Luis de Bethencourt wrote:
> On Tue, Nov 18, 2008 at 11:07 PM, Shawn Walker <[EMAIL PROTECTED]> wrote:
>> Luis de Bethencourt wrote:
>>> That said, we shouldn't accept binary built in an untrust worthy
>>> machine. The process we define has to make submissions be built in our
>>> controled systems. How Launchpad works.
>> Being aware that a "trustworthy machine" is highly dependent upon the
>> machine, the person using it, and so forth.
>>
>> Again, don't forget the exception cases.
>>
>> I don't think anyone here is suggesting that only certain people can
>> build software.
>>
>> If we are, that's sort of silly, since we can't distribute the workload
>> if we do that.

> Can you define distribute the workload?

I'm not talking about the CPU compilation, etc.  I'm talking about 
"people resources" needed to re-do work that's already been done.  Hence 
my reference to certain people doing the build.

For example, if someone contributing a package has been trusted with 
access to a build system, they should be able to build it on that system 
and publish it with approval instead of someone else having to do so.

> As far as I know all Linux distros have a build machine in the
> official repo (the one that gets mirrored) and nobody complains. I
> wouldn't trust a deb package built by someone I don't know and that I
> can't check the sources. The exception to this is gentoo, which makes
> the users be their own build systems in most of the cases.

But you are trusting people you don't know and hence why the Debian 
OpenSSL debacle happened.

-- 
Shawn Walker
_______________________________________________
indiana-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/indiana-discuss

Reply via email to