Boost logo

Boost :

Subject: Re: [boost] Sandbox cleanup
From: Paul A. Bristow (pbristow_at_[hidden])
Date: 2010-11-29 12:44:57


> -----Original Message-----
> From: boost-bounces_at_[hidden] [mailto:boost-bounces_at_[hidden]]
> On Behalf Of Edward Diener
> Sent: Monday, November 29, 2010 3:14 PM
> To: boost_at_[hidden]
> Subject: Re: [boost] Sandbox cleanup
>
> On 11/29/2010 3:27 AM, Vladimir Prus wrote:
> >
> > Folks,
> >
> > it appears that at present, our SVN sandbox is in fairly confused
> > state. A fresh checkout contains objects of very different kinds, in
particular:
> > Comments?
>
> Only that developers should use the correct structure, where the /boost
and
> /libs subdirectories exist under each individual library, and that their
appears to
> be many libraries in the sandbox that are obsolete in the sense that the
library is
> already in Boost but files are still lying around in the sandbox.
>
> How and if the latter can be cleaned up I have no idea. As for the former,
maybe
> Boost should enforce the fact that individual libraries must use their own
/boost
> and /libs subdirectory since the general /boost and /libs directories will
be
> removed at some future date and all the libraries within these directories
will be
> moved to their own individual library.
>
> I too find the sandbox to be pretty messy, but I find it even more
annoying when
> proprosers of possible libraries to Boost put their code in some other
arcane
> place on the Internet rather than in the sandbox.
> So it would be nice if the sandbox could be regularized and cleaned up.

It would be *very* nice if every library followed the correct structure so
it can be moved
easily to trunk or a release download.

I believe the current hassle involved in this is a real impediment to
getting reviewers.

So I think we should begin to make 'tough' noises ;-)

Perhaps one the Boost.Guild people could take it on themselves to start by
contacting offending authors?

And a full description for would-be/existing author of *exactly* what the
structure must be (with examples) would be good too.

Being able to provide a link to a 'how-to' documentation might resolve the
obvious confusion about this.

(Aside: The structure seems quite dotty to me - perhaps someone can provide
a rationale?

But even if that rationale is lost in the mists of time, we have a standard,
and a standard is what we must all follow, however peculiar.)

Paul

---
Paul A. Bristow,
Prizet Farmhouse, Kendal LA8 8AB  UK
+44 1539 561830  07714330204
pbristow_at_[hidden]

Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk