than showing the information. Moved nowhere. 10600. June 24, 2000 NVSS lists Eric Changed the format of the NVSS lists to include one more decimal place for the fluxes - i.e. 0.1 mJy. Replaced the $AIPSTARS files for 1000, 300, 100, and 50 mJy for both 1950 and 2000 epochs. Also created very much larger files for 5 and 1 mJy which will be available from the WWW but not put in the standard distribution. Changed: SETFC Changed the reading of the NVSS lists and flux computation. FLATN Corrected handling of scratch image headers to block out unused axes - OOP has strange ways of doing things that force NAXIS >= 1 but then make NDIM = 7. FACES Changed the NVSS format, improved beam model incl new 74 MHz model. Moved nowhere. 10601. June 24, 2000 $QPGNOT/UVCON.FOR, HLP Leonia The new option: pointing error has been added. The three type of the primary beam are offered: 1. circular dish with the flat illumination (VLA) 2. illumination is 10db down at the dish edge 3. illumination is 15db down at the dish edge The three component of the pointing error are offered: 1.global pointing error, constant in time for all antennas. This error immitates the error of the source coordinate. 2.random among all antennas, but constant in time. 3.random among all antennas and in time. The task can use the model given as CC component using DFT method for the visibility calculation. Moved nowhere. 10602. June 26, 2000 SETFC, FACES Eric The ACOS arguments can sometines get too big - corrected all ACOS uses in these 2 to test for argument size. Moved nowhere. 10603. June 26, 2000 WENSS Eric Added formatted WENSS files for J2000 and B1950 to the WWW STARS area and remarks anbout them to the SETFC and FACES help files. Moved nowhere. 10604. June 28, 2000 SETFC, FACES Eric Added the FLUX adverb to be a cutoff in the product of the source flux and the antenna beam. The file names alone do not provide an adequate flux control. Moved nowhere. 10605. June 28, 2000 VLBAUTIL procedures Chris Fixed a faulty call to the VLBANEW procedure from VLBASUBS. Moved nowhere. 10606. June 29, 2000 TVFLG, SPFLG help Eric The user's value of STOKES does not completely control the default Stokes to be flagged. That default is normally just what is displayed plus relevant crosshands. Corrected the help files on this subject. Moved nowhere. 10607. June 29, 2000 TVLABEL Eric Changed AU5B amd the help files to add GRCHAN as an adverb to TVLABEL and TVWLABEL. Moved nowhere. 10608. July 1, 2000 $APGNOT/MATHS Leonia Added the new option 'ABS' that calculates the map of absolute value of the input image. Moved nowhere. 10609. July 2, 2000 IMAGR improvements Eric In working with IMAGR to correct a bug, it became clear that it is very useful to have NO clean boxes in some fields for a period at the beginning of the Clean, esp in OVERLAP 2 mode. A field is chosen for Cleaning in that mode based on outdated information - in weak fields information that is likely due to sidelobes of the main sources which are Cleaned away before the field is actually chosen. Changed to allow FLDSIZE=-1 to mean no initial Clean box and a size = IMSIZE and to allow the TVBOX option and verb to return 0 boxes (hit button D right off). Changes: IMAGR To have FLDSIZ(*,i) < 0 => no box, size=IMSIZE SCIMG To have FLDSIZ(*,i) < 0 => no box, size=IMSIZE; to declare local includes for TELL separately to avoid unused variables; to correct handling of CLBOX since box 1 for all fields comes ahead of box 2, etc TVMENU Set logical to init TITLE graphics plane from inputs, window size change or previous line longer than current one. Tittles were getting messy. FIELDF Handle round-off for negative FLDSIZE in BOXFILEs. GRBOXS Clear box-drawing graphics when #boxes <= 0; allow NBOX = 0 to be returned. Previously it set NBOX to at least 1 even if no TRC was set. QCLEAN Fixed a bug in 3D Cleaning with OVERLAP<2: it did the histogram only on the first field. This could set a very low flux level for loading to the AP, much too low for larger Clean boxes in later fields. Changed it to do the histogram over all images in this case. Changed it to allow no Clean boxes in a field. Added flag to be set by TELL and the TV saying to QUIT - otherwise the QUIT operation was overridden in some cases and ignored. Added message recommending OVERLAP=2. Added a call to CLNTEL ahead of the Clean loop after images are first made. Added an off scroll for image plane 1 - the scroll was turned off for all other image planes but the one that mattered. Added warning that > MAXPIXEL pixels were loaded to the AP. Helps: IMAGR (fldsize, quick exit from TVBOX) Helps: SCIMG (fldsize, quick exit from TVBOX) Helps: HLPSCMAP (number boxes, quick exit from TVBOX) Helps: HLPCLEAN (number boxes, quick exit from TVBOX) Helps: HLPSCIMG - forgotten real-time help file Moved nowhere. 10610. July 3, 2000 Unused variables Eric Changed 129 subroutines to remove unused variables - most were DATAed things that the Sun compilers do not report. AU2 AU3 AU5A AU5B AU5C AU7 AU7A AU8 AU9A HELPS HIENH HILUT OFMLST PSEUDO RDUSER ANTDAT ANTINI B2JPOS BAKLDS EXTIO LMPIX MAPCOP MAPOPN MAPWIN MDISK MINIT MSGWRT MWFBSM NEWPOS OFMDIR PSFORM SETDO3 TABKEY TAPIO VBOUT ARRAY UVPHAS ZRMTAP ZSSSLK ZSSSOP ZTAPE ZTXREN ZVTPRO ZVTVOP ZVTVRC ZVTVRO ZACTV8 YTVOPN YCREAD DLINTR IMANOT OFMCON OFMGAM TVFLAM TVLOAD TVPSUD TVXPRT ACTFRQ AGINI APLPBI APPLPB ATCONV ATINI BAINI BASAVG BCINI BLINI BLSET BOXBSM BPINI CAINI CALINI CHNDAT CLSORT CMINI CSINI DGETSD FLGINI FNDPOL FNDVAR FQINI FRINI GAINI GCINI GETDEL GETKEY GNFSMO GNSMO GRIDTB HFINI IMCREA IMINI IMREF1 IMSORT INDXIN NDXINI NXTFLG OBINI OTTINI RMEXT SDCSET SNAPP SNINI SNRESC SOINI SOUINI TABAPP TPIOHD WXINI QSPSUB QUVIN QROLL AEGRID GRDCRM GRDSUB UVMDIV UVPDIV UVTBUN UVTBUW Save the COMRPL for the MAXIF change. Moved nowhere. 10611. July 4, 2000 FLGIT, CLIPM, UVLIN, UVMLN Eric Rearranged the reporting of data flagged to use dynamic memory allocation. The forthcoming change in MAXIF was disastrous for these 4, raising memory needs by 16 or 24 Mbytes. Now they will run in a lot less memory in most cases. Moved nowhere. 10612. July 6, 2000 IBLED Eric Rearranged the buffer handling for the calibration and averaging to use dynamic memory. Corrected a hard coded (!!) 28 for the antenna station ID strings. Pie Town caused it to abort. Moved nowhere. 10613. July 6, 2000 APCAL, GETJY Eric Changed GETJY to count sources and then use dynamic memory to allocate the buffers it needs; MAXIF was making this one enormous and it still would only handle 300 sources. Changed APCAL to allow only 10 records per antenna/IF/FQ to describe the gain model in a 2nd dimension. I doubt that > 1 is actually needed at present, so long as MERGECAL is run. It was using 2 * MAXANT (=180) for no good reason at all, making this one of the biggest tasks in the system. Also added option to limit plots to < APARM(5) in sec(z). Moved nowhere. 10614. July 6, 2000 APCAL Chris Removed code that added a scan with 2 system temperature records between every solution interval. Moved nowhere. 10615. July 7, 2000 BLING BLAPP Eric Fixed an error introduced with the DOCALIB=2 change in March: 'Inputs' was misspelled 'Input' causing BLING to die. Changed BLAPP to use dynamic memory for arrays that were previously 2 x MAXIF x MXBASE. Moved nowhere. 10616. July 8, 2000 QEDIUTIL Eric The handling of ALLPOL and ALLIF in EDFLAG was strange - I wonder what I was thinking. Corrected it to handle crowded and uncrowded plots and ALLPOL/ALLIF correctly (I hope). The REDO operation did not init variables it used and otherwise did some odd things. Fixed it to set the parameters in advance, and loop only as needed. UNDO needed a bit of this too. Fixed it to use OBJECT as the default source 1 name. Moved nowhere. 10617. July 8, 2000 FILLM, SPLIT help Eric Added DOWEIGHT to FILLM as a synonym for CPARM(2) = 8. Changed the help file to describe the synonyms. Also did this for DOCAL=2 and the APARM(4) option in SPLIT and SPLAT help files. Changed FILLM to allow the weighting option for any version of the data - the Tant values are only good at rev 25 but the nominal sensitivities were good before then. Moved nowhere. 10618. July 9, 2000 POPS verb Eric Changed POPSDAT.HLP and POPSYM.HLP but NOT NEWPARMS.001 to create a new verb (spelled with a tilde sign). Changed COMPIL to compile it and to correct an error handling the compilation of the = sign in lists. Changed KWICK to call PASSGN and wrote PASSGN, a version of MASSGN to handle the new verb. The tilde verb stores 1 or more values in an array starting at the specified subscript. It does not check the array bounds and so can overflow in the next adverb, whatever that may be. Thus APARM(6) ~ CPARM copies CPARM(1)-CPARM(5) into APARM(6) through APARM(10) and then copies CPARM(6)-CPARM(10) into whatever follows (in this case BPARM(1)-BPARM(5)). This will help a lot with, e.g. RASHIFT, GWIDTH, et all. Thus GWIDTH(1,3) ~ a,b,c sets the width parameters for the 3rd gaussian without changing any other GWIDTH values. The number of values on the right-hand side of the ~ determines how many values on the left are changed. the array is not 0 or blank filled as with the = sign. Moved nowhere. 10619. July 9, 2000 PUVD.INC Eric Changed MAXIF from 28 to 90 to handle interferometers that deal with frequencies in a very general way. Several tasks have already been fixed to use dynamic memory rather than to go up in size by 15-40 Mbytes with this change. Moved nowhere. 10620. July 11, 2000 BOX2CC Eric Corrected BOX2CC.001 to limit NCCBOX to 10 and to set CCBOX to zero first. Created a BOX2CC help file. Fixed a missing comment line in CCEDT.HLP. Moved nowhere. 10621. July 12, 2000 PLAYR Eric Changed order of includes - DATA statements preceded declarations and only HP62 in Bonn told us. Moved nowhere. 10622. July 12, 2000 INDXR Eric The method of computing the antenna and subarray numbers from the baseline code did not allow for numbers epsilon less than ideal and so got wrong answers occasionally. Put in a much less elaborate method that works. Moved nowhere. 10623. July 12, 2000 AFILE Eric AFILE used list-directed READs in several places to read in a variety of variables from internal strings. This is technically illegal in Fortran 77 and the (lesser) HP compiler took today to complain. Changed by adding a routine to use GETNUM to parse the strings and changed the READs to calls to this GETI4. I am hoping that MPI or someone will test this. Moved nowhere. 10624. July 14, 2000 XTRAN Eric Changed to allow DOOUT = -2 to be the same as +2 except no output. DOOUT=-1 matches +1 in its choice of reference pixel etc. Moved nowhere. 10625. July 15, 2000 CALIB Leonia/Eric A logic mistake is found at the ANTUSE determination. As a result only first of listed ANTUSE has been actually selected at a special combination of the ANTUSE list. CALIB could have a problem if number of antennas at the AN table exceeds 50. It has been fixed also. Moved nowhere. 10626. July 17, 2000 $YPGNOT/APCAL.FOR Leonia APCAL has not worked if the TY table lists more sources than SU table. Such a situation takes a place when APCAL is used after UVCOP applied selecting sources. Reading a row of the TY table APCAL has checked the presence of the found source at the SU table and died if the source is not found at the SU table. I have changed the logic here. Now if the source is not found at the SU table, APCAL goes reading the next line at the TY table. Moved nowhere. 10627. July 17, 2000 FRING Eric Changed FRING to use dynamic memory relative to APCORE to see to it that any fringe fit will "fit into" any AP size. Changed $QPSAP and $QSGI versions of QSEARC to take 3 arguments as the 3 offsets needed into the AP. This lets 0 - 3 be to the standard AP memory and the remaining 3 - 0 be to dynamic memory. Added comment to help file. Now sites should consider whether very large AP memory is right for them. FRING previously required it. Moved nowhere. 10628. July 18, 2000 TECOR Chris Fixed two errors in parsing IONEX files, both of which resulted from ambiguous wording in the IONEX definition. 1. The number of maps given in the header is the number of maps of each type not the total number of maps in the file. Assuming the latter may lead to some TEC maps not being read if RMS and HEIGHT maps are interleaved with TEC maps. 2. EXPONENT records can appear outside the header. It appears that where EXPONENT records appear outside the header the new exponent applies to all subsequent data until the end of file or until another EXPONENT record is encountered and is not limited in scope to the enclosing map. This interpretation is based on the sample routines referenced in the IONEX definition document. It does not appear that either error would have caused a problem with any of the data currently archived at CDDIS. Moved nowhere. 10629. July 20, 2000 FRING Eric I had attempted to avoid re-allocating dynamic memory on every scan/pol. But every parameter can vary between scans even the number of rate or delay points so changed it to recompute the pointers and allocate memory if needed in every call to FRNSRC and FRNALL (and deallocate the memory on the way out of course). July 21 - corrected a typo in FRNALL freeing memory. Moved nowhere. 10630. July 21, 2000 PBCALC Eric Broke the primary beam calculation out into a single subroutine to be used by PBCOR, PATGN, SETFC, and FACES. Added Perley's beams for 74MHz and P band and the NVSS beam for L band as well as the old PBCOR beams for higher frequencies. This should still go into the OOP code for IMAGR beam correction.... Moved nowhere. 10631. July 24, 2000 APCAL Leonia APCAL has given the message implemented by Ketan. The message does not have a sense if OPCODE .NE. OPAC. I have left the message only if opacity is uses. Moved nowhere. 10632. July 24, 2000 OHGEO, HGEOM Eric Due to complaints, I have changed the default handling of the new reference pixel adverbs. If they are BOTH zero, then the numeric center of the input image is placed at the numeric center of the output image. Moved nowhere. 10633. July 24, 2000 BPASS.HLP Leonia The default of BPVER (BPVER=0) has been described as BPVER=1 that does not correspond to the actual codes. I changed it: 0 => a new table to be generated. Moved nowhere. 10634. July 25, 2000 IMSEL in APLNOT Chris Re-wrote IMSEL to use IMINIT/TABIM rather than IMINI/IMTAB. This means that the RDATE keyword is no longer omitted from the output table and allows the logic to be simplified since it is no longer necessary to swap table meta-data in and out of common. Moved nowhere. 10635. July 25, 2000 TVLABEL, TVALL Eric Fixed to describe and spell GRCHAN more correctly. Moved nowhere. 10636. July 25, 2000 OHGEO, HGEOM Eric Corrected bug in OHGEO that had the wrong coordinates appearing in the output image - it said the 1st image and should have said those of the 2nd for the first 2 axes. Of course, not long ago it took all axes from the 2nd (e.g. freq, stokes, etc). Also changed the defaults once more: if IMSIZE and the ref-pixel adverbs are ALL 0, the the output size and ref pixel are taken from the 2nd image. Moved nowhere. 10637. July 25, 2000 SETLOC Eric The code trying to recognize coordinates did not do a good job on a legitimate but unrecognized ra/dec pair. Fixed it to still treat them as RA/DEC (or GLON or ELON...) but just do linear arithmetic (i.e. -CAR projection). Moved nowhere. 10638. July 25, 2000 UVHOL Eric Changed help anf fortran to allow BDROP=1 rather than converting that value into 5. 0 still goes to 5. Moved nowhere. 10639. July 26, 2000 IMAGR Eric Changed QCLEAN to sum the restored flux from field n to m and then to report the total clean flux in each field as it exits. Moved nowhere. 10640. July 26, 2000 IMEAN Eric Added another test for the fit of the histogram for rms being in deep trouble. If the fit section does not include at least 10% of the pixels (it should be ~80%) then we have not found the right peak in the histogram or it does not look like a Gaussian. Added a message that all the bail-out tests go to. Moved nowhere. 10641. July 27, 2000 FILLM Chris FILLM does not handle changes in correlator configurations well if they coincide with file boundaries. If FILLM detects an end- of-file condition in an MCRECO READ operation it calls MCRECO with an INIT opcode until it find data that it wants and then call MCRECO with the READ opcode again. If the correlator set-up has changed then the INIT operation detects this and sets a flag indicating that a change of output file is needed but no action is taken regarding the output file since no data has been read. The subsequent READ operation compares the correlator parameters to those it found during the INIT and finds them to be the same and therefore concludes that the output file need not be changed. This results in the first scan from the new file being written to the same output file as the last scan from the previous file when it should be directed to another file. Fudged FILLM so that INIT operations will not overwrite its record of the previous sky frequencies which should prevent most but not all errors of this kind. I suspect that it is unsafe to prevent the update of other remembered data. A better solution would be to determine the output file at the point where the correlator configuration change is detected rather than waiting for data to be written but this would appear to require considerable effort. Moved nowhere. 10642. July 28, 2000 LISTR Leonia LISTR has not understood 'I' as a stokes if OPTYPE='GAIN'. So if the data are 'I' stokes then LISTR considered 2 polarisations and having succesfully finished the first (the only) one died without any explanation because a second polarisation is absent. I have modifyed the routine GANSET adding 'I' in the list of considering stokes and forcing STOKES='I', if ICOR0=1 and if CATBLK indicates the only polarization in the data. Moved nowhere. 10643. July 28, 2000 FITLD Chris While the FITS-IDI definition and the earlier VLBA memo required that epochs be 8-character strings in the SOURCE table the VLBA has been writing the epoch as a double-precision number and FITLD has expected this. FITLD now copes with both the correct form of the SOURCE table and the VLBA variant by checking the dimensionality of the column and assuming the it holds a double-precision number if the dimension is 1 or a character string if the dimension is 8. Moved nowhere. 10644. July 30, 2000 Modeling Eric Changed SETGDS. With the new naming convention, it always numbered the model files from 001 even if there was only 1. This made it impossible to subtract field, say 8, from the uv data. Changed to read IN2CLASS(4:6) == J, and then to number from J to J+NFELD-1. Moved nowhere. 10645. July 31, 2000 FITAB Eric Found an egregious error. the data were opened under LUN 16 and then TABSRT could be called. This uses LUNs 16-25 and routines under it did not care if the LUN was already open in the FTAB, they just kept going. Thus, if one were writing UV data with > 1 piece, the sorting of tables to time order (if needed) would destroy the initial part of the input UV data set. Why has this not been reported - does no one use FITAB? Moved nowhere, lots more to come - should go to patch. 10646. August 1, 2000 UVCOP Leonia I have fixed the bug in UVCOP, that has existed long time. UVCOP has not flagged the data if the selected FREQID>1. UVCOP rewrote the (FREQID) random parameter of a visibility (RPARM(1+ILOCFQ)) to 1(one) independent on the selected FREQID. And this is right. But later inside of routine DATFLG, it compares the FREQID of the FG table with RPARM(1+ILOCFQ), which was recalculated to 1 (one) ealier. It is clear that if FREQID of the FG is not 1, then it does not match with the having selected data and the flagging is missing. I have changed comparison (inside of DATFLG) from RPARM(1+ILOCFQ) to the selected FREQID. Moved nowhere. 10647. August 2, 2000 install.pl Pat Put back a checkpointed version of this script (in $SYSPERL). It's not ready for prime time yet, as I have only tested the menu and input parameter portion of it. Yet to test is the actual meat of the deployment. However, I encourage people to try out the first part and let me know what they think. Moved nowhere. 10648. August 2-3, 2000 ZOPEN Eric When ZOPEN returns an error code 1 (already used in FTAB) this means that some file is open with that LUN. To use that LUN for some other purpose at that point allows that other purpose to overwrite the original file - the users data in FITAB for example. It is never right to mask this error. Changed maksing of this error in: AMERGE EXTINI IMGERR RELPOP TABINI UVPREP UVPROT ALGSUB ONV1 CONV3 GRDMEM UVTBGD PASS1 AHIST BSROT CLIP CORER CORFQ CPYRT DIFRL DIFUV FUDGE HISEQ REMAG SDLSF SDMOD TAFFY TBAVG UVBAS UVCMP UVLSD UVLSF UVMTH UVPOL WTSUM XBASL XGAUS XMOM XPLOT XSMTH XSUM ACCOR AVER AVSPC BASRM BLAVG BLOAT BSFIX BSMAP BSMOD CANDY DAYFX DBCON DECOR DESCM DTCHK FETCH FILLM FILLR FLGIT FXTIM FXVLA GLENS IMFLT IMLIN IMMOD INDXH M3TAR MANDL MK3IN MODVF MULTI MWFLT NINER NNLSQ PATGN PCCOR PHNEG PHSRF RESEQ SBCOR SDTUV SELSD TI2HA UJOIN USUBA UVAVG UVCOP UVCRS UVDGP UVFIL UVFIX UVGLU UVLIN UVMOD UVNOU UVSIM VBCAL VBGLU VBMRG VLBIN WTMOD BLCAL BPASS CALIB CPASS FRING GRIDR LPCAL PCAL SDIMG APCLN APGS APVC SDCLN STEER UTESS UVMAP BLANK BLSUM IMVIM IBLED Moved nowhere - INSTEP4 to be done. 10649. August 6, 2000 IMAGR Eric All values of IMAGRPRM(8) were used in the filtering stage if the initial value was not zero (so that the copy of the original data was made). The help file states that a zero value will mean no filtering at any time. So changed it to test for exactly zero before calling CLFILT. Added emphasis in the help file on this point. Moved nowhere. 10650. August 8, 2000 VLBAUTIL run file Chris Cleaned up procedures defined in VLBAUTIL and added new procedures for splitting out frequency IDs, applying amplitude calibration, and correcting for parallactic angle. Also added some helper procedures and removed VLBAFIX. See the VLBAUTIL help file for more information. Moved nowhere. 10651. August 10, 2000 FITAB, FITTP Eric FITTP had a long-standing failure to pass the error code from PTF3D to higher level routines so that they could quit on tape error while writing tables. This error was copied into FITAB where it is even more important. Moved nowhere. 10652. August 10, 2000 DBCON Eric One of the ZOPEN calls corrected a few days ago had an error test of IRET (set to 1 at that point) while ZOPEN returned IERR. The correction to allow only 0 caused DBCON to have a hard-coded failure. Sigh... Moved nowhere. 10653. August 17, 2000 FITTP Eric The SGI aborted from the use of a real*4 buffer on a real*4 address for a real*8 variable (calling ZRLR64). The buffer has to be forced to a real*8 boundary via equivalence. Moved nowhere. 10654. August 17, 2000 ZWINC2 Eric Created an SGI version of ZWINC2. Oddly, other architectures require POSIX_SOURCE to be true, SGI requires it to be false. Moved nowhere. 10655. August 19, 2000 MULTI Eric Like other tasks in the past, MULTI regarded CL tables with 1,0 solutions as not good and so set weight to 0. Users do not like seeing nothing out now that weights are used. Changed to all weight 1. Moved nowhere. 10656. August 29, 2000 DFT subtraction Bill/Eric When QPTDIV was changed (15APR98), the frequency scaling array was changed to be freq(i)/freq(ref) rather than an incremental freq(i)/freq(i-1) form. Unfortunately, the calling routine was not changed to match. Bill changed VISDFT, VSCDFT, VSFDFT, QXXPTS (comments), QPTDIV, QPTADC to go to the new/better frequency form. I noticed that the 7 words used for point CC, 12 for Gaussians, and 11 for spheres, could be reduced to 4, 7, and 6, resp. So I changed the above for this saving of space and changed QGADIV, QGADIV, QSPDIV, QSPSUB to do the model computation in one loop as QPTDIV and to use the reduced memory per component. the error affected subtraction by DFT of compoennts well away from the phase center when there are > 2 frequencies. Moved nowhere. 10657. August 29, 2000 Gridded subtraction Eric ALGSTB had a habit of looping once through the data for each row swath and each spectral channel in the data set whenever the full grid of model values would not fit in the AP. It attempted to do > 1 channel at once by a very suspicious algorithm which fortunately rarely if ever actually did more than one at once. With a smarter use of the AP and overlapping swaths, it is possible to do all the channels at once. Changed ALGSTB to do this, making it go 5-6 times faster for a 14 channel test data set. Moved nowhere. 10658. August 29, 2000 QCLEAN, QUVUTIL Eric Changed QCLEAN to choose the display field in OVERLAP=2 mode whenever it is about to make available all fields on the TV. Previously, it started with the value of DOTV and continued with whatever field was displayed last. Now it at least hints at what field it will clean nexxt if no boxes are changed. Changed QCLEAN to look for the keyword NFLDRSTR in the CLEAN object and to use that value as the maximum field number to be computed and restored at the end. Any value <= 0, > MFIELD, or if the keyword is not present causes all fields to be recomputed and restored. The overlap from all fields is included in the restored fields. Changed QUVUTIL to check that the input and output uv data sets in OUVSRT have the same compression state. Added Bill's corrections to make most routines quit if IERR not 0 at the start. Added Bill's change to test the 'SCRCNO' keyword. If the keyword is in an object, its value must be tested to make sure it is positive before it is used. Moved nowhere. 10659. August 29, 2000 OOSRT Eric A UVSRT using the OOP sort routines rather than the ones in UVSRT. It is mostly just good as a test of those routines. It is about the same speed. Moved nowhere. 10660. August 29, 2000 OOP Eric Raised the limit in ZOINTD.C from 1000 to 1280 objects. The 512 field limit on images went over the 1000 limit. Moved nowhere. 10661. August 31, 2000 cationex Chris Added the source code for a stand-alone utility to combine several IONEX data files into one for TECOR to APLCONTR. The sources and makefile are in APLCONTR:cationex.tar. Primitive instructiones for compiling and running cationex are in APLCONTR:cationex.html. Moved nowhere. 10662. September 1, 2000 model division Eric For I do not know how long, the model subtraction and division using SMODEL has failed. It chose gridded in all cases and as a result did nothing - so moved the DODFT = DODFT .or. DOPTMD to after all choices between the 2 methods in UVMSUB and UVMDIV. Corrected VISDFT which no longer used the right AP locations forsuch models after the recent change in the CC handling. Moved nowhere. 10663. September 1, 2000 TVCPS Eric Changed TVCPS to make sure that the margin does not erase the picture and that the font scaling does not go to 0. Moved nowhere. 10664. September 5, 2000 COPIXEL Eric Created a new verb COPIXEL to translate between coordinates (COORDINA) and pixel coordinates (PIXXY). Changed AU7 to deprecate the copy of COWINDOW there and move to AU7B with new verb. Changed POPSDAT.HLP and NEWPARMS.001 to renumber COWINDOW and create COPIXEL. Corrected help for COWINDOW - it did not mention INNAME et al (which it used) and created a help for COPIXEL. Moved nowhere. 10665. September 6, 2000 IMINIT Eric Changed Chris' routine IMINIT to accept a missing RDATE keyword. Otherwise UVCOP et al which call IMSEL could fail for this little used table. RDATE was not in the older versions of the table (or at least the IM subroutines) and got lost sometimes. Moved nowhere. 10666. September 7, 2000 MOVE Eric The output user number is needed in HI routines if they have to expand the output HI file. Fixed it to have that number set as NLUSER. Moved nowhere. 10667. September 8, 2000 UVFLG, UVPLT Eric Changed UVFLG to flag a range of elevations rather than just elevations below APARM(4). This means changing APARM(5 and 6) to APARM(6 and 7). Changed UVPLT to offer the option using REFANT to plot elevation, hour angle and parallactic angle wrt to a single REFANT rather than wrt the average of the two antennas of each antenna pair. Moved nowhere. 10668. September 13, 2000 SETPAR Pat Finally made it accept 31DEC99 as a valid version. Moved nowhere. 10669. September 14, 2000 INDXR Eric INDXR and TABLEUTIL thought that RFREQ in MCINI was REAL*4. Everywhere else it was known to be the correct REAL*8. This caused fussy machines to quit but no damage to data. Moved nowhere. 10670. September 14, 2000 CALIB, FRING Eric Changed help and code in CALIB and FRING to have DOFIT allow a negative antenna number to mean all but those listed. Fixed the help file for KRING which already allowed this option. Moved nowhere. 10671. September 21, 2000 UVPLT Leonia UVPLT has not worked with multi freqid/subarray data if all FREQID or/and all SUBARRAY wanted to plot (FREQID/SUBARRAY=0). Choosing all FREQID or/and all SUBARRAY, UVPLT can come across to the combination of FREQID/SUBARRAY which does not exist at the data. UVPLT has died at such a case. I have fixed it making the task go to closing the data file after UVGET('INIT'..) if IRET.EQ.-1 and looking for the next FREQID/SUBARRAY. Moved nowhere. 10672. September 25, 2000 FITTP Eric FITTP had a buffer defined as 8192 for writing out images. FITAB had it right (= MAXIMG). Moved nowhere. 10673. September 28, 2000 UVPLT Eric When asked to do a fixed-scale plot of all subarrays, it would plot the last subarray Nsub times. Moved nowhere. 10674. October 4, 2000 CVEL Eric Fixed CVEL's handling of no source table to avoid a useless and confusing close call. Fixed GETFRQ; it handles the absence of an IF axis but quit if there was not one present. The formats were too small as well. Corrected BPASS - it called GETFRQ with CVPARM once (without initialization to 0.0) and with APARM once (with user's adverbs about all sorts of things not including velocity). Affected VLBA data only. Moved nowhere. 10675. October 6, 2000 XTRAN Eric Fixed XTRAN to (1) use ordinary methods when just copying the image rather than reading it all into core, (2) to do the full parameter regrid (using separate X and Y solutions) even when only 3 parameters are found (this deals with skew some) and (3) to display a second test on the fit results when only header changes are done (or DOUTP <= 0). Changed XTRFIT (1) to display more digits in the answers, (2) to warn when the apparent rotation in the X and Y axes is different, (3) to compute the correct axis increment (it mixed the 2 before which fails when the two rotations are not =). Moved nowhere. 10676. October 6, 2000 FILLM Eric Added info about DOWEIGHT and DOUVCM to the history file. Moved nowhere. 10677. October 10, 2000 XTRAN Eric Corrected the computation of fit angles and increments which were still wrong in XTRFIT. Also added a display of the fit parameters. In XTRANS, made sure the new checking routine got the right header. Corrected handling of BLC, TRC - it changed the ref pixel before the fitting but did not alter the input pixels. Moved that correction to after the fit. Also added the ability to reject user pixel positions outside BLC-TRC. Added fit parameters and INFILE to history file. Moved nowhere. 10678. October 12, 2000 CALIB Eric Fixed handling of DISK2 - a zero was passed to routines like CATKEY that could not handle it. Moved nowhere. 10679. October 13, 2000 MATCH Eric New task - converts source, antenna, FQ numbers of one UV file to match those of another. Moved nowhere. 10680. October 16, 2000 IMEAN Eric Changed the pixel counter to integer. Floats can count to 16 million but then adding one changes nothing so they cannot count by one any higher. Moved nowhere. 10681. October 16, 2000 GETJY Leonia GETJY has not selected in LINUX the right antenna list if the deselected option is used at the ANTENNA. The array ANTS was declared with dimension 50. At the same time this array is calculated for the numbers more than 50 until MAXANT=90. LINUX is more sensitive (probably) to such a "violence". I have changed dimension of the array ANTS from 50 to MAXANT. Moved nowhere. 10682. October 18, 2000 IMAGR Eric Changed IMAGR.FOR to reorder the multi-resolution widths to make 0 first. If there is a 0 width, that is the only one restored. Changed IMAGR.HLP to say this and to allow OVERLAP > 2 to mean recompute all images every OVERLAP "major" cycles. Changed QCLEAN.FOR to implement the OVERLAP > 2 option, to always test for >= 2 rather than = 2, and to write the correct max/min in the output headers (cubes required a forced open, put, then close). Moved nowhere. 10683. October 18, 2000 Misc Eric Changed: HGEOM Help file had typos. IRING Used the axis increment to find the beam area without thinking about sign. This gave a pixels/beam that was usually negative. Put ans abs value in with PPB and with the default for CPARM(2). Moved nowhere. 10684. October 20, 2000 IRING Eric Failed to handle BLC right in computing the coordinates. It corrected the center for BLC (to the subimage) and then corrected the subimage pixels to the full image. Fixed. Moved nowhere. 10685. October 20, 2000 CONVL Eric Failed on images = 8192 due to small buffer sizes - raised them. Failed on images > 8192 due to a need to make 32768 images for them. I do not know why that is so but put in message and early quit for now. Moved nowhere. 10686. October 20, 2000 FILLM Eric Changed it to omit pointing data most of the time unless CPARM(2) = 16. Fixed it to quit with only one end-of-tape diatribe. Moved nowhere. 10687. October 20, 2000 LISTR Eric Used an I5.5 format for qualifiers which always makes ***** for negative numbers. Changed to I5.4. Moved nowhere. 10688. October 23, 2000 FACES, SETFC Eric Changed the format of the NVSS and WENSS files to integer mJy again with an extra digit to fit in the big fluxes. Added Bill Cotton's algorithm for setting face locations and for defaulting cell and image size. SETFC will now write out Clean boxes (inscribed circles) for the fly's eye portion and, optionally, write out Clean circles for the exterior sources. Replace all $AIPSTARS files with NV00.0030, NV00.0100, NV00.0300, NV00.1000, NV50.0030, NV50.0100, NV50.0300, NV50.1000, WE00.0100, and WE50.0100 where the first 2 letters are the survey, next 2 are the last 2 of the epoch, and the last 4 are the flux limit in mJy. Moved nowhere. 10689. October 27, 2000 UVHOL Eric Changed help file to add STOKES (to do 4 polarizations in mode HOLG), to describe a new naming convention. Changed UVHOL to honor user's file name for output (appending pol and IF), to write all 4 polarizations out for HOLG optionally, to do more standard handling of ANTENNAS and BASELINE, to allow single source selection (there were 2 gotchas). Moved nowhere. 10690. October 30, 2000 ZLASCL Pat Fixed a bug in the PREVIEW section: regardless of the type of previewer specified in the PRDEVS.LIST file (ghostscript, gv, gs, or ghostview), one always got the same one due to an error in the logic of the script. Now fixed. Moved to CV operational systems this date. 10691. November 6, 2000 PRTUV Leonia The new option to print random parameters has been added. The list of the random parameters associated with the data is found with help of the subroutines LISRAN and PRMSET. The subroutines remove the REMOVE type random parameters and give the relevant list of random parameters' type. Random parameters or regular data are printed under control of DPARM(6). The default is print of the regular data. The random parameter BASELINE is printed as the column of relevant antennas and subarray. So the first four columns (Vis#, IAT, Base, Suba) look identical at the output print of the regualar data and random parameters. Moved nowhere. 10692. November 8, 2000 COMPRESS Eric Added a message to K2TEXT to explain that it replaces an old compress scratch file with a new one. Moved nowhere. 10693. November 8, 2000 IMAGR Eric The changes to QCLEAN dated October 18 above were never putbck. Redid them to allow automatic re-imaging of all fields and to force the image max/min to disk. Moved nowhere. 10694. November 12, 2000 POSSM Eric Believe it or not, POSSM still had 28 antennas hard coded a number of places. Changed to MAXANT and reduced a few other minor stupidities. Moved nowhere. 10695. November 12, 2000 IMAGR Eric In early July, I corrected the handling of histograms for determining how deep to Clean. Unfortunately, in OVERLAP >= 2 mode, this became a separate histogram on each field with MAXPIXEL applied a field at a time. This caused it to Clean outer noise fields very deeply, which is often wrong. Changed it back to computing the histogram over all fields every time even though all fields have not been re-imaged. In some cases, this will make the Cleaning too slow but MAXPIXEL and FACTOR can then be used to speed it up. Changed IMAGR.HLP to point this out. Moved nowhere. 10696. November 13, 2000 PRTUV Leonia The first source name has not been printed at the recently implemented option printing the random parameters. I have fixed it. Moved nowhere. 10697. November 14, 2000 TBIN Leonia/Eric The message with version of the created table has been added. The name of the output UV file has been added to the history file. Moved nowhere. 10698. November 20, 2000 TBOUT Leonia The table version, the name of the output text file, and date and time have been added to the history file. Moved nowhere. 10699. November 20, 2000 SKYVE Eric SKYVE required that the EQUINOX keyword be put in the HI file by the FITS reader. Since they now understand that word, they put its value in the header and do not put it in the HI file. Changed to allow that word to be present or absent. Moved nowhere. 10700. Novemnber 21, 2000 SPLIT Eric The HI file checked only the old adverb for whether weights were calibrated. Moved nowhere. 10701. November 27, 2000 IMAGR Eric Changed $QNOT/MAKMAP to eliminate the uniform weighting. It was trapping the multi-resolution tapering and redoing the uniform weighting over again. All tasks that use MAKMAP do the uniform weighting and initial tapering before making maps. Changed $APLOOP/UVUTIL.FOR to copy the AN and FQ tables when making a scratch uv file in UVRSCR. This caused the filtering option to fail when a sort was not required. Moved nowhere. 10702. November 27, 2000 UVCON Leonia If an image is used as a model and Jy/beam is the image unit, the following UV data have the wrong flux. I have fixed it up dividing the output UV data by the ratio beam/pixel. The help file has erroniously explained that FACTOR multiplies only the clean component fluxes. Actually FACTOR carries out the multiplication of the image as a model also. So the HELP file has been modified. Moved nowhere. 10703. November 29, 2000 $YPGNOT/PBEAM.FOR, ..HLP Leonia The new task PBEAM fits for the coefficients of an analytical model to raster-scanned data produced by UVHOL. The program fits a polynomial of the form: GO + G1*ARG + G2*(ARG^2) + G3*(ARG^3) + ... where ARG = (X1*EC)^2 + Y1^2 X1 = (X-X0)*COS(PHI) + (Y-Y0)*SIN(PHI) Y1 = -(X-X0)*SIN(PHI) + (Y-Y0)*COS(PHI) X0 and Y0 are offsets from the fiducial center. EC = b/a is eccentricity of the ellipse PHI is position angle of the major axis In the case of circular beam (EC and PHI are not fitted) the fit polynomial looks simpler: GO + G1*ARG + G2*(ARG^2) + G3*(ARG^3) + ... where ARG = (X-X0)^2 + (Y-Y0)^2 X0 and Y0 are offsets from the fiducial center. The fitting parameters are recorded in the output text file. The data points and the fit polynomial plotted at the TV or in the ps file. Moved nowhere. 10704. November 29, 2000 -> 31DEC00 Eric Changed 31DEC99 to 31DEC00 in $AIPS_ROOT/AIPSPATH.SH and .CSH $APLSUB/GETRLS $AIPNOT/SETPAR $HLPFIL/DOCALIB, SETSLICE Fixed ftp address in Y2K.HLP Updated: WHATSNEW.HLP, LSAPROPO.HLP, and ABOUT files for ADVERB AP CALIBRAT COORDINA GENERAL IMAGING INTERACT MODELING OOP PLOT POLARIZA PROCEDUR RUN TASK TV-APPL UTILITY UV VERB VLBI Moved nowhere. 10705. December 1, 2000 FITLD Leonia FITLD has made the confusion with FQ and SU table if DOCONCAT=1, if selection of IF is included and if the data have many freqid. The problem was in the two subroutines: FR2FQ and SO2SU. These routines considered number of IF to be equal to the total number of IFs even if the less number of IFs is selected. I have fixed it using the actual selected number of IFs (NIFFR in FR2FQ, SUBAND in SO2SU, instead of NOBAND). Moved nowhere. 10705. December 3, 2000 IMAGR Eric IMAGR took the average of the 64 first BCOMPs and then put that in BCOMP(1->512-64) rather than BCOMP(65->512). The setting of BCOMP was then messed up for all uses. Moved nowhere. 10706. December 3, 20000 IMAGR Eric Changed $QPSAP/Q1FIN to avoid using rows for the Hermitian parts of the grid from addresses outside the AP grid area. UVGRID was capable of generating a grid down to say row -2 and then wanting to quit (if the lowest abs(u) was row 1). If the swaths were small the grid had a statistical chance to go only to -2 rather than -3 or below. Q1FIN used row -3 for the Hermitian of row 3 in any case - even though it had not been computed and the addresses contained user dat (incl u,v,w values) and the convolving functions. For normal data, this should have made a very obvious problem (image values about = u,v,w in wavelengths). For 4-band, where u,v,w are small and fluxes large, the error might not have been so obvious. The error was only likely for large images in small APs with sparse data. Moved nowhere. 10707. December 5, 2000 PBCALC Eric Rick Perley has provided consistent and good VLA beam fits for all frequencies above and including L band. Changed PBCALC to implement these beams and changed the help files for PATGN, PBCOR, SETFC, and FACES accordingly. Changed the Fortran to implement better history writing and to alter the default selection of IMSIZE in SETFC and FACES fly's eye sections. Moved nowhere. 10708. December 6, 2000 BPCOR Eric BPCOR did not let FQID = 0 in the BP table mean any and insisted on reading a source file even in single-source files. Fixed these two sillinesses. Moved nowhere. 10709. December 7, 2000 FITLD Eric Found errors in selecting IFs for AT and IM tables. It was using NOBAND reset to input number of bands on tables for which the IF selection had been applied. Got rid of unused variables, and changed routines using NOBAND as a local variable. Changed wording in help - FITLD should select IFs correctly even if they are not in ascending frequency order. But that selection will be different than if it were done on the frequncy-sorted output of FITLD. Moved nowhere. 10710. December 8, 2000 DFTPL Eric Added BCHAN, ECHAN, BIF, EIF, FREQID to the inputs and fixed the task to loop over them and use correct frequencies to scale any position shifting. Corrected uninitialized variables that led to NaNs for some reason I do not understand. It suggests that dynamic variables are still used sometimes. Found that it did not include samples in the max/min that it should reject (self-scale but fixed range) but it did try to plot them. Fixed it to handle compressed data. Moved nowhere. 10711. December 8, 2000 HLPB Eric The lines in LSAPROPO are up to 26 characters in length. This service program clipped at 24. Moved nowhere. 10712. December 8, 2000 OHGEO Eric OHGEO copies header parameters mostly from image 1. But it set the number of axes and some other fundamentals from image 2. This is not good when the 2 disagree on these such as image 2 is a moment image. Moved nowhere. *********************************************************************** *********************************************************************** Changes while 31DEC00 (31DEC99) is NEW *********************************************************************** *********************************************************************** 10713. December 12, 2000 Update Eric Update: NEW INSTEP1 updated from 31DEC99 to 31DEC00. Moved nowhere. 10714. December 12, 2000 PAPC.INC Eric Changed the shippedcopy to include the larger sizes of AP but the one that is not commented away is still 5 Mbyte. Moved to NEW this date. 10715. December 12, 2000 HGEOM Eric HGEOM has had an error reading the input whenever it was using only a subset of the rows and was doing more than one plane. It would read 1 too may rows in each plane when the rows were long enough - for some reason it avoided this error on smaller rows. It worked on 512x512 and failed on 1024x1024. GEOM, LGEOM and PGEOM also had this error. Changed all of them to put out 1/10'th as many progress messages. Moved to NEW (31DEC00) this date. 10716. December 12, 2000 COMB Eric Fixed typo in help file. Moved to NEW (31DEC00) this date. 10717. December 12, 2000 VPLOT Eric It reverses the sign of the VLA Y antenna position twice. Models work much better with only once. Moved to NEW (31DEC00) this date. 10718. December 12, 2000 FLATN Eric Changed to do multiple fields of multiple pointings. If there is > 1 pointing, the output image is primary beam corrected. Changed PBCALC to use the LAMBDA value as a clue to init or re-init the parameters. Moved to NEW (31DEC00) this date. 10719. December 12, 2000 VPLOT Leonia/Eric There was an error addressing the X coordinate when plotting a model at the data points in the presence of time averaging. This caused a bad and only partial model plot. Moved to NEW (31DEC00) this date. 10720. December 12, 2000 INSTEP3 Eric Added IMAGR and FITAN to support Y2K as well as DDT. Moved to NEW (31DEC00) this date. 10721. December 12, 2000 PBPARM Eric The new FLATN has a new adverb (to go to PBCOR, FACES, SETFC at least in time) call PBPARM for primary beam parameters. Changed POPSDAT.HLP, NEWPARMS.001, and created a help file. Moved to NEW (31DEC00) this date. 10722. December 13, 2000 INCS.SH Eric $SYSSGI was missing a suitable INCS.SH. Moved to NEW (31DEC00) this date. 10723. December 21, 2000 SETJY Jim Added the 1999.2 flux scale supplied by Taylor and Perley, and changed that to the default. Added a warning about 43-GHz fluxes to the help file. Moved to NEW this date. 10724. December 21 omitted from NEW 10725. December 25, 2000 $APGOOP/SPECR Eric/Leonia SPECR has not worked with compressed data. It has been fixed up. Moved to NEW (31DEC00) this date.