httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Tim Costello <>
Subject RE: [PROPOSED] linkage spec names, final binding vote
Date Thu, 01 Jan 1970 00:00:00 GMT
On Tuesday, 30 May 2000, William A. Rowe, Jr. wrote:
> Please go back to the archives (March, I believe) when I 
> sent the assembly source of some C code examples to illustrate 
> the bottom line issues.

My mistake - I never read that message completely. I guess what I was saying
was that I don't think the performance gain from using __declspec(dllexport)
is worth it. The issue has already been decided though. 

> 1) `use DEF files` and `easiest to maintain` is an oxymoron :-)

That's ok - I respect your opinion (but we still disagree). :)

I think it's better to just have code that is code (which is compiled), and
another file that says what can be accessed from outside (which is used by 
the linker). 

> 2) `simplest method that works` is, as you point out for the
>    case of data, simply incorrect. 

Sorry, I was talking about the export directives only, not importing. 

>    If we would actually define x for y consistently, then we 
>    would use y throughout instead and have no EXPORT(type) 
>    macros whatsoever.

hooray! :-)

> 3) Reconciling those statements, DEF files go away, and
>    the code is maintained in one manner only... either the
>    symbol is exported in it`s declaration, or it is absolutly
>    private.  I`m a big advocate of that approach, but that`s
>    a debate for July, not today, so don`t -even- get me 
>    started ;-)

Ok - sorry to have brought it up again. 


This message was sent through MyMail

View raw message