[Ifeffit] Athena Problems

[Bruce Ravel]

Hi folks,

Alistair said:

> I am having some problems with Athena.
> I reformatted my hard drive the other day (windows XP Pro), and
> downloaded Athena etc. After installing the programs, they work fine.
> However, once I download the updates, Athena refuses to work. I try to
> open the program and all I get a black command-style box filled with odd
> words/text. I do not have a virus (first thing I checked), and all the
> updated versions of the other programs all work fine (e.g. Artemis,
> Hephaestus).

Then Fabio said:

> Once, I had the same problem you reported.
> After downloading the update (ifup_15.zip file) and decompressing it I 
> noticed the installation was succesful. When opening Athena (or Artemis) 
> only the black command screen appeared.
> I went to the following path (C:\Documents and 
> Settings\your_username\Local Settings\Temp\par-your_username) and 
> deleted the content of that folder. After that, everything worked fine!

Then I went searching for the source of this problem.  I am not
thrilled about what I found.

First things first.  If you have not observed this problem, great!
Hopefully you won't.  If you are seeing this problem, try Fabio's
suggestion.  If that still doesn't help, try downloading this
executable:
  http://leonardo.phys.washington.edu/~ravel/misc/athena+artemis/athena.exe
That may improve the situation.


I should mention that, in the course of tracking down the reported
problem, I found and fixed a bug that may cause odd behavior when
importing an Athena project file on Windows (but which does not seem
to be a problem on linux).


Now on to the issue at hand.  I'll explain this as bullet points:

  * I am now using a tool called PAR (http://par.perl.org/index.cgi)
    to create Windows executables.  I switched to this from PerlApp
    (http://www.activestate.com/Products/Perl_Dev_Kit/) because it is
    free.  I could see no reason that I or my boss should purchase a
    software product that no one in my research group needs.

  * A PAR executable is actually a self-extracting, self-executing zip
    file.  It extracts itself to a cache folder in "C:\Documents and
    Settings\username\Local Settings\Temp\par-username". 

  * Each of these cache folders is about 6.4 megabytes.  By default
    they do *not* get cleaned up once you quit the application.  (That
    was news to me.  I did not know that until this afternoon!)  As
    you use Athena, Artemis, and Hephaestus, you build up a mighty
    collection of 6.4 MB folders.

  * PAR has a compiler flag for cleaning up the cache.  Doing so does
    not work well with Athena.  That is, if I use the clean-up flag,
    Athena cannot read project files.  At this time I see no way to
    reconcile those two facts.

  * I cannot include code in Athena to clean up the cache folder
    before it exits.  Windows does not allow a program to delete a
    file it is currently using.

At this time, I would say that my software is behaving like a slow but
nasty virus in that it is slowly filling up your hard disk with
useless stuff.  The best work-around I can suggest at this time is
that you occassionally delete "C:\Documents and
Settings\username\Local Settings\Temp\par-username", as Fabio
suggested. 

I can see a couple ways for this situation to be resolved.

1. The batch files used to launch the codes could be rewritten to
   clean up PAR's caches.  That should work.  Windows prevent's Athena
   from cleaning up the cache, but doing so at the level of the batch
   file should work.

2. Someone could volunteer to buy me a copy of PerlApp and then I'll
   go back to using it rather than PAR.

I should say that I am very busy this month and will not have a lot of
time to spend on this problem.  Indeed, it's not a very pressing
problem to me since the Windows versions of the codes are something I
provide only as a service to many of you.  I am not a Windows user
myself.

This whole situation really underscores the need for someone who cares
about these codes and who is a Windows user to step forward and help
out with these packaging issues.  Matt and I have made many requests
over the past year for help with the software on Windows and we have
been repeatedly disappointed.  Windows users have been relying on Matt
and I to continue providing a high level of support for a system that
neither of us understands well or enjoys supporting.  At some point,
our patience is liable to run dry.  Indeed, at some point one or both
of us are likely to just through up our hands and say "fuck it."  That
would probably not be in the interests of those of you using Windows.

B


-- 
 Bruce Ravel  ----------------------------------- bravel at anl.gov  -or-
						  ravel at phys.washington.edu
 Environmental Research Division, Building 203, Room E-165
 Argonne National Laboratory         phone and voice mail: (1) 630 252 5033
 Argonne IL 60439, USA                                fax: (1) 630 252 9793

 My homepage:    http://feff.phys.washington.edu/~ravel 
 EXAFS software: http://feff.phys.washington.edu/~ravel/software/exafs/