On Sun, May 07, 2006 at 03:22:48PM -0400, Mathieu Bouchard wrote:
On Sun, 7 May 2006, geiger wrote:
However, many other open-source projects are intensely more collaborative than Pd, in such a way that (the core of) Pd, jMax and MAX/MSP look about as collaborative as each other.
I don't have a good metric for this, but it seems that since the patch tracker was started, many more patches have gone into Miller's Pd than used to. It also seems like he has rejected some patches for technical reasons and instead of fixing those technical faults and re-submitting, people have just let the patches languish. One of the ones that makes me sad is your tool tips patch - how cool it would be if inlets had tooltips!
he includes patches when time permits and also explains why he doesn't include others.
I haven't found his explanations to be always particularly explaining, and I don't expect them to always be, but there are some biggies for which his explanations were even more disconcerting than the lack thereof.
Linus Torvalds and Andrew Morton often drop patches completely silently. Andrew Morton gave a really good explanation of why this is neccesary at Linux.conf.au a couple of years back. It's the right of the "benevolent dictator" to accept or reject submissions for whatever reason. And it's the right of the developer to fork if they don't like the direction the mainline is going in. It's also their right to be too busy to explain things, though as we see with Pd this might discourage some people from contributing.
Perhaps it is reactions to his "disconcerting explanations" that is putting him off from giving more detailed explanations? Who knows. My point all along has been that a little bit of politeness and respect can go a long way.
I don't mean just the patch submission system, I also mean discussions that lead potential patch submitters to decide whether submitting patches is worth doing.
Yeah, it seems like Miller doesn't always indicate the development direction he is going in, and if he did that it might help things along. This could be because he isn't employed by someone like the OSDL and is pretty busy as a researcher. Then again, there are a heap of tasks in the TODO file which people seem to be ignoring and working on other stuff instead, often stuff that is quite perpendicular to the development direction of Pd main.
Best,
Chris.
chris@mccormick.cx http://mccormick.cx