problem after installing patch 2.15.1

giannicatanzaro · Joined: 08 Mar 2011 Posts: 10

dear All

after installing 2.15.1a patch on my IRAF 64bit (run on openSUSE 11.3) I got a very curious problem. It seems that there is a problem with the number format, for instance the aperture and order are 20 digits numbers and wavelengths are exponentials.

anyone can help me????????

fitz · Posted: Wed Mar 23, 2011 3:53 pm Post subject:

What task are you using? Is this a plot label, text output, or is the aperture header being being modified?

giannicatanzaro · Joined: 08 Mar 2011 Posts: 10

fitz · Posted: Thu Mar 24, 2011 4:09 pm Post subject:

I think you may be seeing the effects of

http://iraf.net/article.php/20110309004941989

If you installed the patch in the last few days it may be something else, if it was done around the time the v2.15.1a patch was first released then re-applying the patch may solve it.

StarryEyed · Joined: 24 Mar 2011 Posts: 6

giannicatanzaro · Joined: 08 Mar 2011 Posts: 10

fitz · Posted: Thu Mar 24, 2011 6:14 pm Post subject:

To rule out problems in installing the patch, please do a

giannicatanzaro · Joined: 08 Mar 2011 Posts: 10

fitz · Posted: Sat Mar 26, 2011 5:00 am Post subject:

It seems this is a continuation of the earlier problem, i.e. the first fix wasn't correct in this case. I think it is finally fixed now, you can either try to apply the patch file again (sorry for the inconvenience) or else get just the patched x_onedspec.e binary file from ftp://iraf.noao.edu/iraf/v215/support/linux64/x_onedspec.e and install in the noao$bin.linux64/ directory.

Let me know if you continue to have problems.

giannicatanzaro · Joined: 08 Mar 2011 Posts: 10

dvbowen · Joined: 23 Jun 2006 Posts: 5

Just a quick note... I spent the morning working on this problem too, before I found this post. I copied the x_onedspec.e file you linked to, and it looks like the fix does indeed work (yay!). I thought it might be worth mentioning though that the version of IRAF I recently installed is labelled "IRAF V2.15.1a February 2011" so should be a version with all patches included? It looks like I picked up iraf.lnux.x86_64.gz on March-17 2011, so I wonder whether the fix really *is* the most up-to-date version?

Thought the info might be useful!

Cheers

Dave

fitz · Posted: Fri Apr 08, 2011 7:56 pm Post subject:

That is the most up to date version, I simply forgot to change the date on the version strings in all the rush to put the patch together.

nancy-mac

Some time ago, you kindly fixed a problem I, along with some other people, was experiencing with splot in 2.15. Unfortunately, at the time I was so delighted when your fix worked that I didn't pay attention to the 15-digit aperture numbers. The aperture editor provides sensible aperture numbers and writes them in the database file, so I reduced a night of spectroscopic data up to the ecidentify stage, at which point I started getting arithmetic exceptions, segmentation violations, etc.

Now, I've installed the March 11, 2011 version of 2.15.1a. I still get the 20-bit aperture numbers with splot or ecidentify, but the behavior of ecidentify has changed. You can mark features, but in response to an 'f' command or a :read it misbehaves. Before, it crashed with a segmentation violation or an arithmetic exception. Now, it just does nothing.

I am running 2.15.1a under OS 10.6.6 on a 64-bit MacBook Pro. The version test you suggested gives the following.

fitz · Posted: Tue Jun 07, 2011 11:10 pm Post subject:

I checked the binaries in the current distribution files and they seem to still be correct (or at least, they're the same as what was mentioned earlier in the thread). FITSVERIFY does show a number of errors on the FITS files themselves but the pixels look reasonable and I can otherwise use the files.

What DOES seem off however are the image names: IRAF doesn't generally like images with only numbers as names (and '21x' is interpreted as a hex value for 33) and tasks that create database entries or temp files may only use the root name which leads to other problems. The first thing I'd try is to rename "21.fits" to "foo.fits" to see if it makes a difference.

If not, please post the parameters and details of what you did so I can reproduce the problem. If you still have problems I'll have to pass this on to Frank as being the expert.

nancy-mac · Posted: Sun Jun 12, 2011 2:50 pm Post subject:

You are correct! Simply changing the file name from '21.fits' to 'f21.fits' solved both the aperture number problem and the ecidentify problem, at least at first glance. Thanks!