Hey Dan, Thanks for taking the time to write all of that out. Why don't we start a new thread to discuss the timing of merging your work in? It's already in a fairly usable state, so having it in the default branch for active development can get a lot more eyes on it. There are a lot of Linux distros where AVbin just doesn't work at all anymore, so it would be a big improvement even in the current state of development.
Regarding Python 3, I think that one point to keep in mind is that we don't actually have to change anything immediately. In fact, I think that it's probably best that we don't do any major rewrites. Instead, we just free ourselves of the burden of writing and testing under Python 2. We can then start cleaning up and modernizing modules as things slow down. On Sunday, June 18, 2017 at 5:58:04 PM UTC+9, Daniel Gillet wrote: > > Woaw, just found out this thread by coincidence. Great things happening > here! Lots of positive and creative thoughts. I like it. > > Regarding the FFmpeg bindings, it probably does not care too much about > py2 or py3. The initial goal was to merge it with the current branch 1.X. > > As Benjamin said I kind of stuck with some synchronization issues. And > it's terribly annoying because I'm unsure what is going on. And to make > things even more complicated it's Benjamin who experience these problems > (on Linux) while I don't necessarily notice them on Windows. I don't have > Linux here so it's hard to debug for another OS... After all my researches > so far, it seems that on both Windows and Linux frames get discarded once > in a while because the function to display the next frame is called too > late. But looking at the CPU, it's not really that busy. So I'm starting to > wonder if it's not related to pyglet scheduling mechanism which might not > be as accurate as needed for this case? But I don't find a way to prove (or > not) this theory. I might be completely wrong. And I would gladly be, > because it would push me in another direction. > > While doing this binding, I tried as much as I could to make the least > changes to the current code for the media player. But something is > inherently wrong with the current implementation. They basically choose to > synchronize the image with the sound. So if the sound is not played at the > right speed, this could cause also this jittering. The right approach is to > synchronize both sound and image to an independent clock. But this requires > to change many things in the media player. > > Just to say that I could accept some help if someone has some time. > Talking about the problems and deciding on the best way to fix things would > be helpful. Oh, and time... that's another issue right now. I'm pretty busy > and beginning of July I have some vacation. After that it should be better. > > Sorry if I derailed the topic. > > So coming back to the main question, I'm not an expert in OpenGL 3, but I > can see lots of benefits pushing pyglet in that direction. Regarding > support for Python 2, If we make a new branch, maybe it's not a bad idea to > drop py2 support. It's anyway supposed to lose support in 2020, right? > That's in about 2+ years. Now I'm also sure there are probably a > substantial amount of projects based on pyglet and still using py2. The > question is how many are still *active*? On the other hand, releasing a > python3 only version is a strong message towards the community and I think > it's a good thing. Personally I have no problem going forward with python 3 > only. > > Dan > -- You received this message because you are subscribed to the Google Groups "pyglet-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at https://groups.google.com/group/pyglet-users. For more options, visit https://groups.google.com/d/optout.
