;----------------------------------------------------------------------- ; Copyright (C) 2004-2006 ; 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 SOCORRO. 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 31DEC05 is TST *********************************************************************** *********************************************************************** 11753. December 1, 2004 roll-over Eric Changed the files GETRLS, SETPAR, INSTEP1, install.pl, update.pl, UPDCONTROL, AIPSPATH.CSH, AIPSPATH.SH, ZXHLP2.PL to specify 31DEC05. The AIPSPATH ones I changed so that new is 31DEC04 and old is 31DEC03. Moved nowhere. 11754. December 2, 2004 MNJ Eric Changed UPDRSYNC to drop the work-around for an error and AIPSUPD to fix the error, definig $logFile and the error file with full pathnames rather than relative ones. We have been leaving LogFile entries lying around. Moved to 31DEC04 this date. 11755. December 3, 2004 more MNJ Eric UPDCVS and UPDREMOVE also had this work-around. (I hate work-arounds, one should always find the general solution.) Moved to 31DEC04 this date. 11756. December 5-6, 2004 more MNJ Eric Changed AIPSUPD to set the LogFile and ErrorFile logicals with full pathnames inside the version loop. Otherwise they were set with the calling version which might not be the version(s) being done. Changed AIPSUPD, UPDCONTROL, and UPDCVS always to tee every message that goes to ErrorFile also to LogFile. It was inconsistent previously. Changed UPDUPDATE to do the $AIPS_ROOT portions only if UpdDoCopy is YES and to provide the $AIPS_ROOT value as a call arg to AIPSROOT.DEFINE. Changed UPDREPORT to use the logical LogFile rather than a locally relative name. Moved to 31DEC04 this date. 11757. December 6, 2004 $APLNOT/J2B, B2J Leonia The new subroutines J2B, B2J convert J2000 to B1950 and vice versa. Moved to 31DEC04 on 14 Dec. 11758. December 8, 2004 $APLSUB/JABER.FOR Leonia There has been an error calculating the linear speed of rotation of the Earth. TWOPI was missed. The strange OMEGA = 360.9856D0/360.0D0 changed for more recognizable OMEGA = 1.002737778D0 Moved nowhere. 11759. December 8, 2004 Misc Eric Changed: 1. FLATN.HLP There is no reason to limit IMSIZE to 8192 2 UPDCONFIG Wrong home area for MPI. Moved to 31DEC04 this date, #1 to Hawaii, #2 to Bonn. 11760. December 9, 2004 $APLNOT/JPRECS.FOR Leonia The new version of JPRECS can work with both J2000 and B1950 coordinate systems. Originally JPRECS could work only with J2000. The new version converts B1950 mean coordinates (if EQUIN=1950D0 and DIR=1) into J2000 using the new subroutine B2J (see #11757) and recalculates the found J2000 mean coordinates to the apparent coordinates. If DIR=-1, then JPRECS converts the given apparent coordinates into the mean J2000 and then recalculates J2000 in B1950 using the new subroutine J2B (see #11757). Moved to 31DEC04 on 14 Dec. 11761. December 14, 2004 $APGNOT, $APGOOP, $QPGNOT Leonia Change the old subroutine PRECES for the new JPRECS (see 11759) in the following tasks: $APGNOT/DTSIM.FOR $APGNOT/FXVLA.FOR $APGNOT/M3TAR.FOR $APGNOT/MK3IN.FOR $APGNOT/MULTI.FOR $APGOOP/UV2MS.FOR $QPGNOT/SDVEL.FOR Moved nowhere. 11762. December 14, 2004 BLAVG, UVFND Amy Add SUBARRAY selection. Both BLAVG and UVFND use UVGET to get and calibrate data. UVGET does not allow calibration of multiple subarrays (i.e., SUBARRAY=0), but UVFND and BLAVG did not allow the user to select an individual SUBARRAY, causing a frustrating feature, where they would quit with an error telling the user to select a subarray but since there was no SUBARRAY input that was impossible. Moved nowhere. 11763. December 14, 2004 SN2CL Amy The IBM compiler exposed a problem in SN2CL (a subroutine that does the interpolation for CLCAL) in which it relied on floating point arithmetic to give it an exact zero in order to not interpolate FBLANK entries. SN2CL now no longer relies on this nor does it do any arithmetic on FBLANKs themselves. Moved to 31DEC04 this date. 11764. December 14, 2004 AIPSASSN Eric Changed to add VLADATADIR for the VLA site and to add a call to LOCALASSN.CSH or .SH if present. Moved to 31DEC04 this date. 11765. December 14, 2004 FILLM Eric Changed to force the apparent position to be computed by Leonia's new JPRECS. This removes the diurnal aberation included by the VLA but it was only right for one time not the whole data set anyway. Moved to 31DEC04 this date. 11766. December 14, 2004 ATMCA Leonia Change the collinearity limit (APARM(1)=1) from 30 degrees to 45 degrees. Reduce the help file. Moved nowhere. 11767. December 15, 2004 MAKE.*MNJ Eric Changed MAKE.MNJ to copy the MAKE.*MNJ files to the update work area along with the A*, U*. and R* files. Also fixed the message that comes out ahead of doing this from link to copy. MAKE.BMNJ still was not executable in the putbck version. Moved to 31DEC04 this date. 11768. December 17, 2004 FILLM Eric Users may use fixed frequency for calibrator observations followed by tracking frequency for target source line observations. Changed the setting of the velocity type keywords to catch tracking modes should they occur after the first source. Moved to 31DEC04 December 29. 11769. December 17, 2004 CALIB Eric Changed CALIB, UVUTIL, CLBPA, GCALC, GCALC1, GCALC2, NCALC, and ECALC to return the rms from the closure computation and eventually to report the avergae rms and the rms of the rms's. Moved nowhere. 11770. December 17, 2004 WIPER Eric Changed to display the curvalue-like bit in the same location as the menu. This lets the user move it out of the way of the data. Moved nowhere. 11771. December 17, 2004 install.pl, MAKE.*MNJ, UPDRSYNC Eric The former left out HP2 architecture from the list of cvs forbidden things and the latter including a debug printout. Added SYSTEM areas for all architectures to install.pl even for rsync installs. Moved to 31DEC04 this date. 11772. December 19, 2004 PRINT Eric Changed the PRINT/TYPE verb to print the actual length of strings up to the last character rather than including trailing blanks. Sbroutine is KWICK.FOR. Moved nowhere. 11773. December 20, 2004 X band models Amy/Eric Put back models for 3C286 and 3C48 at X band. Moved to 31DEC04 automatically. 11774. December 20, 2004 FRMAP, GETSOU Leonia COMLNK/COMRPL $YPGNOT/FRMAP.FOR and $APLNOT/GETSOU.FOR using the new version of JPRECS. Moved nowhere. 11775. December 24, 2004 WIPER Eric Adjusted the curvalue-like display to keep it inside the boundary line. Moved to patch area and 31DEC04 April 22, 2004. 11776. December 24, 2004 RECAT Eric Changed AU3B to put out a renumber message only when it really renumbers something. Moved nowhere. 11777. December 24, 2004 PLOTR Eric Added CPARM so that one can have lines and or symbols with each data type. Added 2 levels of broken lines as well. This means that one does not have to devote 2 type numbers with identical data if one wants connected symbols. Moved nowhere. 11778. December 25, 2004 PLOTR Eric Put a proper dashed line subroutine into place rather than 3 special cases. Moved nowhere. 11779. December 25, 2004 SNPLT Eric Added adverb SYMBOL to control the points plot symbol used and changed FACTOR s.t. when FACTOR < 0, abs(FACTOR) scales the points plotted and the points are connected with lines. Moved nowhere. 11780. December 26, 2004 IMAGR Eric In QCLEAN.FOR: 1. Changed the limit on interpolated TV load images from -3 to -5. 2. Added histogam for current field only to CLRHIS. This is used to make the switch to/from SDI Clean decision while the histogram over all fields is used for the level to which BGC Cleans go. 3. Changed the flux loaded for SGI to use RSSMAX(ifield) rather than RESMAX which is the incorrect maximum residual over all fields, most of which are not current. It could get trapped in a position of taking almost nothing with each major cycle without this fix. Moved nowhere. 11781. December 27, 2004 BPASS Eric Forgot to fix the call sequence to GCALC et al when I changed them for CALIB. Moved to AOC immediately. 11782. December 28, 2004 PLOTR Eric A subtle error caused the lines to miss the points. Moved nowhere. 11783. December 28, 2004 EXTLIST Eric Updated AU8A for changes to PLOTR and SNPLT. Moved nowhere. 11784. December 28, 2004 SNPLT Eric Changed to allow a plus as well as a cross sign. Changed SOURNU so that its error messages can be blocked which SNPLT tried to do. Moved nowhere. 11785. December 28, 2004 PNTPLT Eric let's clarify this once and (hopefully) for all. Changed PNTPLT to require the diagonal X points in the precursor comments, to use new numbers 2-22 (were 0-20) with a new number 1 = plus sign and 23 = vertical line. Changed AU5B to have the range 1-23 for symbol and to use an X set of corners GAL X corners for new symbol number 3 IRING X corners for new symbol number 3 LOCIT to handle new plot symbol values MFPRT to handle new star numbers PLOTR to handle new plot symbol values SNPLT to do X corners for symbol always, new symbol values STARPL to new symbol values CLPLT to use PNTPL with vertical line default VPLOT to use PNTPL with vertical line default Changed HELP files: COSTAR, LOCIT, MFPRT, PLOTR, SNPLT, STARS, SYMBOL, CLPLT, VPLOT. Moved nowhere. 11786. December 29, 2004 FILLM Eric Added check on gain codes with message and flagging of data. Moved to 31DEC04 this date. 11787. 31DEC04 only 11788. January 4, 2005 install.pl, CDSETUP Eric install.pl was given option -c to mean CDrom as well as the test on logical AIPSCDROM=YES. It was fixed to turn off network when CD and vice versa and to test for previous network or CD specification not made this time. CDSETUP was excessively complicated since it assumed that one might be redoing the size of the AP or the XAS or similar things even when keeping things on the CD. These things cannot be done. I suppose that one could fully offload the CD to disk and then do a normal install rather than a CD one, but not when keeping a light footprint. Moved to 31DEC04 (04 version of install.pl) this date. 11789. January 6, 2005 OTF data Eric OTF data are now written by Linux and so are byte swapped from what they used to be. Changed: OTFUV Detects this now, lies to AIPS about which OS it actually is to get things still to work. Now calls ZBFLIP rather than ZR64RL and ZRLR64 et al. to avoid masking apparent Nan's (could have been a problem before too since strings are present). OTFIN As OTFUV. SDDPTS Changed call sequence to deal with all cases. Calls ZIPAC2 directly. ZIPAC2 C routine had typo in previously unused mode. PRTLIN The lie causes CHLTOU to fail. Changed test to 'Q' and 'q' instead. Moved to patch area this date, to 31DEC04 April 22, 2005. 11790. January 6, 2005 FILLM Eric The gain exponent test was able to make too many messages. Reordered the code so that excess gain values are reported only when they cause data that would have been pass to be instead deleted. Changed the range to -10 through +22. Values up to 15, 16, 17 are in fact used. Moved nowhere. 11791. January 6, 2005 AIPSMEM111 Ed/Leonia Long awaited memo on using ATMCA. Moved to ftp site this date. 11792. January 10, 2005 SPFLG, TVFLG Eric When looping over channels, Stokes, baselines, etc. in CLIP BY FORM and REDO FLAGS, an error in the TV load function due to no valid data should simply be noted and the function continued. Moved to patch area and 31DEC04 April 22, 2004. 11793. January 10, 2005 install.pl Eric Reversed the order of the 2 link statements that link aips and AIPS to START_AIPS. On Macs, the case dependence is lost and one ended up with AIPS rarher than aips, so switched them since the lower case is usually the preferred one. MACS will do both the rm and the ln in case-independent fashion, but when you try to execute them from the shell, then they care about case! Moved nowhere. 11794. January 10, 2005 RUN area Eric Removed BEEP, CYGLOAD, CYGLBOX, CYGEXEC, DOTLOAD, DOTEXEC, DDTHLOAD obsolete procs and CYG and CYGSAVE help files. Also deleted the CYG data from the aips ftp site. Fixed DDT.HLP to omit the useless reference to DDTHLOAD and correct the ftp site reference to Socorro and the latest versions. Moved nowhere. 11795. January 10, 2005 REGRD Eric Changed it to translate the pointing position to the new epoch along with the other coordinates. Moved nowhere. 11796. January 10, 2005 CfA Eric Changed UPDCONFIG for site CFA to 2 machines both direct clients of nrao. Changed AIPSASSN.{CSH,SH} to turn off the message emulator by default for site CFA. Moved nowhere. 11797. January 11, 2005 STQUEUE, AIPS Eric Updated Copyleft in AIPS; corrected error numbers for STQUEUE in AU2. Moved nowhere. 11798. January 19, 2005 TVFLG, SPFLG Eric These old tasks used the non-standard FLAGVER=0 -> 1 default. Corrected handling to use highest. Moved to patch area and 31DEC04 April 22, 2004. 11799. January 19, 2005 CALIB Eric Changed CALIB, UVUTIL, CLBPA, GCALC, GCALC1, GCALC2, NCALC to allow external limitation on how tight we squeeze in the robust solution method. Also BPASS call seq (use 0). Moved nowhere. 11800. January 19, 2005 VLBAFQS (VLBAUTIL.001) Amy VLBAFQS did not handle the situation where there was X, S, 90, or 50 frequency band data but no S/X or 90/50 data correctly. Fixed. Moved nowhere. 11801. January 26, 2005 XGAUS, XBASL, XPLOT Eric Moved these tasks from $APLPGM to $YPGM and added TV plot option. Modernized the code some too. Moved TVLAB and TVVEC with no real changes to $YSUB from $AIPSUB. Fixed ITICS which confused image and TV pixels in setting tick lengths, leading to wierd lengths at times. Moved nowhere. 11802. January 26, 2005 CXPOLN, CXCLN Eric This old procedure and task to clean complex images had not been properly updated to the new naming conventions. They now have and I actually tested them. They still use MX, sigh... Thanks to Uwe Bach for debugging this. Moved to patch area and 31DEC04 April 22, 2004. 11803. January 26, 2005 IMFIT, JMFIT Eric Typo "if" instead of "of" made a complex sentence much harder. Moved nowhere. 11804. January 27, 2005 UFVIX Eric Corrected: 1. Header epoch to 2000.0 when multi-source; single source was done already. 2. History gets the Besselian epoch of observation iff it is ever actually used. Moved nowhere. 11805. February 8, 2005 LISTR Eric Changed the SCAN listing so that the FQ table frequencies are picked up and used when listing the frquencies for all or each source. It used to do that only if there was no FQ table, but the SU table has an FQ id in its header so which FQ to use is clear. Moved nowhere. 11806. February 11, 2005 CLCAL, DEFAULT Eric Changed the default of DOHMS in POPSDAT and AU2A.FOR. Fixed other DEFAULT values in AU2A, several were missing. Upgraded CLCAL.HLP with added explain. Moved nowhere. 11807. February 11, 2005 VLACLCAL Amy Added DOBTWEEN and DOBLANK as inputs. Moved nowhere. 11808. February 14, 2005 CONFI Leonia The minor changes have been done to make CONFI work for very wide field of view (+- 90 degrees). The .HLP file is updated. Moved nowhere. 11809. February 14, 2005 BPASS Eric (1) Added test on number records copied to scratch file for fitting the bandpass. A no data selected problem was not trapped and bad, mysterious things happened later. (2) Changed GCALC, GCALC1, NCALC to return (1,0) whenever an antenna is missing. They would only do this before if INIT was true. (3) Changed weighting - the task seemed not to know when a baseline was missing for a particular channel but not for others. Weights are now channel-dependent but can also be affected by channel-dependent amplitude factors. Trapped the pure (1,0) solutions and blanked - to be fixed by interpolation. (4) VLBA alway interpolated (since shifted) so adverb not needed - add to help file. Moved nowhere. 11810. February 15, 2005 INSTEP3 Eric Add OGEOM, REMAG for Y2K HUGE to list of tasks done. Moved nowhere. 11811. February 15, 2005 3C48 C model Amy New source model. Moved nowhere. 11812. February 16, 2005 LISTR Eric Added DOACOR to inputs; took DOACOR=true from the matrix listing. Then made the AUTO scaling message appear only if DOACOR is true. At the end of the matrix listings, the averages and rmses of the matrix halves came out only if they differred from the first line. Now they will come out unless they are empty. Moved nowhere. 11813. February 17, 2005 FILLM Eric Attempts to make it use a different output file if the input data files were in bad time order failed. This task is very excessively complicated. Added loud messages about time order. Moved nowhere. 11814. February 17, 2005 SPFLG, TVFLG Eric Both would fail if the data are out of time order. Added messages and caused it to stop rather sooner. Moved to patch area and 31DEC04 April 22, 2004. 11815. February 17, 2005 UVSRT Eric Stop it from copying the NX table which should be obsolete. Moved nowhere. 11816. February 18, 2005 WIPER Eric The flag area function would abort if buttons B, C, or D were hit when button A was expected. Also button D would delete data despite what the instructions state. Now, when A is needed B and C are taken to be A and D exits always with no more flagging. Moved to patch area and 31DEC04 April 22, 2004. 11817. February 23, 2005 FIXBX Eric The task handled really messy boxfiles well - but did not work for single-facet cases such as using higher-frequency boxes for the next lower frequency in a generally simple case. Fixed it handle simple cases incl BOXFILE = ' '. Moved to patch area and 31DEC04 April 22, 2004. 11818. February 28, 2005 IMAGR.HLP Amy Clarified explanation of UVWTFN. Moved nowhere. 11819. March 8, 2005 Model subtraction Eric Changed message in GRDCRM to suggest using DFT when a CC file has mixed object sizes/types. Moved nowhere. 11820. March 8, 2005 IMMOD Eric The code cared about the sign of the axis increment when setting flux levels and scaling. Added a couple of ABS values. Moved nowhere. 11821. March 8, 2005 IMAGR.HLP Eric Corrected UVWTFN - Amy misunderstood and the old version was bad. Moved nowhere. 11822. March 9, 2005 SNPLT Eric The computation of LST had an odd extra factor which made both the antenna LSTs and Hour angles incorrect. Moved nowhere. 11823. March 10, 2005 INSTEP2 Eric A typo in an error handler was pointed out by Joe Lazio. Moved nowhere. 11824. March 10, 2005 EGETHEAD Eric New verb: GETHEAD but sets ERROR rather than signalling an exception. Allows testing header for existence of a keyword in procedures. Moved nowhere. 11825. March 28, 2005 DRAWBOX Eric Changed AU5C to have DRAWBOX with non-blank BOXFILE invoke the FILEBOX verb with no changing of boxes allowed and slightly different messages. Without BOXFILE, DRAWBOX does a REBOX with similar differences. Changed FILEBOX to use GRCHAN. Moved nowhere. 11826. April 6, 2005 POSSM Eric The scaling to km/s for velocity was done twice to the velocities listed in OUTFILE. Removed the extra one. It also hard-coded the epoch for the coordinates in the text file output. Moved to patch area and 31DEC04 April 22, 2004. 11827. April 19, 2005 TECOR Eric Changed web address from cddisa to cddis. Moved nowhere. 11828. April 19, 2005 CL2HF Eric Three calls to UV2TAB had an extra argument in them which confused the handling of Fortran string lengths. Caused failure under Linux but not under Solaris. Moved to patch area and 31DEC04 April 22, 2004. 11829. April 19, 2005 CONVL Eric Changed $QPSAP routines QRFFT and QXFOUR to allow them to do 32768 length FFTs. Changed CONVL to test on 2*MAXIMG rather than MAXIMG which I will leave at 16K. Also changed CONVL's computation of scratch file size to avoid overflow. Moved nowhere. 11830. April 19, 2005 CALIB Eric Changed FACSET to recognize the primary flux calibration sources and to sum only those Clean components within a standard radius of the central position. Thus interfering sources, needed for modeling, are assumed not to be included in the official source flux. Moved nowhere. 11831. April 20, 2005 CALIB Eric Changed CALIB to sum and report the average gain modulus whenever amplitudes are being solved and, if CPARM(2) > 0, saving it as MGMOD. Moved nowhere. 11832. April 20, 2005 SETJY Eric Added adverb APARM(3) to allow the user to ask for APARM(3) times the formal Baars et al fluxes. This allows for user corrections for resolution in case the user is not using one of the source models. Moved nowhere. 11833. April 20, 2005 TCOPY Eric This task got overlooked when the tape LUNs were changed around. They should be 129 - tape# rather than 30 + tape#. Moved to patch area and 31DEC04 April 22, 2004. 11834. April 20, 2005 FRING Eric Reduced to optimization level 0 in OPTIMIZE.LIS due to compiler error on HP2s. Moved nowhere. 11835. April 22, 2005 BPASS Eric Gain solutions were not initialized correctly if the first scan had channel 1 flagged. Fixed handling of the INIT parameter. Moved (this fix only) to patch and 31DEC04 this date. 11836. April 25, 2005 MAXIMG Eric I thought I could get away without changing this, but... Changed MAXIMG in PMAD.INC, changed CONVL.FOR to treat MAXIMG as it did before (limiting image to MAXIMG/2). Recompile an awful lot and relink everything. Moved nowhere. 11837. April 25, 2005 Mac version Eric Changed the PAPC.INC in $INC/NOTST/MACPPC to make 80-Mbyte AP and recompiled ZDCHIN, $QOOP, $QPSAP on pippin. The MNJ above will relink everything. Also ren SETPAR. Moved nowhere. 11838. April 26, 2005 UPDRSYNC Eric The cron that builds all of the areas for the MNJ puts a blocking file in $AIPS_VERSION/SYSTEM. UPDRSYNC looked for it in $AIPS_VERSION/$ARCH/SYSTEM and of course never found it. The blocking should be architecture independent, so fixed UPDRSYNC. Moved nowhere. 11839. April 28, 2005 CL2HF, MBDLY Eric/Dave Gordon Installed latest GSFC version of these merged with the few changes we made over the last 5 years.: CL2HF - drop unused FLAGVER, get phases from the SN table, add APARM(3) the integration time. APARM(5)=0 is explicitly deprecated (= 1 => use SN table for residual phases). MBDLY - modified to pass phases from FRING, so APARM(6) has no effect. Also it will pass 'failed' observations with APARM(7) = 1. They will get a quality code of zero later on in CL2HF (and in the Calc/Solve database). Also help files. Moved nowhere. 11840. April 29, 2005 CXPOLN Eric Changed CXPOLN to use vput/vget to save the user's input. Then test and set if needed the output name parameters. Then tput cxpoln, default mx and uvpol; tget cxpoln. Changed help to mention the defaults and encourage users to set the output name parameters (name, seq, disk). Moved nowhere. 11841. May 2, 2005 MNJ Eric Brought the UPDCONFIG file up to date replacing bob with dave. Changed the UPDCOPYSUB to use scp rather than rcp when doing simple copies. The method when using UpdSSH is too complex - using exchanged permission files - so UpdSSH remains NO for miranda. I hope this does not break other people's MNJs. Changed the message about this in AIPSUPD to describe the new situation rather than the old ("this is bad"). Moved to miranda this date, nowhere else. 11842. May 3, 2005 BOXES Eric This task, like FIXBX, could not handle really simple cases although it did the fancy BOXFILEs really well. Added minor fixes to let it do simple BOXFILEs and single fields. Moved to 31DEC04 16 August, nowhere else.. 11843. May 4, 2004 GEOMs Eric I added some comments about accuracy to the help file portions of HGEOM, LGEOM, PGEOM, OGEOM, and OHGEO warning about interpolation order and support size. The last two I also mentioned points per beam which also matters. Moved nowhere. 11844. May 6, 2005 TECOR Eric TECOR has assumed that the ionosphere remains roughly fixed wrt the Sun and so adjusts longitudes in looking up the values for times intermediate between the provided data. In many cases this is bad and it seems that much better results come from assuming the iocosphere moves with the earth. Added APARM(2) to select which model or some intermediate case - APARM(2) about 0, fixed with the earth, about 1 fixed wrt the sun. Pure 0 is taken as 1 to keep the old method. Moved nowhere. 11845. May 6, 2005 IMDIST Eric Added shift parameter display to AU9 and help. Moved nowhere. 11846. May 6, 2005 GETNAME et al. Eric Changed them to set the disk number as well - the printed display is misleading if we do not change disk 0 to something. Also CLRNAME and friends do zero the disk so this is similar. Changed AU8 and the help files. Previously these verbs changed the disk from 0 only if needed to avoid duplicate names (although GETONAME's help file said otherwise). Moved nowhere. 11847. May 9, 2005 UPDCOPYSUB Eric Added 2> /dev/null to stop login messages from by repeated by cron for miranda. Moved nowhere. 11848. May 11, 2005 UVCOP Eric Extended the time range used to copy tables by 15 minutes in each direction in order to insure that there are enough table entries for all the data that have been copied. Moved nowhere. 11849. May 11, 2005 CLCOR Eric/Bryan Butler Corrected OPAC to use the full exponential function rather than a 3-term expansion. Moved nowhere. 11850. May 12, 2005 IMFIT, JMFIT Eric Fixed both tasks to return the baseline parameters in the output adverbs and to use a better print format for them. Changed JMFIT to do something like IMFIT to find the errors for the baseline (curved plane) parameters - varying each parameter individually within a guessed uncertainty and figuring out the width of the resulting rmses. Moved nowhere. 11851. May 13, 2005 POSSM Eric Clarified STOKES default in help file. Moved nowhere. 11852. May 14, 2005 TAPLT Eric Changed it to do a more careful job of labeling the axes. When a column did not have units the axis got ticks but no labels. Changed ATAN and ATAN2 from radians to degrees and made sure any axis with that function was labeled degrees. CPARM(9) and CPARM(10) were forced to 1,0. Restored the option to do an overall scale and offset to the X axis. Added examples to the previously non-existent explain file. Moved nowhere. 11853. May 16, 2005 UVCOP Eric A typo changed it to copy only 30 minutes around the start time rather than extending both ends of the time interval as intended on May 11. Moved nowhere. 11854. May 16, 2005 TAPLT Eric Added BPARM(1) < 0 => plot the sum rather than the average in each bin. Fixed the binned plotting to be self-scaling and corrected errors that led points to be omitted from the bins. Improved the labeling at the top for binned plots and fixed an error that truncated the top label strings. Moved nowhere. 11855. May 17, 2005 FITTP Eric The conversion to true Stokes made a modest mess of the weights. It should be more correct now. Moved nowhere. 11856. May 18, 2005 CLCAL Eric Significant problems corrected: 1. Time smoothing of delays did not check for blanked delay values and so used the magic value to figure out a phase correction for the "change" in delay. 2. Merging SN tables meant only concatenating and sorting. This if SN table 1 had valid solutions for IF 1 and blanked for IF 2, while SN table 2 had the reverse, the merged table would have two records for each time. The merged records would indicate various "failures" which would probably mess up the interpolation to the CL table. Time smoothing (tiny boxcar would do with DOBLANK >= 0) would fix this sort of, but that is inelegant and often forgotten. A new, true merger has been added. If the sorted records have two or more records for the same time/antenna/subarray/source/FQID then any blanked solutions are replaced by good ones, the weights adjusted, and only one record written back into the merged SN table. CLCAL.HLP explain file had a substantial paragraph added about this. Moved nowhere. 11857. May 18, 2005 DDT, Y2K Eric Updated and corrected the DDT, DDTSAVE, Y2K, Y2KSAVE help files. Moved nowhere. 11858. May 23, 2005 AIPSPATH.SH Eric/Pat Quotes were needed to guard against undefined variables in the LD_LIBRARY_PATH area for binary jobs. Moved nowhere. 11859. May 24, 2005 SNPLT Eric Added Ionospheric Faraday rotation to the plot choices and simplified some of the labeling code. Moved nowhere. 11860. May 24, 2005 CALIB Eric SNVER = 0 is forced for single-source files - made this point in the help file. Moved nowhere. 11861. May 24, 2005 TABED Eric When looping over table versions, it set the range of allowed lines based on the first file. This was to get the range into the history - I found a new way to do this over the full list of versions. Moved nowhere. 11862. May 24, 2005 IMDIST Eric Changed AU9 to drop the requirement that the second image have the disk number specified. The name, class, and seq must be specified or the first image is used for both positions, but the disk can do normal defaulting. Moved nowhere. 11863. May 26, 2005 ZEXPN2.C Eric Changed it to zero the memory that is copied to the expanded disk areas and also changed the lseek type code to SEEK_END rather than 2. The latter is recommended although the two are identical. Moved nowhere. 11864. May 26, 2005 MCUBE Eric Added DOCONCAT option to force a SEQ>NUM> axis even when there are differeing coordinates on some one other axis. The HI even gets the ignored coordinate values. Moved nowhere. 11865. May 27, 2005 MCUBE Eric Added a header keyword whan an axis coordinate is ignored while building a cube (while forcing a SEQ.NUM. axis). This keyword then lets a restart - filling in more of an existing cube - to also add that coordinate to the history on the restart. Moved nowhere. 11866. June 1, 2005 CLCAL Eric The new version of SN2CL interpolates SN solutions to times before the first one and after the last one by a continuing linear extrapolation. This is fine up to a point but gets ridiculous at times - making weights negative for example just because the two weights differed. Changed it to limit all interpolation factors to the range -0.5 to 1.5 and the factors for solution weights to 0 to 1. Moved nowhere. 11867. June 9, 2005 PBEAM Leonia PBEAM requires the first column of the input files changing slower than the second one. If it is opposite, PBEAM creates a lot of plots with a few points and works very slow. Now PBEAM can switch the first two columns of the input files under control of APARM(7). PBEAM uses UV data AIPS file only to attach a plot file, if it is planed to create it (DOTV.LE.0). PBEAM is modified now to allow the user exclude the UV data putting OUTNAME=''. However if OUTNAME='' and DOTV.LE.0, PBEAM switches DOTV to 1 and prints out the relevant message. Moved nowhere. 11868. June 9, 2005 CALIB Eric When CALIB created a new flag table (DOFLAG > 0) and a new SN table, the latter seemed to be forgotten. This is probably because the task used two in-memory copies of the image header. One was referenced for the SN and the other for the FG. Changed to use the same one for both. This has been working fine for single-source files but not for multi-source. Moved this fix only to 31DEC04 16 August, nowhere else. 11869. June 10, 2005 PBPARM Eric Added comment that when PBPARM(2) <= 0, the default VLA or ATCA beams are used to the help files for SAD, IMFIT, JMFIT, PBCOR, BOXES, FACES, FLATN, PATGN, SETFC, LTESS, MOMNT, STESS, UTESS, VTESS, and XMOM. Apparently the old wording was not clear. Moved nowhere. 11870. June 16, 2005 READLINE, install.pl Eric Removed the 4 SHR files in $SYSUNIX that implemented the old version of readline and put back a new READLINE.SHR containing all of a modern version. Changed install.pl to look for libreadline.so and libreadline.dynlib (Macs) as well as libreadline.a when finding whether the OS already has what we need. Also changed it to drop the unpacking of RLSUPP.SHR which is now contained in the main SHR file. Changed $UPDUNIX files UPDUPDATE and UPDCONTROL to ignore changes in READLINE.SHR. An installed site does not need to react to changes we may make in READLINE. Moved the old READLINE files to $APLCONTR to keep them around. Moved nowhere. 11871. June 19, 2005 READLINE Eric The file contained rcs-recognized $Id, $Revision, $Date, $Author strings that the mnj prep script cause to change every day. So READLINE.SHR was being checked in every day. Removed these lines. Moved nowhere. 11872. June 19, 2005 XGAUS Eric Found a spot where the catalog header for the residual image was being updated on disk even if there was no residual image being written. This trashed the AMPL1 header and image structure which coud cause bad failures and caused an error about history files to be written. Major COMRPL yesterday since not all subroutines affected by MAXIMG change were compiled - also affected XGAUS. Moved nowhere. 11873. June 20, 2005 misc Eric Changed help files for 1. MF2ST meaning of star plot symbol numbers was not updated. 2. FITTP adverb CATNO applies at all times not just on DOALL. Moved nowhere. 11874. June 21, 2005 UVCON Leonia The following enhancements of UVCON are added: 1. The information defining the antenna noise temperature (TSYS) and efficiency dependence on elevation is read by UVCON from an auxiliary second input file (IN2FILE), supplied by the user. 2. The new connotation of the frequency channel bandwidth is added: APARM(8)<0 => one channel data are created with different U,V,W for each frequency. Such data are useful for simulating of multi frequency imaging. APARM(8)>=0 =>it is frequency increment and the created UV data are AIPS' standard multi channel data. The data can be used for different transformations in AIPS. In particular such data can be used for simulating of smearing effect using AIPS tasks UVAVG for time averaging and AVSPC or SPLAT for frequency averaging. 3. The error (help file) in uniform sphere notification is fixed: 3 (not 2 !!!) => uniform sphere. Moved nowhere. 11875. July 3, 2005 HIIO Eric In task MOVE, HIIO often reports that the attempt to increase a HI file size failed since the apparent size did not go up. Added to HIIO a test to see if the size got bigger than we had thought it to be previously. If so, we now ignore the error. Moved nowhere. 11876. July 6, 2005 FRING Eric Changed wording of multiple subarray condition message. Moved nowhere. 11877. June 6, 2005 3C286 C Model Amy/Eric New calibrator model for 3C286 at C band (6 cm). Moved nowhere. 11878. July 6, 2005 AIPSLetter Eric June 30 AIPSLetter put back and mailed today. Moved nowhere. 11879. July 8, 2005 IMAGR Eric Improved the help file description of NCHAV, CHINC. Moved nowhere. 11880. July 8, 2005 CALIB Eric Changed CLBPA to mark failed solutions with weight -1 for real failure and -2 for "no data". Changed CALIB to count only those at -1 not those at -2 as "failed". Moved nowhere. 11881. July 11, 2005 PNTPLT Eric The box definitions were confused - the ends of the major and minor axes are needed not corners. Changed PNTPLT to allow the simple functions to follow this rule and added a cross with gap since the other became plus with gap. Changed box in call to PNTPLT in AU5B (COSTAR), CLPLT, GAL, IRING, LOCIT, PLOTR, SNPLT, and VPLOT. Changed help files for all these plus STARS. Moved nowhere. 11882. July 13, 2005 install.pl Eric If a tarball cannot be found in the usuaal places, ask the user if the user intended a binary installation. Moved nowhere. 11883. July 14, 2005 SAD Eric Added the option to read in an image of RMS to use for setting the error bars on the components. The usual global RMS is used in the earlier portions of the task. The task RMSD may be used to compute the rms image. That image does not have to align with the image being fitted. The coordinates of each source are used to select and interpolate inthe rms image. Moved nowhere. 11884. July 15, 2005 UVnTYPE Eric Added a reference to all help files for UV gridding function types. Moved nowhere. 11885. July 15, 2005 SPLIT Eric When doing simple frequency smoothing and using CHINC the increment in the FQ tale should be computed using CHINC not the count of channels averaged. SPLAT does not do the same sort of thing. Moved nowhere. 11886. July 19, 2005 SNPLT Eric/Mark Kettenis JIVE noticed that SNPLT was not plotting PC tables correctly. To plot an IF greater than BIF one must increment the pointers by NTONE not 1. This applies to phase, amplitude, rate, real, and imaginary. Also added tests so that BCOUNT is in range of 1 - NTONE. Moved nowhere. 11887. July 19, 2005 Mac stuff Eric/Wes Added to LDOPTS.SH LINK="-L/usr/lib -lSystemStubs" Added to $APLLINUX/ZCPU.C #include which the latest OS requires, previous did this for us. Moved to pippin $SYSLOCAL, nowhere else. 11888. July 19, 2005 MSGSRV Eric The screen goes away too fast when it has trouble starting. MOved the delay call in $APLUNIX/ZMSSRV.C to make it delay before doing any of the final closes. Maybe we can then read the error messages. Moved nowhere. 11889. July 21, 2005 LSAPROPO Eric Changed HLPA to include the task, verb, adverb, etc. names themselves as symbols for the APROPO verb. Changed LSAPROPO to bring it up to date and to add all help file names as symbols. Changed AU1A also - apropos had a small error in setting the output string making for bad spelling. Moved nowhere. 11890. July 21, 2005 CLCOP Eric NEW TASK - copies right cal to left, left to right, or swaps the 2 polarizations in SN or CL tables. Moved nowhere. 11891. July 25, 2005 CCOPTS Eric The SOL architecture had not been fixed to do 64-bit files by default. If people do SOL rather than SUL, this hurts. Fixed the default. Moved nowhere. 11892. July 26, 2005 XMOM Eric Typos found while doing SPIXR - LAT and LONG axes would not have been recognized as equivalent to RA and DEC. Moved nowhere. 11893. July 26-27, 2005 BOXES, FACES, SETFC Eric Added code to go through the files doing formatted reads only every 25 records until we are nearly in range and then quitting when we are out of range. This only works on files that are in strict RA order - those in $AIPSTARS. But they are the long files. Moved nowhere. 11894. July 29, 2005 IMMOD Eric Changed task from one that added a single component in flux and arc second units to one that adds up to 4 components in the brightness and pixel units (and adverbs) used by IMFIT/JMFIT. Moved nowhere. 11895. July 29, 2005 SPIXR Eric NEW TASK: Finds the spectral index along the X axis of a cube which can be either a true frequency axis or a SEQ.NUM. axis with frequencies in the history file written by MCUBE. Moved nowhere. 11896. July 29, 2005 TI2HA Eric Changed it to use the hour angle of antenna 1 rather than the average of the two hour angles in the baseline. For VLBI, where this matters the old usage took data from the same time and put them at separated times. Moved nowhere. 11897. August 1, 2005 WINDF, CCEDT Eric Changed $APLSUB/WINDF and CCEDT to check rectangular boxes for reversed corners and reorder them to the expected BLC, TRC order. WINDF is used in IMAGR et al. Local versions of WINDF in CHKFC and FIXBX were also modified. Moved nowhere. 11898. August 1, 2005 IMAGR Eric Found a typo in TVUTIL.FOR which should have made it hard to get TXINC for TV loads. Changed QCLEAN to separate OVERLAP 2 into a separate cleaning subroutine from overlap 0 and 1, thus simplifying the code. Changed UVPREP to catch and signal the no visibilities error. QCLEAN had some odd code that limited the interpolated image to < 520 pixels for some unknown reason. That code also allowed the TXINC, TYINC to become -1 which would then load different increments in X and Y. Moved nowhere. 11899. August 1, 2005 Spectral indexes Eric Fixed SPIXR to use the reference frequency as the reference for spectral index when the axis is a regular FREQ axis. Changed UVMOD, adding DPARM to specify spectral indices and curvatures for 4 sources and the noise level. Moved nowhere. 11900. August 4, 2005 FITLD Eric Phil code: Two JD dates obtained by dufferent means were differenced and the difference was required to be >= 0. A difference of -1.0D-6 caused the task to skip the data file! To add more injury, more than one subroutine did not initialize the version number before calling the ext. table INI routine. That was fine on the first call and if there is only 1 version but not if there are multiple subarrays or whatever. But this is the first 2005 fix of FITLD!!!!! Moved to 31DEC04 16 August, nowhere else. 11901. August 4, 2005 REAMP Eric NEW TASK: Cleaned up version of FUDGE, rescales vis amplitudes and weights. FUDGE does this too but the new task is cleaner. Moved nowhere. 11902. August 5, 2005 CLCOP Eric Enhanced so that it can make a 2 polarization file from a 1 polarization file. Fixed it to do weights, reference antennas, clocks, multi-band delays, etc. These were overlooked previously. Moved to AOC Linux this date, nowhere else. 11903. August 6, 2005 CCEDT Eric Corrected numerous problems: 1. Made CCBOX be in a common so that its dimensions could change should the automatic splitting be done. 2. Fixed history to tell the boxes correctly in both arc sec and pixels. Fixed formats to fit. 3. Put tests in to make sure that the number of boxes and the number of output files are the same when writing multiple output files. 4. NCCBOX as a negative bvalue could be used in the past to create data in an object. Error fixed by going to the common. 5. Changed to set CCBOX in arc sec to look like like BLC/TRC in those units and corrected the test for component inside a box. The old test assumed boxes in the correct order and axis increments < 0 in X and > 0 in Y. 6. Changed code to do all the work of summing fluxes only when CUTOFF is being used. 7. Some counters assumed that there were <= 10 boxes despite the parameter MAXBOX used elsewhere. 8. The automatic algorithm changed the size of search boxes by too much for each minor iteration. Limited the change to 10%. This gets more realistic boxes. 9. Changed the automatic algorithm to set CCBOX and NCCBOX - so that the history file could contain the correct boxes rather than some adverb values that were ignored. 10. Changed the automatic algorithm to do only one pass - previously it would do overlapping boxes - such as a large one with a smaller one taken out of the middle! Moved nowhere. 11904. August 6, 2005 IMAGR Eric New QCLEAN subroutine CLNUV2 was able to return an error code = -2 indicating that the selected field was too weak and there were no more fileds. Suppress this as an error. Moved nowhere. 11905. August 9, 2005 FLAGR, FINDR Eric Changed ASOLVE to return negatives on truely failed solutions and magic blanks on non-existent data. Similarly, changed VSOLVE to return pure zero on failed and blanks on no-data. Changed FLAGR and FINDR to flag on things <= the cutoff rather than < cutoff so that zero will flag the failed solutions - previously they were blanked and ignored and so not flagged. Changed help files to match. Moved nowhere. 11906. August.10, 2005 Intel compiler Eric Changed $SYSLINUX versions of CCOPTS.SH, LDOPTS.SH and FDEFAULT.SH to insert the new Intel compilers icc and ifort and the parameters that Benjamin Mort (Oxford) found for them. Moved nowhere. 11907. August 11-12, 2005 FITLD Eric Changed: (1) removed a large set of unused FORMATs. (2) Added code to recognize station Y with gain 1.0 as needing the Tant set to -1.0 rather than magic blank. This tells later code that the Tsys is really Tsys/Tant (values less than 1 rather than in 2 or 3 figures actually). (3) 8/12 changed to test sensitivity rather than gain curve since that is 1.0 frequently. Moved to 31DEC04 16 August, nowhere else. 11908. August 16, 2005 BOXES, FIXBX Eric IMAGR can also set boxes and fields with adverbs as well as a boxfile. Changed: FACES - Fixed bug in reading stars files SETFC - Fixed bug in reading stars files BOXES - dropped SHIFT adverb that was not used, added NFIELD, RASHIFT, DECSHIFT, fixed bug reading stars files NFIELD if > 0 controls the max number of facets no matter what the BOXFILE contains. FIXBX - Added NMAPS, NOISE, RBSIZE, NBOXES, CLBOX to describe old imaging parameters along with INFILE, added NFIELD, RASHIFT, DECSHIFT to describe new imaging parameters along with BOXFILE. Fixed incorrect wording in help file. Moved nowhere. 11909. August 22, 2005 Dynamic memory Eric We need to use LONGINT for address pointers in dynamic memory since 64-bit computers may use more of the virtual address space than can be represented in ordinary integers. This means declaring the index pointer a LONGINT which is translated to INTEGER on 32-bit addressing machines and to INTEGER*8 on 64-bit addressing machines. Changed: ZADRSZ New - returns size of an int pointer (4 or 8) A copy is in $SYSUNIX/ZTRLOP.C already. ZMEMRY Change INTEGERs to LONGINT. Fixed QUIT opcode to delete all entries ignoring all other args - FRAL deletes only those of named routine. ZMEMR2 $APLGEN stub fix declarations ZMEMR2 $APLUNIX C change int * to long * for offset. DIETSK Added call to ZMEMRY with opcode 'QUIT' ZDBPC $INC changed PBPBUF, PBPIBF, PVLBUF, ANTNX, ANTPNT pointers to LONGINT BPASET Changed comments (calls ZMEMRY with above), changed to 0 ANTNX with DO loop since FILL may not be okay. BNGET Changed local variables that are set from pointers. DATBND Changed local variables that are set from pointers. IOBSRC Comments CALADJ Changed OFFSET to LONGINT - local dynamic memory. UVGET Changed comments, frees PBPBUF and references it. INTERPLATE Changed PTIMAG to LONGINT. TVUTIL Changed subroutine OTVBOX pointer WINP to LONGINT. ACFIT Changed pointer VISOO and variable K (to KLONG) computed from it to LONGINT ATMCA Fixed to call ZMEMRY only if really needed and to free the memory once used. It otherwise created two memories with only a single pointer even when they were not needed and never freed them. BPSMO Changed secondary pointers LWT, LBP CLIPM Changed pointers FLGIT Changed pointers GETJY Changed pointers MSORT Changed pointers OTFUV Changed pointers and local variable SNFLG Changed pointer SPLAT Changed pointers STFUN Changed pointers UVLIN Changed pointers UVMLN Changed pointers XTRAN Changed pointers - are call args from main SDMOD Changed pointer - is call arg from main, also local version of pointer UVMTH Changed pointer BLAPP Changed pointers, added FRAL call CCEDT Changed pointer FIXWT Changed pointers FXPOL Changed pointers BPASS Changed pointers CALIB Changed pointers in 2 routines DEFLG Changed pointers FRING Changed poointers in two places. In the dynamic pseudo-AP, changed to LONGINT pointer, then set AP pointer to LONGINT one, and if not equal branch to error message and exit. I*4 AP pointer will not be equal I*8 pointer if allocated memory is too far away. KRING Changed numerous pointers BLING Changed multiple pointers and corrected their usage - 1 has to be added to all references. BSAVG Changed pointer and secondary pointer IMAGR Changed pointer in common and 2 secondary ones OMFIT Changed numerous pointers - it appears to be allocating many small areas which is not a good idea. Removed unused variables. SCIMG Changed pointer in common and 2 secondary pointers WFCLN Changed pointer in common and A secondary pointer DELZN Changed pointers in common ELINT Changed pointers in common IBLED Changed pointers OBPLT Changed pointers VPLOT Changed pointer and numerous secondary ones WIPER Changed pointers FRPLT Dropped DBPC.INC BSCLN Changed pointers Moved to CfA promptly for testing. 11910. August.23, 2005 FINDR, FLAGR Eric Changed dimension of certain time arrays from 1000 to 15000, putting them in the same common to save space. Corrected the VRFI routines to check counters against this limit also. Moved nowhere. 11911. August.26, 2005 ANTAB Eric Increased max FQID from 12 to 25. Moved to AOC this date, nowhere else. 11912. August 26, 2005 REUSE Eric NEW SERVICE TASK: Converts from user public catalogs to user private catalogs, renaming the CA, SG, and all data files. Moved nowhere. 11913. August 26, 2005 Public catalogs Eric Changed: AIPS Will create and init public catalogs if missing AIPSB Will create and init public catalogs if missing FILAIP Dropped 2 questions - will always do private catalogs and 10 tape lock files. NVOL not needed at all. FILAI2 Changed to do private catalogs always - this program not used now I think. install.pl - changed its use of FILAIP. Moved nowhere. 11914. August 28, 2005 RESEQ, MATCH, PRTAN Eric Changed MATCH help file to make it clear that it effectively concatenates the AN, FQ, and SU information. Thus, careful applications of MATCH will allow separate VLBI data sets (each missing some antennas, sources, and/or FQIDs) to be made so similar that they can be safely merged. Task RESEQ is meant to renumber the various ground station "antenna numbers" for the space antenna into a single antenna number. It had the attributes: 1. It used a non-standard way to read and write UV files which failed when I tried it and died with no message in its set-up phase when Craig tried it. A FUDGE clone would have worked. 2. The tables subroutine did not set a LUN. 3. The polarization checking code used a wrong reference pixel. 4. The subarray number was dropped from the output baseline. 5. The antenna number column in AN tables was not recognized as such. Took a FUDGE clone to copy the UV and made RESEQD (DIDDLE) do the UV copy wit antenna renumber and possible baseline direction swap. Copied the TABLE handling routine from the old RESEQ and cleaned it up and made it work. The new task will be able to renumber antennas in complicated schemes, completely swapping things around if desired. PRTAN confused AN row number with antenna number, causing the VLA display to be messed up when the 2 are not equal. Moved nowhere. 11915. August.28, 2005 3C48 L-band Model Amy Putbck new model for 3C48 at L band 1.426 GHz. Moved nowhere. 11916. August.30, 2005 LINUX Eric Added -axNP option the $SYSLINUX CCOPTS.SH, LDOPTS.SH, and FDEFAULT.SH for the Intel compilers. This should get us load modules that will work on generic machines but also take advantage as much as possible of the current PIV chip. Moved to AOC Linux this date, nowhere else. 11917. August 30, 2005 FRING Eric The formula for how much memory is needed in the fring search was wrong. This caused the dynamic memory routines to ask for slightly too little memory when doing enormous fring searches. Moved nowhere. 11918. August 31, 2005 $SYSLINUX, XAS Eric More changes for the Intel compiler: 1. -xW overrides the -axNP so drop it in CCOPTS.SH, FDEFAULT.SH 2. Force linker to use AIPS-provided run-time libraries with -Xlinker -rpath -Xlinker $LIBR/INTELCMP 3. Add comments to Makefile in XAS.SHR as guides to someone using the Intel compiler. Moved nowhere. 11919. September 2, 2005 TYPUVD Eric TYPUVD is set to identify types of UV data. UVMOD, UVCON, MX, and UVMAP took +1 to be -NCP when the correct value is -1 for -NCP. The last two exited on TYPUVD > 0 at one point and then tested for +1 later. Moved nowhere. 11920. September 2, 2005 CELGAL Eric On UV data, AU7 (CELGAL verb) changed RA/DEC to GLON/GLAT but then back to RA--/DEC-. Changed it to test properly and return with or without the -'s. Moved nowhere. 11921. September 6, 2005 Single dish Eric Changed UVPGET to allow Galactic, Ecliptic, and az/el coordinates for random parameters and axis parameters but only for single-dish data. Drop projected coords. Changed OBJMNGR to call all coordinate types that are not projected 'SD'. Changed UVUTIL subroutine U2IDES to test coordinate types and use what comes in for SD data only. These two will take care of SDGRD I hope. Changed SDIMG to use whatever coordinate comes in rather than forcing RA-- and DEC-, but add the -'s if needed. Made it check that the random parameters and the axis coordinates are the same type. Removed SELSD and GRIDR, they were fully replaced by SDIMG and SDGRD and do not handle compressed data and are otherwise is need of work. This allows us to drop projected coordinates from single-dish UV data entirely. SDVEL so far tests for RA/DEC and quits if it is not. It could do everything with a subroutine to convert Galactic and Ecliptic to RA/Dec. Moved nowhere. 11922. September 6, 2005 ANTINI Eric/Bill A more careful approach to finding the columns pointers is used most places but was not used here. Bill has pointed this out and provided the revised code unsolicited. Moved nowhere. 11923. September 6, 2005 CCEDT Eric Fixed error - I took CUTOFF to be used only if positive, but one may want to allow minor negatives just not big negatives. Moved nowhere. 11924. September 11, 2005 PRTAB Eric Changed it to allow for negative times, correcting range tests and formatting. Moved nowhere.. 11925. September 12, 2005 FITLD Eric Wrote new include DCTV.INC and new subroutines CTINI, TABCT, and CTREFM to init, read/write, and reformat (add a time interval column) CT (Calc) tables. Changed FITLD to invoke the reformating and then to add the time intervals after the files has been read. Moved nowhere. 11926. September 13, 2005 INTERP Eric INTERP is too nice a subroutine name. The $APLNOT/INTERP.FOR subroutine is not used anywhere but it is a good routine. So changed $APLNOT/JPLEPH.FOR internal subroutine name INTERP to JPLINT. Also changed comments and/or subroutine name internal to AIPS, AIPSB, AIPSC, DAIP, KWICK, POPSGN, UVCON, and APCAL. I do not see how this was missed earlier - the Mac MNJ failed last night for this. Moved nowhere. 11927. September 15, 2005 MCUBE, SPIXR Eric Changed MCUBE to write an FQID axis when forced to combine frequencies that are not well aligned. Moved FQINI and TABFQ from $APLNOT to $APLSUB with a substantial improvement to the code (see e.g. ANTINI above) for use by MCUBE. Changed SPIXR to read the FQ table when axis 1 is FQID; corrected the frequency chosen for primary beam correction, made the reference frequency be that from the header on the FREQ axis (or what is left of that axis). Changed TRANS, CPYRT, REMAG to copy all table extension files. Moved nowhere. 11928. September 16, 2005 EXTDEST Eric Changed AU8 to write EXTDEST happenings into the HI file. Moved nowhere. 11929. September 21, 2005 CLCOR Leonia The new OPTYPE='EOPS' has been implemented: CLCOR reads the CT table, interpolates the eop parameters (UT1-UTC, WOBXY) into the the CL table row time, calculates the delay taking into account the EOP correction. This is the delay (possibly wrong) which used by correlator. Then CLCOR reads the USNO file, interpolates the eop parameters (UT1-UTC, WOBXY) into the the CL table row time, calculates the delay taking into account the EOP correction. This is the correct delay. Having calculated the two delays CLCOR applies their difference to correct the given CL table row delay, rate, phase. So the new OPTYPE='EOPS' allows to correct the wrong application of the eop parameters by the VLBA correlator. CLCOR has erroniously corrected the total geometrical delay for OPTYPE='ANTP' and 'ANAX'. This correction has been removed. Moved nowhere. 11930. September 22, 2005 UVCON Leonia W term has been always considered zero. I have added W term calculating visibilities. It works only if the model is given as a set of CC components (CMETHOD='DFT ', CMODEL='COMP') Moved nowhere. 11931. September 28, 2005 CLCOR Leonia CT table first group has been printed out if the all groups are identical. Nothing has been printed out if the groups are not identical. Now all CT table groups are printed out if the groups are not identical. The warning about using OPTYPE='EOPS' only for VLBA correlator has been added. Moved nowhere. 11932. October 3, 2005 FLGIT Eric via Amy Fixed loop over subarray. The closes were outside the loop but the UVGET ('INIT') is inside. Moved nowhere. 11933. October 6, 2006 VLBAUTIL Amy Compressed procedures by putting variable declarations on one line where possible. VLBALOAD - add ability to load from disk. VLBAMPCL - add check on repeats in the ANTENNAS inputs, a user used ANTENNAS 5, which filled the ANTENNAS array with 5s and since I needed to add the reference antenna to the list this caused it to croak. So now it checks for repeats and changes things like ANTENNAS=5 to ANTENNAS=5,0. VLBAEOPS - new procedure to run CLCOR to correct EOPs. If INFILE is left blank then the USNO file with final EOPs is automatically downloaded into the /tmp directory and used with CLCOR. The file is them deleted. Moved nowhere. 11934. October 9, 2005 UVFLG, TACOP Eric Changed FLAGUP to do a JTRIM on the reason in an existing FG table before comparing the reason from the user adverb. Changed error messages in TACOP to be a bit more explicit. Moved nowhere. 11935. October 10, 2005 LISTR Eric The MATX listing limited the time averaging to > 0.0001 days or about 7 sec. Fixed to something much less. The GAIN listing forced 2 second averaging - changed to 0.01. Moved nowhere. 11936. October 10, 2005 MFPRT Eric Changed the basic printing and the LIST printing to include the file INNAME, INCLASS, and INSEQ. Useful whenlisting many files. Moved nowhere. 11937. October 10, 2005 VLBAUTIL Amy Add VLBAEOPS to VLBAUTIL.HLP and added a couple information lines at the end of VLBAEOPS (i.e., CL# ? contains the EOPS correction...). Moved nowhere. 11938. October 10, 2005 QINP Eric New verb - restart INP or INPUTS (or QINP) at the last page that was showing, including any new adverb values. Changed AU1A.FOR, NEWPARMS.001, POPSDAT.HLP and new QINP.HLP Moved nowhere. 11939. October 11, 2005 CLCOR.HLP Leonia The explanation of CLCORPRM for OPCODE='EOPS' included JD. This confused people who could think that JD information has to be provided by the user. Actually it has not. I changed the help file with the help of Michael Bietenholz. Moved nowhere. 11940. October 11, 2005 GETNAME Eric The GETNAME error message for bad volume used the wrong variable so complained that 0 was > 4. Fixed AU8. Moved nowhere. 11941. October 13, 2005 FRING Amy When the Intel compiler optimized the loop that calculated the frequencies (FREQS) the accuracy was degraded enough (errors on the order of a few hundred kHz), that it triggered a "FREQUENCIES NOT ON A UNIFORM GRID" warning. Changed the FREQS array to double precision, which seems to force the Intel compiler not to optimize the do loop in question, and the calculated FREQS now have a high degree of accuracy. Moved nowhere. 11942. October 13, 2005 XAS and UNIX SERVERS Eric It was discovered that some sites need quotes around the process name to be executed inside the xterm. Moved nowhere. 11943. October 14, 2005 FRING Eric A MOD operation conflicted when DF was REAL*4 and FREQS was double. Made a DPDF to use inside the grid test routine and set it to the real*4 DF on the way out. Made FTOL be properly known as fractional error (one use was as absolute ~ 1 MHz) and fixed some comments. Moved nowhere. 11944. October 15, 2005 UPDUPDATE Eric Add section to copy $SYS$ARCH to $SYSLOCAL except for LDOPTS, CCOPTS, and FDEFAULT and to make noise about them. This will let me put no-quote versions if XASERVERS and UNIXSERVERS in $SYSSUL. Which I have done. Moved nowhere. 11945. October 15, 2005 $APGNOT/RMSD Leonia The new OPCODE='mean' has been added at request of Rob Ivison from Edinburgh (UK). RMSD with this opcode creates the image of MEAN at the vicinity of the pixel. The existing OPCODE='rms' allowed to create the image of RMS at the vicinity of the pixel. Moved nowhere. 11946. October 19, 2005 $RUNSYS/VLBAARCH.001 Lorant After Amy added reading disk files with VLBALOAD (upon my request), the VLBALOAD help file changed, causing INDXR in ARCHLOAD to break. Fixed by adding INFILE='' to the code. Moved nowhere. 11947. October 19, 2005 UPDUPDATE Eric Tests of this on a Linux box allowed == for an equality test. Unfortunately, this mean Linux is nicer than other boxes which insiste on the correct = test. Moved nowhere. 11948. October 19, 2005 Slices Eric Fixed: SLICE to use a better format in the output text file SL2PL to remove typo that drew an extra line at x=0 AU8A to adjust pointer to max & min in SL listing SLFIT to say physical units for the 2nd output list Moved nowhere. 11949. October 20, 2005 VLBAEOPS Amy When CLCOR was run INFILE pointed to the wrong directory ($FITS rather than /tmp). Fixed. Moved nowhere. 11950. October 20, 2005 MOVE Eric Changed limit on userid from 8192 to 46655. Moved nowhere. 11951. October 24, 2005 RSTOR Eric Added FACTOR adverb to allow components to be removed instead of added. This still works only when components are exactly on pixels, components are points, and the image is 2^n in size. Moved nowhere. 11952. October 25, 2005 tape Eric Added a message to ZTPOPN suggesting a mount when AMT0n cannot be translated. Changed TPMON to add error reports after every call to a Z tape routine. Moved nowhere. 11953. October 26, 2005 SYSTEM Eric Added new adverb SYS2COM to allow SYSTEM commands to be very long. Changed: AU1A Inputs now only show the characters in the string, not all the trailing blanks AU3B To pick up the new adverb SYSTEM Help file to describe usage. ZSHCMD $APLGEN changed to support longer strings ZSHCM2 $APLUNIX and $APLIBM changed to support longer strings SYS2COM New help file POPSDAT Add adverb NEWPARMS Add adverb Moved nowhere. 11954. October 27, 2005 CALIB Eric Changed the subtle adjustments made to times - the scans were extended by 1 sec in each direction, now by 0.05 sec. Other adjustments were changed from 0.1 to 0.05 sec. The VLA is producing data at 0.41666 sec intervals and the VLBA can also do short intervals. Moved nowhere. 11955. October 28, 2005 UVHOL Amy At Rick Perley's request I added the option to print the elevation and azimuth at the beginning and end of scans to output file. APARM(4)>0 triggers this new option. Moved nowhere. 11956. October 30, 2005 FILLM Eric FILLM took bandwidth code 8 to be the same as 9 and to be 50MHz/(2^9). Actually both modes are 50/(2^8). This led to wrong reference channels and other less important errors when data were taken in mode 9 (very narrow spectral line mode). Changed help file to say that NCOUNT is ignored by the ON-LINE version. It already says that STOP will always be needed or a stop time in the ON-LINE version. Moved nowhere. 11957. October 30, 2005 install.pl, UPDCONTROL Eric Changed UPDCONTROL to put out messages about recompiling ZTRLOP, PP, et al only for non-binary jobs. Changed install.pl to make an AIPS symlink only if not MACPPC - that system confuses AIPS and aips. Moved nowhere. 11958. October 31, 2005 FITLD Eric Rearranged code in UT2UV so that files will be closed on error to allow later file management to occur, trying to die gracefully. Moved nowhere. 11959. October 31, 2005 VPLOT Eric VPLOT had a limit of 1000000 visibilities and when it got above that limit it kept counting and reading the data. Then in plotting it thought it had much more data than it had. Changed to test the limit more carefully and to quit counting and reading at that point. After testing this I raised the limit to 2000000. Note that this is via dynamic memory anyway. Moved nowhere. 11960. November 1, 2005 CALWR Eric Changed it to force the BLC and TRC to retain the center pixel as the center pixel in the output. Moved nowhere. 11961. November 1, 2005 Cal models Eric Redid all the calibrator models so that the center pixel is the center pixel of the subimage actually shipped. A feature in the imaging causes subimages that do not have this property to shift the object in self cal. Moved nowhere. 11962. November 2, 2005 TECOR Amy Added the the ability to use a directory name, rather than just a environmental variable in INFILE. This was needed for the VLBATECR procedure. Moved nowhere. 11963. November 3, 2005 VLBAUTIL/VLBATECR Amy New procedure that 1) looks at the header and NX table calculates the date and number or files it needs to downloads; 2) downloads the files using SYSTEM and the system commands wget, curl or ncftpget (depending on what is available) to the /tmp directory; and 3) runs TECOR. Moved nowhere. 11964. November 7, 2005 BPASS, CPASS Eric BPASS began with a call to CLSORT on CL table 1 for a VLBA data sets. Then, if VLBA and ACONLY, it called a routine that handled shifting of the observation. That called FRQUPD but told FRQUPD that the CL table was already open and sorted. The latter was true but not the former. Dropped the call to CLSORT and changed the call to FRQUPD so that FRQUPD will open and sort the CL table if it is going to use one. FRQUPD is smart enough to handle the absence of a CL table without complaint. CPASS for no visible reason called CLSORT as well. Moved nowhere. 11965. November 8, 2005 CLCOR Leonia If the INFILE (opcode 'EOP') is loaded long ahead of the observation, then the last date with real data in the file can be too close to the OBSDAT or even precede it. The using the INFILE may be wrong in such a case. So CLCOR is terminated now in such a case and the relevant warning is issued. Moved nowhere. 11966. November 9, 2005 install.pl Eric Added code for AMD 64s to (1) recognize the architecture as LINUX but also set a new variable IS64, (2) remove -malign-double from FORTOPT when IS64 is YES, and (3) change the library in XAS' Makefile to lib64 when IS64 is YES. Now nothing special has to be done for the AMD 64s. Moved nowhere. 11967. November 10, 2005 CVEL, BPASS Eric CVEL has been dying in the Intel compiler. With debug messages added but not actually output, it does not fail. I hate magic fixes. BPASS was corrected recently - previously it was supposed to read the CL table for autocorrelation bandpasses on the VLBA, but managed to avoid doing so. When I fixed it to try to do so, all sorts of conflicting LUNs showed up. Changed the LUN for table sort to 89 and 88 in TABSRT. Changed 2 LUNs away from 49 which is used in FRQUPD and never closed. Moved nowhere. 11968. November 14, 2005 subimage self-cal trouble Eric When self-cal (or UVSUB or ...) with a sub-image, the tasks get confused if the original central pixel is no longer the central pixel. Also, the tasks cannot tell if the imaging was done with 3-D or not. I have decided to add 3 header parameters to tell the type of imaging (KIITY) with value 1 for non 3-D and 2 for 3-D (and 0 says not to use the other 2 parameters). The other two parameters are the offset of the reference pixel from the numeric image center in the original image. Changed: DHDR.INC Add the 3 pointers VHDRIN Compute the 3 pointers CATN2L Send the new header data to-from networks IMCREA Set the 3 parameters in the header IMHCOR Set the 3 parameters in the header GRDAT If the type is 1 or 2 get the XPOFF, YPOFF from the header, else compute as before (wrong for subimages) SETGDS If header is 2 count image as do3d true SETDO3 If header is 2 count image as do3d true AU7A GETHEAD and PUTHEAD on added keywords - PUT is dangerous but will allow subimages to be corrected PUTHEAD help file updated - include making new keywords GETHEAD help file updated EGETHEAD help file updated CALWR add line of image type,xpoff, yposs FITTP add line of image type,xpoff, yposs FITAB add line of image type,xpoff, yposs DFIT.INC add 3 parameters for reading tasks VFIT.INC add 3 parameters for reading tasks Moved nowhere. 11969. November 16, 2005 UPDUPDATE Eric Changed the grammar of an or statement in the script - Solaris will not by or statements so far as I can tell. This led to error messages that seem harmless if strange when I fixed the g77 options in $SYSSUL/FDEFAULT.SH. Moved nowhere. 11970. November 17, 2005 UPDCONTROL Eric Did not make a special report file for binary jobs any more but then tested the logical to see if it exists and is bigger than 0 bytes. Solaris (again) does not like this. Moved the file creation outside the if binary section. Moved nowhere. 11971. November 17, 2005 DELZN Leonia/Eric The source file has had low dimension (30). It caused corruption of the OUTFILE if the actual source number exceeds 30! Fixed it increasing the dimension to MXSOU=100. Moved nowhere. 11972. November 17, 2005 STARPL Eric The reading of the background image could get caught with a pixel slightly outside the image which would then get a bad window return from the read-init operation. Fixed to suppress this edge effect. Moved nowhere. 11973. November 18, 2005 DELZN Leonia/Eric Leonia made a mistake at the yesterday DELZN correction. Moved nowhere. 11974. November 18, 2005 CALIB Eric Changed it to calibrate weights when calibrating the single-source output file. Moved nowhere. 11975. November 18, 2005 CHKFC Eric Changed it to allow F cards in the BOXFILE. This requires adding the COORDINA adverb. So also changed it to take NFIELD, RASHIFT, DECSHIFT, NBOXES, CLBOX and do the usual things with them. This means that BOXFIL can be blank and the adverbs are used unless overridden by the BOXFILE. Moved nowhere. 11976. November 20, 2005 3C286_L.MODEL Amy/Eric Checked in Amy's new model Moved nowhere. 11977. November 20, 2005 UVPLT Eric Added FACTOR < 0 to tell it to connect the pplot points with lines. Moved nowhere. 11978. November 20, 2005 Misc Eric WIPER - changed to free all dynamic memory rather than just one. Some users reported troubles at exit. Also did not die when some of the dynamic memory did not allocate properly. UVMLN - typo caused it to do history only when there was an error, this was backwards from desired. POSSM - added antenna number un upper right corner of appropriate plots (along with station info in upper left). Moved nowhere. 11979. November 20, 2005 SNSMO Eric Changed VLBI to average rates only over polarization, previously rates were average over IF and polarization. Added SMOTYPE VLRI to average rates (only) over IF also and explicitly have VLMB do so also. Moved nowhere. 11980. November 21, 2005 LSAPROPO, ABOUT Eric Updated HELP files LSAPROPO, and about files ADVERB ANALYSIS AP CALIBRAT CATALOG FITS GENERAL IMAGE-UT IMAGING MODELING ONED PLOT POPS PROCEDUR RUN SINGLEDI SPECTRAL TABLE TASK TV-APPL UV VERB VLBI Fixed REAMP.HLP which declared CALIBRAT not CALIBRATION as its secondary keyword. This messed up the SHOPH task. Also update the tab completion file HLPIT.LIS. Moved nowhere. 11981. November 21, 2005 FIXBX Eric Added NGAUSS adverb. Changed it to honor the input adverbs NMAPS and NFIELD rather than depend on the input box files. Changed the reading of those box files to include Clean box field numbers to set the number facets if the adverbs are zero. Previously it ignored the input adverbs and depended solely on the box files. BOXES already does most of this. Moved nowhere. 11982. November 21, 2005 Flagged FG table records Eric WIPER copied FG table records but did not test for such records being flagged. Instead it died silently. Added messages and omit the copy of flagged records. Also corrected subroutine FGREFM and tasks CLIPM, SNFLG, UVMLN, VPFLG, CALIB, DEFLG for this over sight. Moved nowhere 11983. November 22, 2005 WIPER Eric An addressing error (when BIF > 1) accessing dynamic memory arrays caused the freeing of the memory that was not used to abort the task. Corrected the addressing error and the free now works. Curious... Moved nowhere. 11984. November 28, 2005 MK3IN Eric The precession change made a year ago had a typo - the output variables from the 2nd JPRECS call were supposed to be different than those of the 1st call and they were not. Moved nowhere. 11985. November 29, 2005 IMSTAT, TVSTAT Eric AU6D required that the number of iterations be > 0 as well as the Clean beam area be > 0. Dropped that excess requirement (not in IMEAN). Moved nowhere. 11986. November 30, 2005 CookBook Eric/Amy Update text of CookBook: 0 Drop recipe 2 Change EXIT to KLEENEX 3 Add QINP, drop some about deleting files outside AIPS, added KLEENEX to some EXIT references 4 Changed to emphasize model usage in VLA calibration, CLCAL discussion add INVERS for range of SN tables,, SNPLT SYMBOL adverb, baseline correction reference web page pather than defunct vlais. 5 typos, Add sentence on FLAGR, add references to RSTOR. 6 XGAUS et al under TV as well, deprectae TEK, TAPLT subtle reword 7 TV under XGAUS, UVMOD spectral index, IMMOD up to 4 sources, SPIXR paragraph 8 Paragraph MCUBE DOCONCAT = 2 option 9 www addresses, TECOR changes incl VLBATECR, VLBAEOP, CLCOR 12 Useful verbs and tasks for procedures section. 13 Update ABOUT files B Corrected CLCAL entry to SAMPTYPE BOX and warning about changed adverbs. C www addresses, added flagging info for external telescopes, EOPS and TECOR helpful procs D Fix www addresses - including change to AIPSprovided cal models I Updated Also COOKBOOK.PDF and html files. Moved nowhere. 11987. December 1, 2005 CTINI Eric Corrected CTINI - one of the keyword types was not set due to a typo. Why did this work? Anyway, also added code to close the table on error so the LUN will not stay open. Affected FITLD mostly. Moved nowhere. 11988. December 1, 2005 UVPRT Eric Change format of Weight scaling message. Moved nowhere. 11989. December 1, 2005 3c$8 X band model Eric Compressed the CC file. Moved nowhere. *********************************************************************** *********************************************************************** Changes while 31DEC05 is NEW *********************************************************************** *********************************************************************** 11991. December 4, 2005 TVFLG, SPFLG Eric The gridding routine needed a +0.001 to avoid numeric errors with computing the row to grid. Moved to 31DEC05 this date. 11992. December 6, 2005 APCAL Eric Bad call seq to PLCLOS caused Intel load modules to abort. Moved to 31DEC05 this date, nowhere else. 11993. December 6, 2005 IMAGR Eric MINPCH was allowed to change to match window sizes etc. Fixed to use a local value based on the user's MINPATCH without changing the user's value in routine CDECID in QCLEAN. Moved to 31EC05 this date, nowhere else. 11994. December 8, 2005 install.pl Eric Made grammar changes suggested by Chris Phillips (ATNF) and also added a read after an information display (of available OS flavors) so that the user has time to read the display. Moved changes to 31DEC05 this date. 11995. December 9, 2005 DSORC Eric Olaf Wucknitz (Jive) pointed out an error in the routine that calls the source changing so that the changing was done on the input not the output. Moved to 31DEC05 this date. 11996. December 9, 2005 CCRES Eric NEW TASK - like RSTOR except that it uses brute force methods. Thus it can subtract and/or add Celan components from/to any image without requiring the components to be on cells or the images to be powers of 2 in size. A model file of varying Gaussian widths may be subtracted or added also. Moved to 31DEC05 this date. 11997. December 12, 2005 TVCPS Eric Corrected it to read more than one image from disk when more than one is visible. Note that this works only when the images are on different image planes and are the same in general image size and coordinates. Moved to 31DEC05 this date, nowhere else, 11998. December 16, 2005 CPASS Eric Changed it to accept error code 3 from the solver as normal, error code 2 is reported but the answer accepted anyway (no converge), and so it dies only on error code -1 - input parameter error. Fixed help file to tell defaults and suggest that many iterations and larger tolerance is needed. Moved to 31DEC05 this date. 12001. December 29, 2005 *GEOM helps Eric The explain files for LGEOM, HGEOM, and PGEOM used the APARM nomenclature for the order of interpolation in part and then used the actual order number in the rest of the explanation. Changed it to use the APARM(n) nomenclature solely. The order is APARM(n)/2 where n=1 for HGEOM and n=6 for the other two. Moved to 31DEC05 this date. 12003. January 4, 2006 VLBAUTIL, VLBALOAD, VLBAEOPS, VLBATECR Amy Fixed VLBALOAD, the final INDXR did not work if people used the new load data from disk option. Fixed VLBAEOPS and VLBATECR to set SYS2COM and SYSOUT to '' at begining Moved nowhere - later to reopened 31DEC05. 12004. January 5, 2005 XASERVERS, UNIXSERVERS Eric There are too many sites for which the quote marks do not work in the xterm -e 'message-server ; sleep 3 ' commands and only one site that I am sure requires them. Of course the sleep 3 lets the sites that cannot use the quotes to read the message. I have dropped the quotes so some sites will now have an xterm flash by with no way to see what is wrong. Moved to a re-opened 31DEC05 this date. 12006. January 6, 2006 TVUTIL Eric Missed one variable that needs to be LONGINT - caused DOTV in IMAGR to abort on AMD 64s. Moved to 31DEC05 this date. ------------------------- PATCHES ----------------------------- 12011. January 17, 2006 DBCON Eric Fixed code to handle FQ entries that do not have the same channel width (in magnitude or sign) as the header. This matters when the reference channel is changed to 1. Moved nowhere - moved to 31DEC05 patch 21 Feb. 12014. January 19, 2006 DSMEAR Eric This VLBI de-smearing routine did not handle FQIDs correctly. It should test for a match or for either the CQ table or the calling routine FQ <= 0. It only tested for a match so when there is only 1 FQ and hence no FQID parameter, the routine failed. Made fixes to BPASET, BPGET, CSLGET to require > 0 for both FREQ IDs or a match. Moved nowhere - moved to 31DEC05 patch 21 Feb. 12016. January 23, 2006 SAD Eric Fixed the routine that computes a string of the RA/Dec in sexagesimal. The round-off terms were not reduced when the accuracy was increased - the resulting display was 0.0005 seconds of time and 0.005 seconds of arc too large. Moved nowhere - moved to 31DEC05 patch 21 Feb. 12021. January 30, 2006 WIPER Eric Corrected source number initialization to -1 since 0 is what one gets for single-source files and elevation etc require a source parameter lookup. Added a call to IAXIS1 to display axis labels whenever the full image is visible. Moved nowhere, source number to 31DEC05 patch 21 Feb. 12035. February 20, 2006 SETFC Eric Found an error in setting the boxes around NVSS sources. The X position was computed by dividing the RA difference by the cos of declination. That should be a multiplication when converting to pixels. Moved nowhere - moved to 31DEC05 patch 21 Feb. 12037. February 21, 2006 FLATN Eric Changed INTERPLATE.FOR - several intermediate variables needed to be LONGINT. Moved nowhere - moved to 31DEC05 patch 21 Feb. 12038. February 22, 2006 IBLED Eric Changed: 1. UVGET had a bad error message which had mismatched variable list and format 2. IBLED did not set error returns in the init routine and did not check for them in main 3. IBLED did not provide a sensible default for NMAPS and used numerical values for call arguments that are in/out 4. IBLED perhaps due to changes in FACGRD failed to detect missing model images (it looks for I, R, and L) and quit instead if all 3 were not present. Moved nowhere - IBLED went to patch 31DEC05 23 Aug 2006. 12047. March 6, 2006 TABF3D Eric Fixed TABF3D subroutine so that the default column count is 1. It was returning zero. Linked FITLD, UVLOD, IMLOD, etc. Moved nowhere, went to patch 31DEC05 23 Aug 2006. 12105. May 16, 2006 MBDLY Eric A call sequence got messed up with an extra argument. Caused it to have bad antenna number, time, rate, source ID and source name. Moved nowhere, went to patch 31DEC05 23 Aug 2006. 12165. August 23, 2006 CCEDT Eric Changed the automatic boxing code a bit more - removing an infinite loop situation occasioned by the old code combined with the newer code which reduced the amount that DX and DY are adjusted in each cycle. Dropped error messages for things that are really not errors. Note however that CCEDT automatic code gets different answers than it used to. Moved nowhere, went to patch 31DEC05 23 Aug 2006. 12166. August 23, 2006 IBLED Eric Found an error in which the window limits were computed as if IYP(1) < IYP(2) immediately following a computation that had them opposite! This caused error bars on phase plots to produce TV errors killing the task. Moved nowhere, went to patch 31DEC05 23 Aug 2006.