Re: [Boost-build] intel-win issues

From: Jürgen Hunold (hunold_at_[hidden])
Date: 2006-04-05 05:22:05

Hi Stefan !

Am Middeweken 05 April 2006, 09:57 schrieb Stefan Slapeta:
> Jürgen Hunold wrote:
> > I'd suggest to keep this options for intel < 9.0 (or 8.0) until someone
> > with more knowledge comes with a better solution.
>
> First of all, this is _not_ dependent on the version of Intel Compiler,
> but on the version of (emulated) VC! If makes no sense at all to provide
> options for Intel 9/8.0 that are not provided for lower versions.
>
> So far, I and several others have run boost tests for Intel now for
> several years and analyzed/submitted many DRs making this compiler free
> of any errors for the last few boost releases. This was based on a
> configuration which always has been considered by anybody of us to be
> the best for this compiler. 

Well, you obviously use intel regularly and have much experience. On the other 
I've never gotten around to test it in depth... :-(( but was the only one 
using it with V2 at that time...

> I don't know which positive effect at all 
> could be expected by introducing undocumented and unsupported compiler
> options that are known to be considerably dangerous. I'm glad that
> Vladimir removed them again.

Well, I used them to get intel-6.0 compile some sophisticated code it wouldn't 
compile in vc-6.0 compatibility mode. You probably would call this an evil 
hack.

I neither have vc-6.0 nor intel-6.0 anymore so I really don't care about 
support for this combination :-))

So, please go ahead and get the current compilers working.
Maybe I'll find time to test the current intel compiler someday ;-))

Yours,

Jürgen

-- 
* Dipl.-Math. Jürgen Hunold  ! Ingenieurgesellschaft für 
* voice: ++49 511 262926 57  ! Verkehrs- und Eisenbahnwesen mbH  
* fax  : ++49 511 262926 99  ! Lister Straße 15
* hunold_at_[hidden]          ! www.ive-mbh.de