[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: Parser Maker
From:       "Thomas Schilling" <snuffeler () gmx ! net>
Date:       2001-10-24 19:51:19
[Download RAW message or body]

> IMHO wether or not you do this parser maker, depends on your reason for
proposing
> it in the first place.
> [...]
> On the other hand, if you are doing this because the subject interests
you, you
> may want to do it is if you could find lots of similar things.

The idea of doing this came to me as I tried to implement (or experiment
with) the, IMO, very important code completion for KDevelop (I read that
there were experts missing until KDevelop 2.0 so I'd like to become one)
which base would also allow code hinting and better code browsing. So this
idea of a parser maker came up and I asked (indirectly) if there's already
something existing in this direction.
But as I was told that there exists something that is good enough for my
purposes, why reinvent the wheel? That's the big sickness of software
development - that's why I also advocate using 'brutal' approaches to make
software componental.
-- We currently have at least 10 text editors running on KDE, why can't we
have ONE that is a (configurable) component or, better, consisting itself of
many smaller components (renderer, grep, find, tool-manager, ...). But
that's another topic and will maybe get subject for KDE 4 or so. --

> Whatever the reason for you doing something, don't allow yourself to be
overly
> influenced by public opinion. To many good things are sacrificed on the
alter of
> public opinion.

As I said, I think it's better using existing work. The problem is to find
it/them.

Well, there are two main reasons why this 'rule' could be broken:
 1) a different approach was choosen.
 2) you cannot find fast enough into the code - there missing a lot of tools
making code more understandable (good UML programms - with round
tripping!! - code hinting - that's what I wanna work on)

> If it's something you want to do then just do it. I'm sure that your
project will
> find some people who find it useful and if not... why should you care.
Just let
> your imagination and creativity run free. Your spirit will thank you.

I think the point I wanna work on will interest a LOT of people (hopefully
all KD programmers :-)

Thomas

 
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<

[prev in list] [next in list] [prev in thread] [next in thread]