| 1 |
On Thu, 20 Aug 2009 13:36:30 -0400 |
| 2 |
Andrew D Kirch <trelane@×××××××.net> wrote: |
| 3 |
> > > I look forward to seeing Funtoo's creation of EAPI funtoo-2. |
| 4 |
> |
| 5 |
> Obviously you don't get it. We aren't going to spend time writing |
| 6 |
> this sort of spurious and unnecessary specification documents. |
| 7 |
|
| 8 |
Why not? Are you saying you don't want the way Funtoo's package manager |
| 9 |
works to be documented? I'll remind you that the feature in question |
| 10 |
isn't documented anywhere. |
| 11 |
|
| 12 |
It's a five minute job, and it will mean Funtoo's tree can be used by |
| 13 |
any of the many third party tools that work with repositories. |
| 14 |
|
| 15 |
I find it hard to believe that one of Funtoo's goals is to use |
| 16 |
undocumented features and not have any way for contributors to know what |
| 17 |
the package format is. |
| 18 |
|
| 19 |
> Nor do you, as I have proven, have standing to file such a bug as you |
| 20 |
> are not on the council (even as an alternate), and the SOLE option |
| 21 |
> for packages violating PMS per the council is a council vote to mask |
| 22 |
> the package. |
| 23 |
|
| 24 |
That's not the sole option at all. The first thing to do is to get the |
| 25 |
issue fixed. Masking is the nuclear option. |
| 26 |
|
| 27 |
> Portage is a tool used by users, repoman is a tool used by developers |
| 28 |
> for tree QA. |
| 29 |
|
| 30 |
repoman isn't an effective way of getting messages to overlay |
| 31 |
developers. |
| 32 |
|
| 33 |
> Considering the zeal with which you are pushing this |
| 34 |
> "accidentally works with Portage until Portage is fixed", I believe a |
| 35 |
> reasonable person is going to look at the b.g.o bug, and at the |
| 36 |
> Paludis bug and realize that you're more interested in process than |
| 37 |
> innovation, and that you simply don't care about throwing needless |
| 38 |
> confusing warnings at users (indeed a prima facia examination of |
| 39 |
> Paludis would seem to confirm this, and my concerns WRT Paludis and |
| 40 |
> the development methods are well known). I think they'll also |
| 41 |
> realize that throughout this process you've been less than honest, |
| 42 |
> and a huge impairment to the work going on at Funtoo. |
| 43 |
|
| 44 |
The process for innovation is EAPIs. This allows innovation to go |
| 45 |
through in a well defined manner that doesn't cause highly random |
| 46 |
breakage every time a new feature comes along. Unfortunately, your |
| 47 |
early misconception that EAPIs are about Paludis is getting in the way |
| 48 |
of you taking a few simple steps to turn this from exploiting |
| 49 |
undocumented behaviour to a well defined feature. |
| 50 |
|
| 51 |
EAPIs are there to help you, not hold you back. |
| 52 |
|
| 53 |
-- |
| 54 |
Ciaran McCreesh |
Archives