Re: [Tinycc-devel] the most useless question on this list about basic c++ extensions

[mobi phil]

Hello Pancake, hello All,


If I am not wrong you are the guy behind radare. Really cool tool,
however bit disappointed that never ever got any reply to mails sent
to the list...


> C++ is a very different weird and hardly complex language compared to C.
Indeed it is complex, for sure handling mangled symbols with radare
would cause pain to some radare users :). But I think this does not
meen it is necessarily weird. I can call C as weird, as 'natural'
inheritance or other "object orinted extensios" were simply not added
to the original design. (I know the workarounds, no worry ). I also
could say that C is weired, as in lot of situations I have to accept
that it does not provide features that seem to be so obvious to have,
after using so other languages. Instead I have to write workarounds
that simply cost me time. I miss templates (simple ones, not the
crashes-my-compiler-heavy-metaprogramming). There is workaround for
templates, but you need more, and often you end with uggly macros etc.

> Are you saying that TCC will support it? In mainstream? As an extension?
> I would really prefer to not have such support..and in case, just think on
> object-oriented extensions, but not c++.
So... it seems that you need object oriented extensios... I think you
are/were missing them when writing radare 1 and refactoring for radar
2, isnt't it? So add this, add that and consider a bit forces that
push for minimal compatibility... you end up with c++. Anyway there
are few other things that could be added even on top of c++. And
indeed, the key is extension.. I think tcc is a very interesting
platform to play with ideas like prototypes, mixins, traits, etc. etc.
If any of them you did not need in any of your works, this does not
mean, that they are usefull etc. etc. As the design of tcc is
relatively simple, maybe the idea is to add some plugin mechanism to
the parser, where one could add hooks for some extensions etc. Why
not, one will end up with C.2 a super language that will offer more
than c++ etc, with superfast compiler etc. etc.. Prototype languages
are booming, prototypes are powerful feature. I saw recently Lisaac, a
new language. I like some ideas of lisaac,
(http://www.mobiphil.com/2010/05/lisaac-is-almost-complete, but the
cons list is much longer than the pros...

That's all for the moment:)



rgrds,
mobi phil

being mobile, but including technology
http://mobiphil.com