dmd 1.041 and 2.026 releases

Walter Bright newshound1 at digitalmars.com
Sat Mar 7 10:53:14 PST 2009


Daniel Keep wrote:
> 
> Walter Bright wrote:
>> Leandro Lucarella wrote:
>>> Walter Bright, el  5 de marzo a las 11:37 me escribiste:
>>>> Tomas Lindquist Olsen wrote:
>>>>> Compiling on linux from source is broken! Looks like you forgot to
>>>>> include the total.h file!
>>>> You don't need it, I'll fix the makefile. total.h is for precompiled
>>>> headers, which don't even exist on gcc.
>>> I don't know exactly what you mean, but:
>>> http://gcc.gnu.org/onlinedocs/gcc/Precompiled-Headers.html
>>>
>> I suppose it's been added since the last time I looked. gcc has
>> thousands of switches.
> 
> There are days when I can get upset at you for not implementing feature
> X.  I really need X, damnit.  Do you have any idea how badly I want X?!
> 
> ...
> 
> And then I do something that requires me to look at the man page for
> gcc.  And then I die a little inside.

Every command line switch that a compiler has is a failure of design. It 
reminds me of early jet engine controls for aircraft. There were all 
kinds of tweaks for it, and the pilot often got it wrong. The thing was 
eventually boiled down to a fantastically complicated mechanical box, 
that had one lever on it - "gimme more power".


> It's strange at times, but I do appreciate that you take a considerable
> amount of convincing.  It's a bit like the situation with Archchancellor
> Mustrum Ridcully from Discworld; he holds the view that if someone is
> still trying to explain something to him after about 2 minutes, it must
> be worth listening to, and if they give up earlier, it was not worth
> bothering him with in the first place.

The C++ standards committee has a good system for that. They won't 
consider any ideas unless they are submitted as papers. The idea is that 
if someone doesn't care enough about a feature to write a paper about 
it, it isn't worth the committee's time.

I'm not willing to go that far, but working on D is a full time job for 
me. New ideas and proposals come in *every single day*. I cannot even 
*read* them all, let alone prepare an intelligent reply even in saying 
it's not a good idea.


More information about the Digitalmars-d-announce mailing list