[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index]
[Thread Index]
- Subject: Re: Lua 5.0 and globals (long)
- From: "Peter Hill" <corwin@...>
- Date: Mon, 3 Mar 2003 04:17:39 +0800
Peter Hill:
> Sigh. Now THAT is going to be a long and torturous trip :-(.
> A search for "globals" found 574 pages. A summary of the final
> conclusions would be nice.
Peter Shook
> Please, go back and read the list. You don't have to read all the
> questions, just read the responses by the three authors and some
> of the other old timers.
> Just before the time of release 4.0 is a good place to start.
Well, so far I'm up to message 426 (out of 577) in the search of "Globals"
(about July 2002 & Lua5-alpha) and I still haven't got to the good part
yet... but onwards I trudge :-)
I guess what I would like to see in these cases would be, at the end of a
significant discussion (for example, one which actually results in a feature
change), that someone involved might factor out the salient points and list
the results on the Wiki... so that late-comers could quickly get up to speed
on the issues.
Not that I'm volunteering of course! Luckily for me I haven't been involved
in any notable outcomes so that I'm in the clear so far :-)... though I
suppose I should list a few of my threads.
> It's good to have a fresh pair of eyes look things over and question
> things, but please don't suggest changing things without considering
> the history behind it. Otherwise we may end up going in circles.
I'm hoping that such Wiki updates might avoid the 'circles'.
In this case, though, I didn't suggest change, but merely asked for a
summary of the design decisions that lead to the new global concept.
> Some people wanted variables to be local by default. Probably due to
> over exposure to PHP and TCL. Others thought it would be good for Lua
> to have better facilities for programming in the large. It was thought
> that declaring globals might solve both of these issues.
>
> Lua almost had a "global" statement that sort of worked like a Pascal
> "with" statement, but one little asterix spooked a lot of people,
> and it was canned because it was too complicated.
That's about where I am up to so far...
> Now we have a compromise that should have made the PHP and TCL people
> happy but we haven't heard them say anything about it.
> Unfortunately this has made another camp unhappy, but I think
> this is mostly due to coroutines being new and experimental.
Huh? The coroutines seem delightfully simple (given the reasonable limit
that they run in-Lua). People are unhappy?
> I'm sure our three musketeers will save the day in a future episode.
> You just have to keep watching.
All-for-one...
Peter Hill.