|[14 earlier articles]|
|Re: Is this a new idea? firstname.lastname@example.org (1992-11-12)|
|Re: Is this a new idea? email@example.com (1992-11-11)|
|Re: Is this a new idea? firstname.lastname@example.org (1992-11-16)|
|Re: Is this a new idea? email@example.com (1992-11-16)|
|Re: Is this a new idea? firstname.lastname@example.org (1992-11-17)|
|Re: Is this a new idea? email@example.com (1992-11-18)|
|Re: Is this a new idea? firstname.lastname@example.org (1992-11-20)|
|From:||email@example.com (Stavros Macrakis)|
|Organization:||OSF Research Institute|
|Date:||Fri, 20 Nov 1992 16:13:26 GMT|
firstname.lastname@example.org (Clyde Smith-Stubbs) said:
> ...you can't parse code that contains references to such things as
> typedefs that occur earlier in the code, if the typedef is missing
> or in the middle of being edited....
I (email@example.com (Stavros Macrakis)) replied:
> This is a C-specific problem: a bug in the design of C's syntax. Most
> other languages don't commit such foolishness, except of course those
> with extensible syntax.
firstname.lastname@example.org (Clyde Smith-Stubbs) continued:
...the solution is now clear. We should all immediately abandon C
Sounds like a good idea.
and any languages with extensible syntax,
Not many of those around these days.
and perform an appropriate penance for having used such
ideologically unsound tools. A month of COBOL programming sounds
I don't suggest penance. I suggest improving the choice of language.
And of course by switching to a non-extensible language we will
immediately get the benefit of instant compilation. I'm talking about
BASIC of course.
Recompilation is actually quite fast if you do it on appropriately-sized
chunks, e.g. individual functions or modules. Even for C, with its weak
notions of inter-module declarations, it is possible to precompile
definition headers (the processing of which is a major cost in compiling
C, especially if you use large libraries like X).
More seriously, the typedef problem in C does cause parsing difficulties,
but C is here to stay. Rather than pointing fingers at its design, it
would be better to look for ways of solving the problem at hand (which was
speeding up re-compilations).
Working on better tools for existing languages doesn't preclude working on
better tools _and_ languages.
Return to the
Search the comp.compilers archives again.