On 15 Jan 2003, seth vidal wrote: > > You don't mention yum version, but in the latest, > > clientStuff.py at line 173 looks to be uncompressing the > > header payload of something local, and has an uncaught > > exception, which triggered the traceback -- [dumping a stack > > trace is a 'pythonism' default, when there is an uncaught > > exception, similar to leaging a 'core' file] > > he was running 0.9.3 - in the fine print. > gzip was returning a crc error from a header he downloaded - b/c the > packages are handled by gzip - only the headers. heh -- As the author of the code, you know the codepaths already -- I was desk tracing, and scoped it out <grin>. RFE: Can the exception be caught, and a more helpful error message to identify the corrupt file (or maybe, emit a warning, and set a flag to rm and repull the offending file a time or two [it might be corrupt in the remote archive, so a limit on repull attempts, and a notation of the problem file make sense here])? <how about _that_ awkward compound sentence?> -- Russ Herrold
