Category archives: Snaking
Discussions of programming, particularly of programming Python
Memory leaks in TwistedSNMP (Tracking things down that shouldn't leak...)
Written by
on
in
Snaking.
Seems as though there's something leaking memory in TwistedSNMP (I've been tracking down Cinemon's leak). Doesn't appear to be particularly large, but with the number of queries Cinemon does, could be a contributor. Haven't yet figured out where the silly error is inside TwistedSNMP (more precisely, I've determined which test cases are leaking, I just ...
Words formed from serendipitous (Natasha's obsession...)
Written by
on
in
Snaking.
Natasha caught wind of Alan's wife's "serendipitous" words challenge. She's apparently up to 250 words... running the word-finder script, however, says that there's something like 1845 words in total (though that includes abbreviations and proper nouns)... hopefully it's not distracting her from writing the novel.
Still a leak... (Though much slower now)
Written by
on
in
Snaking.
Cinemon still has a leak somewhere. It's pretty large still, about 6MB/hr. That should exhaust system memory in about two weeks and cause a reboot :( . Much better than 19 hours, but nothing like acceptable (no leak is acceptable). Grr.
Next most likely culprit would be PySNMP's message cache, I suppose. After that, possibly ...!-->!-->
A new PyDispatcher release (Hopefully this fixes the memory leak in Cinemon..)
Written by
on
in
Snaking.
Found a memory leak in how Dispatcher deals with it's back-reference table. Basically it would allow you to replace a registratation of a receiver, but would not remove the replaced receiver's back-references. It also wasn't removing the back-reference when a signal was explicitly disconnected.
Both of those operations are fairly rare, but Cinemon does them ...!-->!-->
Thinking as I try to find the Cinemon leak... (Considerations as the test-suite trundles)
Written by
on
in
Snaking.
I think it's time to move PyOpenGL off of SWIG and on to ctypes. Rene did some preliminary work on this... well... it's a while ago now. The basic C-style API isn't that hard, it's replicating the Pythonic API that causes the problems. It's likely worth the pain just to reduce build problems, the flexibility ...
Oh, the day in hotspots (Suppose I should mention the real work...)
Written by
on
in
Snaking.
Think I've gone about as far as I can go without knowing what I'm trying to build for the hotspot project. I put up a quick framework, started filling in page templates with approximately what's needed, but without knowing what business information is needed/wanted, I can't really write the code to go behind the templates. ...
wxoo fixes (Wooing developers into using BasicProperty, PyTable and wxoo)
Written by
on
in
Snaking.
Realised when asked a few days ago that I'd left wxoo sitting right in the middle of a major rewrite. The rewrite was necessitated by the wxPython 2.5 API changes. Those changes were... let's just say "nontrivial", and unfortunately, the new wxPython APIs changed as I was in the middle of doing the upgrade (reversing ...
Linux rises from the grave of the dead tablet (Well, almost...)
Written by
on
in
Snaking.
Decided to give another shot at fixing the wacom tablet in my Gentoo install. I've got a fix, but it's pretty ugly. The problem was that the tablet would, when the pen approached it, start running the cursor along either the left or right side of the screen up and down (very fast). From the ...
Hmm, last post sounded very negative... (Don't post before the first coffee of the morning)
Written by
on
in
Snaking.
Looking at the problem again, it seems obvious that there is a story trying to get through the mess of the presentation;
"Descriptors: from functions to decorators".
That is, a sort of creation myth of the generalisation of the wart/exception in Python 2.1 into a generic, but somewhat difficult to use mechanism in Python 2.2 ...!-->!-->!-->!-->
Feedback on the presentation (Apparently far worse from the other side...)
Written by
on
in
Snaking.
Got some feedback from John via email. In summary, needed to be designed more, have less information (always one of my problems), and use more anecdotes.
All good points. Doing it again, I'd probably have focused solely on descriptors. Don't know about his preference for story-centred organisation. I tend to come from the "theory first, ...!-->!-->