Category archives: Snaking
Discussions of programming, particularly of programming Python
GUI Library Evolution (Or, why we have "event" objects...)
Written by
on
in
Snaking.
This popped out of Matthew Scott's comment; seemed a little too involved to be writing in a comment (and it's a slow news day).
Context: I'd noted that Matthew's code wasn't using one of the strengths of PyDispatcher, particularly it's ability to allow for unifying parameter names and filtering out extraneous parameters before passing ...!-->!-->
And so it ends... (Pretty much, anyway)
Written by
on
in
Snaking.
Just checked in the last official changes to Cinemon. Didn't get a chance to work on the importer fixes, spent the whole day (and an extra 1/2 day for that matter) converting just about everything to the new-style UI. Need to get to bed now so I can get up early and get the shopping ...
Should I take a week off? (Unpaid, but a chance to catch up on languishing projects...)
Written by
on
in
Snaking.
I'm considering whether to take a week of unpaid vacation to work on the various Open Source projects on which I've been falling behind. Particularly PyOpenGL and OpenGLContext, but also wxOO, PyTable and BasicProperties. PyOpenGL is the project that needs the most TLC, but all of them are needing some attention.
Thing ...!-->!-->
Cinemon work draws to a close (Hard deadlines focus the mind admirably...)
Written by
on
in
Snaking.
So, at the management meeting on Wednesday we realised that we just can't afford to keep developing Cinemon with the current business arrangement. So, today is the last day of work on it until we manage to secure some funding.
The effect has been marvelous. No more working on long-term features, no more worrying about ...!-->!-->
Is Linux an infinite time sink? (Another evening disappears with no progress toward a useable desktop...)
Written by
on
in
Snaking.
Actually, some serious regression...
Knowing that there were 4 or 5 security fixes needed for the Linux kernel, I booted into Gentoo to update. Actually, I booted in to see if Eric3 could create a UML diagram for Cinemon for Tim, but once that failed (well, once Tim's finger slipped and destroyed the layout) I ...!-->!-->
PyDispatcher in the wild (Always nice to see work-product get used...)
Written by
on
in
Snaking.
"Golden Spud" has posted an introduction to his attempt to bring PyQt's signals/slots mechanism to wxPython. He's not using all of PyDispatcher's capabilitites (for instance, he's matching the consumer and producer signatures manually instead of letting PyDispatcher unify the arguments), but it is interesting to see it get used. (I didn't create PyDispatcher, I just ...
You say you slog? (What a coincidence, old chap, I've been known to slog a bit in my day...)
Written by
on
in
Snaking.
I really need to add view support to PyTable one of these days. Spent half an hour building yet another mock-view out of a classmethod. The method-based approach is nice in some ways, as you can customise all sorts of parameters to give you a nice query, but there are literally dozens of these things ...
One of those things about which I should probably be excited (Tuple spaces... why?)
Written by
on
in
Snaking.
Okay, what am I missing about tuple spaces? It seems to be a fairly basic data-sharing protocol which requires a server process somewhere. I mean, that's nice, but what are the advantages of this particular system over an RPC or formal message-passing/routing system.
If you're going to do all your networking using a (relatively ...!-->!-->
Slogs on toast (Perched on the edge of the sublime...)
Written by
on
in
Snaking.
Beyond UI cleanups and tracking down (or rather trying to track down) some bugs exposed during monkey-testing, I've been seriously pondering how to handle what amounts to a huge change in approach to the app without disturbing it too much just before the demo next week. Thing is, I would really like this approach to ...
I love to slog (Nothing better...)
Written by
on
in
Snaking.
Think I've finally got the ping scanner behaving properly. Now working on UI issues. First was cleaning up the error-log view to make the error tracebacks a little more pleasantly formatted. Second is testing the node-range mapping import. It seems to be working fine still, but it's linked to network-object views being the primary focus ...