[daip] CVEL in AIPS

[Eric Greisen]

Ross Burns wrote:
> Dear Eric
> 
> Thank you for responding so quickly.
> 
> On 18 December 2013 07:13, Eric Greisen <egreisen at nrao.edu 
> <mailto:egreisen at nrao.edu>> wrote:
> 
>     Let me ask a question - the 31DEC13 version that you are using -
>     when did you last update it (with the MNJ)?  A bug was put in the
>     code March 12, 2012 and corrected after October 13, 2013 (correction
>     put in both 31DEC12 and 31DEC13 then).  This bug was tricky since it
>     affected 32-bit Linux only (binary installations) - the very real
>     error in the software did not cause the 64-bit Intel compiler or
>     those on Macs or GNU to do anything bad.  So any 31DEC12 version
>     taken between these dates or 31DEC13 version taken before October 21
>     will have an error.  If this does not affect you please let me know
> 
> 
> We've all run into the same issue using both 32-bit and 64-bit machines. 
> Most, if not all, of us here are running Ubuntu Linux, though I have 
> also had the same problem on my Mac OS X. I dont think any of us have 
> used the MNJ, probably from fear of change or a "Dont fix what isnt 
> broken" mentality - which is fine when everything works. I guess we are 
> being punished for being wimps. So I would suppose (with regards to my 
> lab members) each person's AIPS version differs depending on when he/she 
> downloaded/installed it. Like I said, our old procedure for CVEL worked 
> fine until about Jan2013 at least (some members are still using ~2009 
> AIPS versions), and we only saw the problem appear at around May2013. 
> The problem is still seen in recent downloads of DEC12 and DEC13. It 
> seems that the introduction of the bug you mention did not cause us any 
> problems because I had a good 31DEC12 downloaded and installed at around 
> Jan'12, but that our problem could have started around the time that the 
> bug you mention was fixed, maybe the patch? What do you think?
> 
> Kind regards, and thanks again for your help.
> Ross

Almost all computers are 64 bit these days - the question I am really 
asking is what architecture of load modules are you running LINUX or 
LNX64?  The former had the problem and I am uncertain whether Macs had 
the issue or not.

There was also a correction made to SETJY September 21, 2013 and patched 
into 31DEC12 - the radio velocity stored in the SU was wrong because the 
system velocity was not added in.  This error probably started in 
September 2012.

Could you please try a MNJ on a 31DEC13 version (or install 31DEC13 over 
again now) and see if this error is still occurring?  If it is, I will 
need to have a copy of your UV data to try this myself (a small piece 
will do).

Note - if you have had bugs since May it would have helped if you had 
asked about them then - although nothing happened anywhere near May to 
any of this code.

Cheers

Eric Greisen