Rektide, I told you exactly what has to happen if you're interested in pursuing this:
>>> If you think that such a thing absolutely MUST be provided by the core library, then that belongs in a github issue. Be prepared to make a VERY strong case for it. It's a huge change to the architecture, which was tried in the past and turned out to be not a good idea. You'll need to be patient (since it will take a while), be willing to help work on it, and have a plenty of new data showing that it's actually a valuable change to make in the core, and why the problems encountered last time can be overcome. I can tell you that it won't have any chance of landing before 2.0, and you'll have to (at least) convince Ben Noordhuis and Bert Belder that it's worth doing. They're both rather skeptical right now, and stubbornly fixated on "evidence" and "use cases" and "reasonable trade-offs", so that'll be an uphill climb, but I've found them easy to convince if you can provide these things. If you convince them, that's probably the best way to convince me. >>> Familiarizing yourself with the node and libuv source code would be a good place to start. You should treat this as an experiment: your goal could be to figure out why several people close to the Node project are so against a threading API, or resurrecting Isolates. Collect data about what is possible using threads, which you can't do now using child processes, and what you'd have to do to make libuv thread-safe. Experiment with TAGG, and perhaps see how far you can get with a purely userland approach to the problem. Figure out what the limitations of TAGG are, and where you'd actually start implementing it. If you come back with sufficient data and understanding, I'd be happy to reopen this issue. I think it's more likely that, armed with that data, you'll see that threads are just not a good fit for a JavaScript platform, at not worth the trade-offs required in libuv. But you shouldn't not do everything people tell you not to do. Sometimes you just gotta try the crazy thing ;) > I humbly propose not participating in things which annoy you or you find > overly distracting. > Isaac: I fully confess to being an ignorant childish distracting useless > retrogressive drag on the community, and I'm sorry if this discussion in fact > hurts us. Also, seek trained medical professional help. These comments are trollish and inappropriate. On Tue, Sep 18, 2012 at 11:51 AM, rektide <[email protected]> wrote: > >> Multi-threading, even without zero copy, would allow for faster context >> switching as well. Context switching is bad, process-per-thread is good and >> helps avoids context switching, but there is a gain even if there is not an >> easy zero copy benefit to be had. > > > *facepalm* s/process-per-thread/process-per-core/ > > -- > Job Board: http://jobs.nodejs.org/ > Posting guidelines: > https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines > You received this message because you are subscribed to the Google > Groups "nodejs" group. > To post to this group, send email to [email protected] > To unsubscribe from this group, send email to > [email protected] > For more options, visit this group at > http://groups.google.com/group/nodejs?hl=en?hl=en -- Job Board: http://jobs.nodejs.org/ Posting guidelines: https://github.com/joyent/node/wiki/Mailing-List-Posting-Guidelines You received this message because you are subscribed to the Google Groups "nodejs" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/nodejs?hl=en?hl=en
