It seems that Luke has ended up in Ben's explicit or implicit killfile. I think that's unfortunate. Therefore, I will make an attempt to summarize what I see as the issues in this thread. I'm not going to add any new ideas, so feel free to skip everything after this paragraph if you feel you understand the issues. But if you do so, please take this as a vote to consider seeking better solutions than just orphaning pyxpcom.
As I understand the debate here, everyone acknowledges that maintaining pyxpcom would be several times easier if done by Mozilla developers as part of the release cycle, with automated testing, than by Luke from his position outside Mozilla. But, while Luke feels that such maintenance is worth the higher amount of time, Ben and Mozilla do not feel that it is worth the smaller amount of time. But that's not the actual debate. The actual debate, the reason Luke is tearing his hair out and others have classed him as a troll, is that Luke (the person actually doing the work) believes that Ben (the person who doesn't want his team to be responsible for the work) is overestimating how much work it would be. That is because some of the initial reasons Ben gave for not wanting to do the work are inaccurate in Luke's estimation. Luke pointed this out, and evidently feels that Ben should acknowledge his mistakes. The Mozilla team seems to feel that Luke is just whining; as if A asks B to the dance, is told "I have to walk my dog", and then repeatedly offers B other dogwalking solutions despite being pointedly ignored. The point is that, as with the dog, the question of exactly how difficult it would be for Mozilla to maintain pyxpcom doesn't actually matter. <arbitrary illustrative numbers>Whether the job is 10 times easier for Mozilla, as Luke believes, or only 2 times easier, as Ben believes (or believed last time he responded),</arbitrary illustrative numbers> the fact is that it is perfectly possible that Mozilla still decides it's not worth it and Luke still decides it is. And if the question of relative ease were resolved, the debate would immediately move on to discussing importance. (I happen to agree with Luke on the importance question, even though I'm not knowledgeable enough to take sides on the technical question. But that's beside the point.) Debating isn't going to resolve this. How do such questions of comparative advantage usually get solved? Through some kind of trade. But it seems unlikely that money is going to change hands here. If Mozilla does this for Luke, is it likely that he will eventually someday find some way to repay it through coding? I'd say it's very likely, even though we can't point to an appropriate coding payback task today. I can say that, I really can't see why it wouldn't be worthwhile to give Luke a full hearing, actually responding to his technical points. Yes, he's socially inept, and it's hard to keep an even temper when he's flying off the handle, no matter how justified he may be by his lights. But at the end of the day, it would be a huge, huge pity if his social ineptitude caused Mozilla to make the wrong decision. Even if he is a troll, he's also a damn good engineer, and that caliber of troll isn't so common that it's a huge task to respond to all of them. And if, on the other hand, he's even half right, then the work maintaining pyxpcom will be repaid several-fold by: broadening the community (with pyjamas and possibly sugar); an additional code-path to catch bugs sooner; having Luke himself, a smart engineer, helping you instead of annoyingly arguing with you; and finally, far from least of all, all the (in my opinion) extremely impressive things that could be done with the technol ogy itself if it weren't for the festering bitrot. OK, enough of my kibbutzing. I hope that my good-will contribution is appreciated. No offense intended to any of the parties involved. Jameson _______________________________________________ dev-embedding mailing list [email protected] https://lists.mozilla.org/listinfo/dev-embedding
