I think this illustrates well the dangers of blind reliance on automation - you cannot assume that decisions made by automatic scripts are necessarily the most sensible ones! It goes without saying that the design of robust automatic scripts is a non-trivial task. In this case the script clearly should not make the decision without further checks that simply because a file exists with the right name a) that it is the same one that it created on a previous run and b) that it contains valid data, and therefore that it's OK to use it. If you're lucky the file format will be invalid and the program will simply exit abnormally, as happened here. If you're unlucky, the file format will be valid and the program will run apparently to normal termination but with garbage data, and the usual "garbage in, garbage out" will apply! The question of course then is how much time will you have to spend as a consequence trying to figure out what happened (and will you ever realise that the fault lay with the script)? I think it's incumbent on the script to do at least one of the following:
1. Make robust checks that the file format is valid and does contain the right data - unfortunately in this case probably the least practical option, or
2. Always assume that Sod's law applies (if anything can go wrong it will, and moreover will do so in a way that maximises your inconvenience), i.e. don't re-use any intermediate files and re-run the whole thing from scratch, or
3. Place the burden of making decisions concerning re-use of intermediate files entirely on the user - this of course means that the script can hardly be called "automatic"!
Given the choice I know which option I as a user would choose every time (i.e. #2), after all what's a few hours of CPU time against even the remote possibility of me wasting days, weeks or even months of my time!
Have fun!
-- Ian
-----Original Message-----
From: Eleanor J. Dodson [mailto:ccp4@ysbl.york.ac.uk]
Sent: 30 April 2003 10:02
To: raji
Cc: ccp4bb@dl.ac.uk
Subject: Re: [ccp4bb]: AmoRe Fails to Run !
I think you probably have an incomplete or old version of the hkl file
... Try deleting /mnt/data1/kllab/raji/gh4pa7-home/ccp4i/gh4pa7z-p1.hkl
then rerun auto amore
Eleanor
raji wrote:
> Hi Everyone,
> This is my very first time using CCP4i and I am trying to run AutoAmoRe.
> After several aborted trials foll. memory allocation problems, now the trial
> has terminated with the foll. message in the logfile I cannot quite
> understand.
> ***************************************************************************
> The program run with command: amore HKLPCK0
> /mnt/data1/kllab/raji/gh4pa7-home/ccp4i/gh4pa7z-p1.hkl CLMN0
> /mnt/data1/kllab/raji/gh4pa7-home/ccp4i/gh4pa7z-p1_0.clmn MAPOUT
> /mnt/data1/kllab/raji/ccp4/gh4pa7z_10_rot.map TABLE1
> /mnt/data1/kllab/raji/gh4pa7-home/ccp4i/mr_MR_trial...tab CLMN1
> /mnt/data1/kllab/raji/gh4pa7-home/ccp4i/mr_MR_trial...clmn HKLPCK1
> /mnt/data1/kllab/raji/ccp4/gh4pa7z_10_3_hkl.tmp ROTING_MI 1000000 ROTING_MC
> 4400000
> has failed with error message
> start: end of file
> apparent state: unit 7 named
> /mnt/data1/kllab/raji/gh4pa7-home/ccp4i/gh4pa7z-p1.hkl
> lately reading sequential unformatted external IO
> ***************************************************************************
> #CCP4I TERMINATION STATUS 0 "start: end of file apparent state: unit 7 named
> /mnt/data1/kllab/raji/gh4pa7-home/ccp4i/gh4pa7z-p1.hkl lately reading
> sequential unformatted external IO"
> #CCP4I TERMINATION TIME 29 Apr 2003 21:57:52
> #CCP4I TERMINATION OUTPUT_FILES
> /mnt/data1/kllab/raji/gh4pa7-home/ccp4i/mr_MR_trial...tab gh4pa7z
> #CCP4I MESSAGE Task failed
> ------------------------------
> Can someone help?
> Thanks a ton,
> Raji
>
> Raji Edayathumangalam
> Dept. of Biochemistry & Molecular Biology
> Colorado State University
> Fort Collins, CO 80523 USA
> Tel:(970)491-4614
> Fax:(970)491-0494
>
>
>
CONFIDENTIALITY NOTICE
This email contains confidential information and may be otherwise protected by law. Its content should not be disclosed and it should not be given or copied to anyone other than the person(s) named or referenced above. If you have received this email in error, please contact I.Tickle@astex-technology.com.