[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index]
[Thread Index]
- Subject: Re: Lua Ecosystem
- From: steve donovan <steve.j.donovan@...>
- Date: Sun, 24 Nov 2013 12:37:12 +0200
On Sun, Nov 24, 2013 at 11:48 AM, Pierre Chapuis <catwell@archlinux.us> wrote:
> Which is why there is also a point about discoverability later on. And it
> is no coincidence that Lua Toolbox is inspired by Ruby Toolbox.
Thanks, Pierre. I have now read the slides and feel able to actually
contribute to the discussion (RTFA as they say on Slashdot). And Lua
Toolbox looks like a _fantastic_ idea which we as a community should
get behind.
I would really have loved to be present at your talk but my work year
did not permit it. A summary of the resulting discussion would be
very useful to those of us who could not attend.
> Yes... Also this results in conflicts if you use modules that depend on
> different implementations.
You advise that would-be module writers should do some research and
see if there's an on-going project that already satisfies their itch,
and put resources in it. Sometimes maintainers are genuinely MIA, in
which case after making a reasonable effort to contact a person can
keep a living version going and publish it on LuaRocks. On the package
maintainer side, maintainers should make the corresponding effort to
encourage participation.
Your point about API compatibility is important. A number of
implementations can coexist if they're mostly plug-in replacements for
each other, with different strengths (e.g. a speed demon LuaJIT
version, a C version, and a pure Lua version for those who cannot use
binary extensions)
> If the module that decodes barcodes is not perfect, that's not great, but
> better than if it doesn't exist.
Yes, especially if it's sufficiently clear and documented so that it's
fixable, and the maintainer accepts patches, and we can find it easily
enough. So I hope that Lua Toolbox helps with the community-rating
and discoverability aspects.
steve d.