So we need a new maintainer? Or a team?

[qfox]

Since @zaach is busy...

Can we get a team with many people who can review branches, merge requests, and publish new versions?

[nolanlawson]

I'm happy to take this up, if @zaach is willing to delegate to me. I can take the time to cleanup the website, merge PRs, review stuff... I don't have infinite time, but I really love Jison and would be happy to help. :)

[qfox]

btw, I've found another fork with a lot of work done already: https://github.com/GerHobbelt/jison

/cc @GerHobbelt

[yosbelms]

@nolanlawson, if @zaach says nothing about it you still can make a fork in a separated repository.

[qfox]

@yosbelms Since NPM repo owned by @zaach and linked to this repo would be great to get a response from him. Just to have publishing access and merging power to this repo. ;-)

Also guess we should review all existing work and merge it all to the one repo. There is a lot of good work that helps a lot with creating grammars. Yes, we can do it a fork but still better to have it in one place with all the power of community. ;-)

[zaach]

Hi folks– I'm back from the abyss. I see y'all have been busy! I've merged some things and have some more to look over before I push a new version.

@nolanlawson Help appreciated. We also need to update our contributing practices to make collaboration easier, e.g. commit message format, linting, test coverage, and CI. I can file some meta issues around that.

[qfox]

@zaach 👍 Issues would be great. What you think about parsing lexer/grammar files with jison-generated parsers? ;-)

https://github.com/GerHobbelt/lex-parser/ — this one looks pretty good to me ;-)

[zaach]

@zxqfox That's currently how it works :)

The prolific @GerHobbelt made a fork of my lex-parser. I should check out what he has in there.

[qfox]

@zaach By my feelings: it has more stricter syntax (blocker), better error reporting.

👍 for CONTRIBUTING.md, commit message linter, jscs/jshint/eslint stuff, travis, etc.

[nolanlawson]

@zaach Cool, welcome back! :) I'll try to help as much as possible; would be great to have code style, test coverage, linting, etc.

Also if you've gotten a chance to look at Jison debugger, I think it might be a good replacement for the current "try Jison online" page (if you approve). Would need some work to merge it into the existing codebase, though.

[GerHobbelt]

I won't be fit to be the maintainer, I'm sorry to say: jison gets waves of attention from me as it's used as the main grammar compiler in a/the major project of ours; from me there can be silences of a few months easily as priority shifts my attention around a few sub-projects (jison being one of them).

Yes, there's some stuff done in that @GerHobbelt branch/fork/clone (nothing major to the LALR parser engine core; it's more the dressing-up around it, though "one glorious day" I want to see if I can integrate and use PEG parsing in there, for I'm running headlong into the LALR(1) boundaries with my main grammar and GLR is, well, ... not making me happy either, but the PEG bit is currently only 'vaporware/future noise')

On the subject of pullreqqing this stuff I did/do: trouble is I have a few 'Mother Of All Commits' commits in there (yeah, 'twas some of those days ;-) ) which should, ideally, be pulled apart to become Real Nice And Suitable For Our Daughter(tm); I have no idea how comfortable you are with less-then-ideal pullreqs. ;-)

From experience I have learned that apparently I'm more comfortable with 'dirty' and blame it on the power of my toolchain (git + TortoiseGit + Beyond Compare) which I'm very happy with;
it also helps when you've been using Beyond Compare for 15+ years (almost as long as it exists), even before there was any git in my life.

Point being: I can't expect others to go at it the way I do: reading visual diffs has become easy unfortunately. So I guess I'll have to go in an untangle those Mothers... )

Anyway, I'll stop blabbing; more (or less?) coffee for this guy and then we'll see what can be done within the gaps of my todo list.

Cheerio!

[techtonik]

So, this got stalled.. http://nolanlawson.github.io/jison-debugger/ is awesome, but it is good as a separate project that can be linked from the static try page, which is good, because you can easily see how it works without complicated build stuff.

My problem with Try page is that I submitted PR to gh-pages branch - #160 - I thought that gh-pages is manually maintained, but it looks like it is generated. I am on Windows, so I don't have make to rebuild it and whole JS stack is strange. So for me a good start is to improve development workflow and do it in small steps so that @zaach can review things faster.

[GerHobbelt]

🛎️

Can we revisit this, please? I'ld like to keep jison alive, particularly given that it's the best available option out there by far (unless you wish to really kick the bottom out of your toolchain and go the asm.js/llvm way via 'C' output, but that's not for the, ah, faint of heart among us. Nor would it be cross-platform (dev platform that is) like the current jison+node toolchain -- which is an other aspect I like about jison!

Meanwhile my fork has seen some activity y'all might be interested in: particularly in terms of generated parser performance a lot has been done as 'my' jison goes and strips out that it possibly can to make the parsers execute as fast as possible (the lexer hasn't seen this level of activity, but then you can make that one way faster by replacing it with a custom lexer, so folks looking for maximum speed can go that route and get some nice parser/lexer frontend speed.

I know there's still a lot to be done, particularly documentation about the little things that make a jison grammar go voom, but meanwhile I wonder if @zaach is still interested in keeping this alive?

(Off topic bemoaning: github should allow pull requests on wikis! Then I can fork that one, not 'pollute' the main repo and still get some decent docs going and live in the belief that merge-back would still be doable... The current github alternative is to do everything in gh-pages or some such, i.e. in the code repo itself. 🌵 )

[GerHobbelt]

OT: Regarding active development in fork: see also #338. Not exactly happy with this approach (READ: how to see, manage and/or compare forks and useful activity on github) but those goods are available via npm now too: package name: jison-gho.

Would love to close several old issues which are either resolved already or plain outdated; saves me going through work done by others and looks better overall for a tool that we use in production. 😉

[Ravenwater]

+1 *Dr. E. Theodore L. Omtzigt* CEO and Founder Stillwater Supercomputing, Inc. East Coast Office : (617) 314 6424 West Coast Office: (415) 738 7387 Mobile phone # : (916) 296-7901 3941 Park Drive, Suite 20-354 El Dorado Hills, CA 95762

…

On Tue, Jan 31, 2017 at 8:42 AM, Ger Hobbelt ***@***.***> wrote: OT: Regarding active development in fork: see also #338 <#338>. Not exactly happy with this approach (READ: how to see, manage and/or compare forks and useful activity on github) but those goods are available via npm now too: package name: jison-gho <https://www.npmjs.com/package/jison-gho>. Would love to close several old issues which are either resolved already or plain outdated; saves me going through work done by others and looks better overall for a tool that we use in production. 😉 — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#297 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAce0cMr9q2TiKdRg9GX5ioMBX4SOiCGks5rXzpggaJpZM4FsEaB> .

[techtonik]

This branch is 1146 commits ahead of zaach:master - https://github.com/GerHobbelt/jison - this at least needs a reference in README.md