Re: [gentoo-dev] Unmasking modular X

[Donnie Berkholz]

Mark Loeser wrote:
>>> On Mon, 23 Jan 2006 23:06:12 -0800 Donnie Berkholz
>>> <spyderous@g.o> wrote:
>>> What's wrong with the original idea of just making any unported ebuild
>>> pull in all of modular X (minus drivers)? Yes, it means that some
>>> people will pick up unnecessary deps until all packages are ported, but
>>> it avoids anyone having to see flashy red errors.
>> The problem with that is that it removes all motivation to ever port the
>> packages. They'll just stay that way forever, where forever means "until
>> I threaten to remove that from the virtual," in which case we'll be in
>> the same scenario we are now. Why? Because people have better things to
>> do than fix stuff that isn't broken.
>
> It'd be nice if you reconsidered this as it will minimize any breakage that
> may occur.  Knowing that >800 packages are broken, and going to unmask it
> knowing that just doesn't seem acceptable in my eyes.  ~arch isn't meant to
> be "things are known to be broken."  It's meant to mean, we think all of this
> is ready to be stable, which it certainly won't be in this case.

No, it won't. It will just postpone the same breakage, as I said above.
You haven't provided any logic or backup to your contrary statement,
just said that somehow a large portion of the other 800 will magically
get ported.

Let me break this down again: of that 800, about 250 are unmaintained
packages according to metadata.xml or lack thereof. About 200 are games.
About 150 more belong to largely inactive herds. That's roughly 600 that
we already know will not get ported in a timely fashion, if left to
their maintainers, all because of lack of manpower. What do you propose
to deal with them? All I've heard besides mine is proposals that delay
the same breakage, not actually get anything fixed.

Thanks,
Donnie