Hide Forgot
abrt version: 1.1.17 architecture: i686 Attached file: backtrace, 3338 bytes comment: I was running hg clone against the Mozilla auroroa repo component: mercurial executable: /usr/bin/hg kernel: 2.6.35.11-83.fc14.i686.PAE package: mercurial-1.8.2-1.fc14 reason: demandimport.py:76:__getattribute__:AttributeError: 'module' object has no attribute 'parse_index' release: Fedora release 14 (Laughlin) time: 1302876463 uid: 500 How to reproduce ----- 1. hg clone http://hg.mozilla.org/mozilla-aurora
Created attachment 492543 [details] File: backtrace
It is hard to imagine how that can be caused by Mercurial. Do you have any extensions enabled? Can you try to disable them all (and perhaps update them or file a report elsewhere)?
(In reply to comment #2) > It is hard to imagine how that can be caused by Mercurial. Do you have any > extensions enabled? No no extensions, I'm a very simple user of Mercurial using only what I get from the Fedora package. I use it only to contribute translations to Mozilla. My guess is that the shear size of the clone might have been a reason. > Can you try to disable them all (and perhaps update them or > file a report elsewhere)? I've managed the clone the directory subsequently. I'll leave it to you to decide if this is worth pursuing or upstreaming. I won't have time to do that. And I won't be offended if you decide its not worth it :)
The funny thing is that Mercurial in version 1.8 replaced parse_index with parse_index2. So the reason you got that error must be that there are some outdated code somewhere. It would be nice to know what the reason is, and if it turns out to be a real problem that can be reproduced it would be worth pursuing. But taking a closer look at the stacktrace I can see that you do have the color extension enabled. Which extensions do you have enabled (hg showconfig extensions)? But I can also see that the function name in the stacktrace doesn't match the line numbers. That indicates some serious trashing. Could it be a memory corruption due to bad hardware (or cosmic radiation) or other problems? Can you reproduce the problem consistently? It is a big repo, but I can clone it without any problems.
(In reply to comment #4) > The funny thing is that Mercurial in version 1.8 replaced parse_index with > parse_index2. So the reason you got that error must be that there are some > outdated code somewhere. It would be nice to know what the reason is, and if it > turns out to be a real problem that can be reproduced it would be worth > pursuing. > > But taking a closer look at the stacktrace I can see that you do have the color > extension enabled. Which extensions do you have enabled (hg showconfig > extensions)? extensions.hgk= extensions.color= > But I can also see that the function name in the stacktrace doesn't match the > line numbers. That indicates some serious trashing. Could it be a memory > corruption due to bad hardware (or cosmic radiation) or other problems? I don't use the machine that much anymore, just for Firefox translations. Could very well be that it was busy upgrading Mercurial. > > Can you reproduce the problem consistently? It is a big repo, but I can clone > it without any problems. I haven't tried to reproduce. I tried another clone and it worked.
Simultaneous upgrade of Mercurial would be a very good explanation.