[lxml-dev] Any estimates for next lxml release?

Hello guys, Are you planning to release a new version soon? Several segfaults fixed in the dev version are really useful for me and I would feel better if I used an official release (I'm using lxml 0.7 in production) than an snapshot. Thank you, -- dharana

dharana wrote:
A release is long overdue; this is mostly because I've been really busy and I'm the only release manager of lxml so far. Thanks for the prodding. :) There's one bug concerning iterators that I'd like to have fixed before we do a release. I'll try to find some time next week to look into it, and push out a release. Note that I was also hoping for windows installer work being done by some volunteer before then, but I guess if nobody steps up that can also be done by someone afterwards. More broadly on the installation front, I recently looked at python eggs/setuptools/easyinstall and it'd be really cool if we could get lxml into an egg so it'd work with easyinstall. Regards, Martijn

On Fri, 2005-09-23 at 18:32 +0200, Martijn Faassen wrote:
That's the case why I'd like to see any bugtracking system for lxml installed. Could you please explain bug? May be I or someone else can help about it?
I remember, I volunteered for that work, but now times changed and I do not have to deal with windows any more at all, so I guess, that should be someone else to do it.

Andrey Tatarinov wrote:
You're right. I need to request this from the codespeak admins. I'll have a chat with them next week (I'm at a zope 3 sprint in Austria right now and a bit distracted).
Could you please explain bug? May be I or someone else can help about it?
See the 'Python interpreter segfault setting element tail' thread. What happens is that the iterator retains the next node (not element) in the iterator. If a tail gets set, this next node is not necessarily valid anymore (perhaps it's removed or a text node gets added). The iterator still refers to the original node and may iterate off into deep space. I tried to fix the bug but it was a bit more involved than I hoped for. Retaining the element that was *last* iterated to might work, but getting everything to work the way expected (have the tests still pass etc) is not trivial.
Okay -- a numer of people volunteered in fact. Good for you that times changed. :) On the general install front I'm in a discussion on distutils-sig about making an egg out of lxml, so those interested could check with that. Regards, Martijn

I recommend using JIRA ( http://atlassian.com/software/jira/ ). It's free for opensource projects and a charm to work with. I use it myself and I also know that at least the apache group uses it for some of its projects: http://issues.apache.org/jira/secure/Dashboard.jspa Martijn Faassen wrote:
-- dharana

Speaking of bug trackers, I'd prefer to use trac (http://projects.edgewall.com/trac/) for its great integration with Wiki and Subversion, which is currently used. On Sat, 2005-09-24 at 16:43 +0200, dharana wrote:

dharana wrote:
A release is long overdue; this is mostly because I've been really busy and I'm the only release manager of lxml so far. Thanks for the prodding. :) There's one bug concerning iterators that I'd like to have fixed before we do a release. I'll try to find some time next week to look into it, and push out a release. Note that I was also hoping for windows installer work being done by some volunteer before then, but I guess if nobody steps up that can also be done by someone afterwards. More broadly on the installation front, I recently looked at python eggs/setuptools/easyinstall and it'd be really cool if we could get lxml into an egg so it'd work with easyinstall. Regards, Martijn

On Fri, 2005-09-23 at 18:32 +0200, Martijn Faassen wrote:
That's the case why I'd like to see any bugtracking system for lxml installed. Could you please explain bug? May be I or someone else can help about it?
I remember, I volunteered for that work, but now times changed and I do not have to deal with windows any more at all, so I guess, that should be someone else to do it.

Andrey Tatarinov wrote:
You're right. I need to request this from the codespeak admins. I'll have a chat with them next week (I'm at a zope 3 sprint in Austria right now and a bit distracted).
Could you please explain bug? May be I or someone else can help about it?
See the 'Python interpreter segfault setting element tail' thread. What happens is that the iterator retains the next node (not element) in the iterator. If a tail gets set, this next node is not necessarily valid anymore (perhaps it's removed or a text node gets added). The iterator still refers to the original node and may iterate off into deep space. I tried to fix the bug but it was a bit more involved than I hoped for. Retaining the element that was *last* iterated to might work, but getting everything to work the way expected (have the tests still pass etc) is not trivial.
Okay -- a numer of people volunteered in fact. Good for you that times changed. :) On the general install front I'm in a discussion on distutils-sig about making an egg out of lxml, so those interested could check with that. Regards, Martijn

I recommend using JIRA ( http://atlassian.com/software/jira/ ). It's free for opensource projects and a charm to work with. I use it myself and I also know that at least the apache group uses it for some of its projects: http://issues.apache.org/jira/secure/Dashboard.jspa Martijn Faassen wrote:
-- dharana

Speaking of bug trackers, I'd prefer to use trac (http://projects.edgewall.com/trac/) for its great integration with Wiki and Subversion, which is currently used. On Sat, 2005-09-24 at 16:43 +0200, dharana wrote:
participants (3)
-
Andrey Tatarinov
-
dharana
-
Martijn Faassen