| 1 |
On 3/29/06, Marius Mauch <genone@g.o> wrote: |
| 2 |
> On Thu, 30 Mar 2006 08:30:17 +0900 |
| 3 |
> Jason Stubbs <jstubbs@g.o> wrote: |
| 4 |
> |
| 5 |
> > On Thursday 30 March 2006 01:21, Marius Mauch wrote: |
| 6 |
> > > Marius Mauch schrieb: |
| 7 |
> > > > So after manifest2 is in, I'll revive the other issue that IMO is |
| 8 |
> > > > a requirement for 2.1: enforcing dependencies needed to use the |
| 9 |
> > > > tree (see old threads or glep44 for reasoning). |
| 10 |
> > |
| 11 |
> > Can you summarise the reasoning again please? |
| 12 |
> |
| 13 |
> a) avoid massive breakage when certain new features are introduced |
| 14 |
> (past examples being cascading profiles or new-style virtuals) |
| 15 |
> b) similar to a) allow people to use new features without having to |
| 16 |
> wait for a year or two |
| 17 |
|
| 18 |
It's worth noting that the massive breakage (virtuals fex) can be |
| 19 |
dealt with a bit saner by writing robust code. The virtuals horkage |
| 20 |
during cache transfer being the prime example of that... |
| 21 |
|
| 22 |
> These docs will then tell the user about the following options |
| 23 |
> a) use the secret override |
| 24 |
|
| 25 |
This seems like a case where --force would be a helluva lot better |
| 26 |
then some random nasty env var that is documented on a webpage (assume |
| 27 |
the user has no browser due to bootstrapping the system without a web |
| 28 |
connection). |
| 29 |
|
| 30 |
> That should be an extreme exception though, there will be some rules |
| 31 |
> regarding format bumps (like "only use formats where that haven been |
| 32 |
> supported by stable portage for n months"). |
| 33 |
|
| 34 |
Knowing the rules up front would be useful to evaluate this patch's |
| 35 |
usefulness... |
| 36 |
|
| 37 |
-- |
| 38 |
gentoo-portage-dev@g.o mailing list |
Archives