lua-users home
lua-l archive

[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index] [Thread Index]


Now that I've really stepped back and analyzed why I'm doing this, I'm
realizing how ridiculous my wishlist really is. :-) Pretty much the
only thing I really think has a chance of getting in (besides the
"Really useful generic stuff" section) if I want to keep libmc light
and simple is the option parser. The algebraic/geometric stuff you are
right about - it's not something most people need, and would be better
in a separate library. And as for unit conversions, it wouldn't be
that it's so hard to implement, but the conversion factors would be
just really bulky.

The XML generation involves something along the lines of xml.p {class
= 'greeting', "Hello, world!"} - really, just a quick & dirty way of
generating XML. Though I used Genshi as an example, a better idea of
what I was thinking is at
http://dev.pocoo.org/projects/werkzeug/browser/werkzeug/utils.py, in
the HTMLBuilder class. It implements that in just a couple of KB,
whereas Genshi is a big library that manipulates the DOM and stuff.

Anyway, remember - it *is* a wishlist. :-)

Regards, LeafStorm
"There are 10 types of people in the world - those who understand
binary and those who don't."



On Sun, Aug 30, 2009 at 8:55 AM, steve donovan<steve.j.donovan@gmail.com> wrote:
> On Sun, Aug 30, 2009 at 1:53 PM, Leaf Storm<leafstormrush@gmail.com> wrote:
>> and Penlight. Some of the ideas on my wishlist page notwithstanding,
>> the libmc idea is to fill in the stuff Lua doesn't provide, and
>> Penlight's is to do that and add stuff - like the lexer, the
>> 'pl.input' module, the file tools, etc.
>
> I've been looking at the wishlist and wondering about that as well ;)
>
> http://bitbucket.org/leafstorm/libmc/wiki/Wishlist
>
> And a fine set of wishes they are - but once you have implemented
> them, you are going to get that 35K up to 190K which is the size of
> the Penlight source.  I like small & focussed myself, hate huge big
> frameworks, so I understand the motivation.  The way out of the
> (apparent) difficulty is a philosophy which C++ people call 'Don't
> have to pay for what you don't use', i.e. don't bring in the kitchen
> sink if the user just wants a split() function.  PL uses lazy loading,
> that is, require 'pl' does not bring in all that 190K, but as soon as
> say tablex is referenced, that module will be loaded.
>
> I see you mention 'algebraic and geometric' formulae, definitely
> useful, but not to everyone. That probably ought to be a standalone
> module of its own. It would e.g. be cool to have the basic statistics
> functions like std dev, etc available.
>
> XML generation - I don't know the Python library, but it's a pity that
> there is no DOM implementation AFAIK that could be used with luaexpat,
> or no XML prettyprinter that could work on the same LOM format.  In
> that case, I would probably try to motivate it as an add-on to
> LuaExpat (i.e. something like lxp.dom and lxp.writer).
>
> Unit conversions: documentation is a little patchy (par for the
> course) but this seems to work very well:
>
> http://luaforge.net/projects/luadc
>
> steve d.
>