Thanks Glyph. That's exactly what I was waiting for. I'd like to add one other important thing for keeping pyjs going: 3 Some people willing to answer (dumb or not so dumb) questions. Preferably a native speaker, but not necessarily.
I've been extremely busy, as I know others will have been too (not just at work but also in my spare-time). Didn't use pyjs for a while (personally, though we're using it in our company and depend on it). There's a different translator available, which is in principle way more pythonic in accessing attributes. The default translator is 'proto'. It depends on the javascript way of creating classes (using 'prototype'), which binds attributes on defining the class. The other translator is 'dict' (--use-translator=dict), which resolves the attributes runtime. (Note, I just committed a fix to get this translator going again). This is in most cases somewhat slower. It lacks the support of generators (i.e. yield statement) and is a total re-implementation. If this is something people are waiting for, I'll pick this up again and implement yield (will probably take 16 to 24 hours of work). So, maybe you (Glyph) and Lex (and others if they need this) can have a look if it is worthwhile or not. BTW dropping support for old-style classes is impossible. We don't support them currently. The syntax is supported, but they are handled as subclasses of object (new-style classes). - Kees On Tue, Oct 8, 2013 at 10:52 PM, Glyph <gl...@twistedmatrix.com> wrote: > > On Oct 7, 2013, at 1:02 PM, Lex Berezhny <l...@berezhny.com> wrote: > > Here is my two cent recommendation/suggestion to save the project: > > > Here are my recommendations to save the project: > > > 1. Fix bugs. > 2. Add features. > > > :-) > > I think the list of things you've proposed are all possibly good ideas > (particularly, I think "100% compatibility with Python" is really > important). However, changing project structure around and setting project > policy are not going to increase the amount of work done if nobody is > available to do the work or interested in doing it. If there is any lesson > from the recent "hijacking" (as the blog post so dramatically put it) it is > that just adopting new development tools or practices will not necessarily > fix any problems. > > The important thing is to just keep actually writing code, submitting pull > requests, commenting on bugs and generally moving things forward one little > piece at a time. Form should follow function: if the project structure > impedes that happening, then it should be changed, but otherwise it is > probably fine. > > Narrowing the focus is only useful if there are lots of people doing work > but the effort is dissipating because it's not being concentrated in a > useful / exciting area, and they're getting demotivated. Setting goals is > only useful if people are going to do some work but don't know what to work > on. > > It's tempting to try to sketch some Big Ideas (and it can provide some > motivation, so I wouldn't say not to do it at all!) but I would definitely > caution against trying to debate these back and forth endlessly rather than > just submitting one or two small pull requests to try to move things > forward. > > Things like splitting out the widget library into a separate project from > the language translator might be more useful later, especially if PRs that > fix language things end up being blocked on widget-library things. Doing > that by itself, though, definitely won't fix anything by itself, and I can > prove it, because someone already did that and it didn't work: < > https://pypi.python.org/pypi/pyjsmin/>. You can see it's been twice as > long since someone has committed to that: < > https://github.com/nathants/pyjsmin>. Splitting out the code in such a > way that both things still work is itself a bunch of effort, and only worth > doing if it is really holding things back. > > So, initially at least, rather than a general, negative focus (let's rip > out the widget library! let's simplify the project!) I think these > discussions should center around a specific, positive focus (let's make the > language runtime more python-compliant by fixing < > https://github.com/pyjs/pyjs/issues/799> and < > https://github.com/pyjs/pyjs/issues/787>! let's make it more accessible > by fixing <https://github.com/pyjs/pyjs/issues/742>!). > > If those just happen to be the bugs I've filed, it's because they're > totally the most important bugs that should be fixed immediately ;-). > > -glyph > > -- > > --- > You received this message because you are subscribed to the Google Groups > "Pyjs.org Users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to pyjs-users+unsubscr...@googlegroups.com. > For more options, visit https://groups.google.com/groups/opt_out. > -- --- You received this message because you are subscribed to the Google Groups "Pyjs.org Users" group. To unsubscribe from this group and stop receiving emails from it, send an email to pyjs-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/groups/opt_out.
