libstdc++-v3

libstdc++-v3

[Ryszard Kabatek]

Is it possible to say, when the libstdc++-v3 will be available?
What about the C++ keyword `export'?

Ryszard Kabatek
--
Martin-Luther University Halle-Wittenberg
Department of Physical Chemistry
Geusaer Str. 88, 06217 Merseburg, Germany
Tel. +49 3461 46 2487 Fax. +49 3461 46 2129

libstdc++-v3

[Gabriel Dos Reis]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 469 bytes --]

>>>>> «Ryszard», Ryszard Kabatek <rysio@rumcajs.chemie.uni-halle.de> wrote:

Ryszard> Is it possible to say, when the libstdc++-v3 will be available?

Work is in progress. Patience. A snapshot next week, Nathan ?

Ryszard> What about the C++ keyword `export'?

I don't know the status about this. But I think its support is more
important than `restrict' in C++ (assuming one come up with a
reasonnable semantic and has solved related problems iin C9x).

-- gaby

Re: libstdc++-v3 (and "export")

[Nathan Myers]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1075 bytes --]

Gabriel Dos Reis wrote:
> 
> >>>>> «Ryszard», Ryszard Kabatek <rysio@rumcajs.chemie.uni-halle.de> wrote:
> 
> Ryszard> Is it possible to say, when the libstdc++-v3 will be available?
> 
> Work is in progress. Patience. A snapshot next week, Nathan ?

We're hoping to do one next week.

> Ryszard> What about the C++ keyword `export'?
> 
> I don't know the status about this. But I think its support is more
> important than `restrict' in C++ (assuming one come up with a
> reasonnable semantic and has solved related problems iin C9x).

There is no current work on export that I know of.

It is by far the most important piece left; it implements in a standard,
portable way the useful qualities of what people have been calling 
"precompiled headers".  Without it we'll be looking at a noticeable 
fraction of a minute per .cc file just to reread all the standard 
library template definitions.

Anybody who wanted the world's everlasting gratitude probably wouldn't 
choose export to work on, but if they wanted *mine* that would do it.

Nathan Myers
ncm@cantrip.org

Re: libstdc++-v3 (and "export")

[mal]

Nathan Myers writes:
 > There is no current work on export that I know of.
 > 
 > It is by far the most important piece left; it implements in a standard,
 > portable way the useful qualities of what people have been calling 
 > "precompiled headers".  Without it we'll be looking at a noticeable 
 > fraction of a minute per .cc file just to reread all the standard 
 > library template definitions.
 > 
 > Anybody who wanted the world's everlasting gratitude probably wouldn't 
 > choose export to work on, but if they wanted *mine* that would do it.
 > 
 > Nathan Myers
 > ncm@cantrip.org

I've thought about export but I'm not sure how a good design for its
implementation would look like. It would definitely need a database
and that opens a whole can of worms with respect to version control,
concurrent access, invalidating cached versions when the source
changes etc. 

We use SunPro at work with one template DB for the project and we have 
to nuke the DB very regularly and recompile the world. Alternatively
you can have one DB per developer but then making sure everything is
in sync when some template definitions change is very tricky.

The export feature is necessary but I haven't seen a good
implementation/design of it yet.

Lieven