| 1 |
Ciaran, |
| 2 |
|
| 3 |
Admittedly, I'm new to this PMS thing so in many cases I'm speaking |
| 4 |
from a position of ignorance, but I guess I need to jump in |
| 5 |
somewhere.... |
| 6 |
|
| 7 |
I think that standardization is a good thing and interoperability |
| 8 |
between paludis, portage, pkgcore and others is something we should |
| 9 |
strive for. If at all possible, I think that this standardization |
| 10 |
effort should be "multi-lateral" in the sense that Gentoo and pkgcore |
| 11 |
are also active participants in the standardization process. |
| 12 |
|
| 13 |
Also, I don't think that the council itself needs to be involved |
| 14 |
directly, as a standards/spec project can be created and worked on, |
| 15 |
and the conformance of Portage to this standard can be measured, and |
| 16 |
if desired Portage developers can tweak portage so that it is more |
| 17 |
conformant to this standard. This can be done voluntarily by all |
| 18 |
parties, as they deem it useful. |
| 19 |
|
| 20 |
The goal would be to have the ability to measure a package manager's |
| 21 |
behavior against a known standard, rather than force a certain package |
| 22 |
manager to behave a certain way. I would expect that the general |
| 23 |
concern for interoperability within the Gentoo community will |
| 24 |
encourage package manager developers to work towards resolving any |
| 25 |
interoperability problems that do exist. |
| 26 |
|
| 27 |
I think standardization should focus on real interoperability issues, |
| 28 |
rather than esoteric technical issues. I think a good way to start |
| 29 |
would be to create some kind of test/regression suite in the portage |
| 30 |
tree that can be used to measure the package manager's functionality. |
| 31 |
Some stuff would be of an obvious pass/fail nature but other things |
| 32 |
can be couched in more subjective terms - like "will unmerge device |
| 33 |
nodes - yes/no " |
| 34 |
|
| 35 |
That at least would allow us to measure where we are in terms of |
| 36 |
interoperability, and identify future areas of improvement. |
| 37 |
|
| 38 |
Like I said, I'm just getting familiar with all this but those are my thoughts. |
| 39 |
|
| 40 |
-Daniel |
| 41 |
|
| 42 |
On 2/20/07, Ciaran McCreesh <ciaranm@×××××××.org> wrote: |
| 43 |
> On Tue, 20 Feb 2007 10:22:14 -0800 Brian Harring <ferringb@×××××.com> |
| 44 |
> wrote: |
| 45 |
> | Perhaps not all of the council; distinctly recall diego pushing about |
| 46 |
> | it though. Quick look through council logs, robbat2 was asking about |
| 47 |
> | timeline also (jan. meeting specifically). |
| 48 |
> |
| 49 |
> You've gotta ask *why* certain people are so keen on pushing it |
| 50 |
> through... Perhaps if they explained why they needed it in such a hurry |
| 51 |
> we would prioritise it differently. |
| 52 |
> |
| 53 |
> | > Several council member have access to it - including me - and |
| 54 |
> | > are quite confident about what is already done. |
| 55 |
> | |
| 56 |
> | The question was specifically in regards to timelines; completion so |
| 57 |
> | that ongoing paludis vs pkgcore vs portage crap can be put to rest. |
| 58 |
> |
| 59 |
> *shrug* I don't see PMS as being viable until there's a fully |
| 60 |
> conformant independent implementation, personally. So that more or |
| 61 |
> less means that for me, PMS will become a priority at around the same |
| 62 |
> time that Paludis 1.0_pre is released. |
| 63 |
> |
| 64 |
> -- |
| 65 |
> Ciaran McCreesh |
| 66 |
> Mail : ciaranm at ciaranm.org |
| 67 |
> Web : http://ciaranm.org/ |
| 68 |
> Paludis, the secure package manager : http://paludis.pioto.org/ |
| 69 |
> |
| 70 |
> |
| 71 |
> |
| 72 |
-- |
| 73 |
gentoo-dev@g.o mailing list |
Archives