RE: [Axiom-developer] Fwd: [M#73697383] Re: Disk-quota Request

[Bill Page]

On September 14, 2006 5:02 PM Gabriel Dos Reis wrote:
> 
> "Alfredo Portes" <address@hidden> writes:
> 
> | This is the response from Google to my request for more space:
> ... 
> | Your project certainly looks mature and has a healthy community
> | around it, and we'd like to see it hosted at Google Code.

That's the good news. :-)

> | However, it also looks like there's a lot of extra 'junk' in the
> | repository... for example, the zips/ directory contains dozens of
> | releases(?) of the same package?  (And even random odd Arch packages,
> | like 'tla'?)

They've been reading my posts to axiom-developer, haven't they? ;)
I think they have a point.

> | 
> | We'd like people to *not* use Subversion as a package-distribution
> | system, but as a source distribution system.  (We're working on
> | creating a dedicated 'downloads' feature right now, though!)  At the
> | moment, we'd be more comfortable giving you another 100MB of disk
> | quota... perhaps you could clean up your repository's history? Trim
> | it down to just necessary source code, and move the release .zips
> | somewhere else for a while?
> 
> Tell them, we're actively working on reducing zips (the can have
> a look at the build-improvements branch).  Furthermore, what it
> contains is really necessary for building Axiom for many, common,
> target configuration.  And, we cannot move the zips somewhere else.

Throw away the tla tarball, and I agree that the build-improvements
branch is probably the minimum source code distribution.

> 
> 1G is good for the moment.
>

They said +100MB that makes 200MB total, right?

If our primary motivation right now is to diagnose/solve the svn
checkout problem, then maybe just uploading the build-improvements
branch to Google Code would be sufficient for now.
 
> Perhaps they think we are not serious?
> 
> They certainly will not object to the current 12G of GCC -- Google
> invests lot of resources into GCC.
> 

I doubt that Google would host GCC using this mechanism so the
comparisonn is rather moot. The problem is, I suppose that Axiom
is not big enough to warrant direct Google investment but (almost?)
too big for the Google Code project.

Regards,
Bill Page.