[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index]
[Thread Index]
- Subject: Re: Lua 5.2 and continue
- From: Alexander Gladysh <agladysh@...>
- Date: Mon, 1 Feb 2010 10:56:16 +0300
On Mon, Feb 1, 2010 at 10:47, pan shizhu <pan.shizhu@gmail.com> wrote:
>> If we start to add defines on language feature basis, Lua would cease
>> to exist as one language, and would become a blurred set of dialects.
> Each games have differnent libraries and all Luas are different.
Again — it is so only if you want it. This is not true for the code I
usually work with.
> But in principle they have "everything is a table" philosophy and I think
> they are the same language.
"Everything is a table" is not everything that Lua is (but a good part
of it, I agree).
If someone patched the core language (as apposed to the libraries, to
the sane extent) — I believe, the result would be not Lua, but a
custom dialect of Lua.
This is terminology problem.
>> I believe this would really hurt the language. Even now it is still
>> barely possible to find good 3rd party Lua library (in broad sense)
>> for a given task. It would be impossible with this change.
> IMO it is the 3rd party libraries who hurt the language.
You mean the lack of them?
> Look at python, that most libraries are accepted by official
> distribution, so we always have the standard implementation of many
> features. If we stripped off everything, and make most python
> libraries as 3rd-party library, python would not become so sucessful.
Lua is successful enough for me. Part of its success is lack of
"pythonic" bloat.
We surely need a set of "blessed" libraries for all purposes. But,
IMO, they should not get into the core language.
>> You need fancy feature, not in core language and you can't talk Lua
>> Team into implementing it?
> It is different from "need fancy feature". The implementation of
> "continue" is already there, but it is not in the official release.
A lot of other features also exist in this form.
> I think it would be better to include it in official release, even if
> we must define compile-time argument to enable the "feature", a
> disabled-by-default feature is much better than a 3rd-party-patch.
I understand you, but I disagree. See above.
Alexander.