Category archives: Snaking

Discussions of programming, particularly of programming Python

RSS feed of Snaking

Embrace the Singularity: Sunday Bahen 1180

Presentation is shaping up nicely. I'm still way-over on slides (71), but that's normal for me at this stage. I'll do some ruthless cutting as I start into full read-throughs. The biggest thing that's shown up in reviews so far is that probably 15 of the slides are fodder for an hour's discussion... and we'll ...

Continue reading

Has it really been 18 years since I did Java?

I've definitely had to touch Java in the intervening years, but the last big Java project I can remember working on was in 1997. And that was a VR Mutech that we ripped out and replaced with C++ and Javascript ASAP. Java code always seems somewhat like someone is pulling my leg...

  • You wrote 43 ...

Continue reading

That point where you wish you'd asked for an hour

There's a point in every presentation where you realize "I should have asked for an hour (or 8)"... that's about 1/2 way through in my experience, it's where you've filled the work-table with a thousand items, taking each bullet point you wanted to cover and expanding it to full-lecture size, and now you start the ...

Continue reading

Finally moving to Django migrations

So it has taken far too long to get this done. As our migration operations were tightly coupled to South migrations, South was holding us back for far too long. Today I've pretty-much finished the migration. I wound up embedding a virtualenv inside the firmware for each product that has stub packages, the south migrations, ...

Continue reading

Multiple "output" parameters in PyOpenGL

So Moisés has been sending me bug-fix patches for PyOpenGL, and todays (actually Sunday's) pointed out that the multiple-return-parameter case was plainly broken in the PyOpenGL wrapper code. Basically the code would only return the *last* parameter which was setOutput()'d so that the low-level wrappers would return somewhat pointless values rather than returning the tuple ...

Continue reading

Neural Nets on the CPU Are (as Expected) Pointless

I finally got my OpenDeep Dataset implementation for the TEDLIUM dataset feeding into a basic generative RNN... and as expected, it's pointlessly slow. Expect to go, say, 100 epochs... and it's already taken ~ 2 hours on the first epoch. Seems like I really do need to get an Amazon account and use that to ...

Continue reading

Just got my Pycon.ca Ticket

PyCon.ca tickets are now on sale. $99 for self-sponsored tickets, $49 if you're a student, 2 days conference plus 1 day sprints, November 7th and 8th. IIUC they've sold out every year, so you likely want to get your order in soon. (Obviously I waited until after I had my tickets to advertise that they ...

Continue reading

Old School GPGPU is Frustrating

So I burned the evening on trying to get a working GPGPU implementation for my desktop/workstation and laptop.

Started off with PyOpenCL, but wound up blocked on that due to libcffi failing with some missing headers. So switched back to playing with raw GPGPU on GLSL. The actual setup/GL part went well, but when I ...

Continue reading

Raspberry Pi PyOpenGL

So since Soni got me to setup raspbian on the old raspberry pi, I got PyOpenGL + GLES2 working on it today. There is a bug in raspbian's EGL library (it depends on GLES2 without linking to it), but with a work-around for that bzr head of PyOpenGL can run the bcmwindow example/raw.py demo. I ...

Continue reading