This is the mail archive of the
mailing list for the GCC project.
Re: gcc compile-time performance
- From: Hans Aberg <haberg at matematik dot su dot se>
- To: dewar at gnat dot com (Robert Dewar)
- Cc: pkoning at equallogic dot com, gcc at gcc dot gnu dot org, help-bison at gnu dot org
- Date: Tue, 21 May 2002 00:44:14 +0200
- Subject: Re: gcc compile-time performance
At 17:50 -0400 2002/05/20, Robert Dewar wrote:
><That makes no sense to me at all. If I could only fix one typo at a
>time, progress would be quite a bit slower.
>Well if you have an instantaneous compiler and an integrated environment
>which puts you at the error, immediately ready to correct, it's a
>reasonable approach. It was the way UCSD Pascal worked, and many people
>found it quite convenient.
If one should carry out the suggestion as fundamentalisticly as the quote
above, then one should get down to a compile time of about no more than a
few seconds; with anything more than that, it becomes annoying.
There is a system for TeX (BlueSky for MacOS) that compiles the code
continuously, whenever one types and reports the errors. For such a system
to very practical, one needs a very fast computer, plus preferably a
language/compiler combination especially designed for that.