| 1 |
On Monday 18 of May 2009 16:52:19 Ciaran McCreesh wrote: |
| 2 |
> On Mon, 18 May 2009 17:47:52 +0300 |
| 3 |
[snip] |
| 4 |
> The definition of soft behaviour allows soft blockers to be treated |
| 5 |
> identically to hard blockers. We had to do it this way because |
| 6 |
> Portage's rules for soft blockers are too fuzzy and arbitrary to turn |
| 7 |
> into a formal specification -- they were a "code first, think later" |
| 8 |
> solution with which we can't do anything useful. |
| 9 |
|
| 10 |
Not sure who is 'we' there, but Portage team already made is useful. Basic |
| 11 |
portage rule for soft-blocks behaviour is "no longer referenced (a'ka 'soft') |
| 12 |
blocked package can be uninstalled cleanly without user intervention" - it's |
| 13 |
well defined behaviour and possible subject of formal specification - it's |
| 14 |
just up to PM to implement block resolution algorithms for corner cases (those |
| 15 |
would not be the subject of formal specification of course, it's just an |
| 16 |
implementation detail like whether to apply rule like order set by '||' |
| 17 |
operator takes precedence over '!' block or order of block appearance in |
| 18 |
RDEPEND sets block precedence) - Zac did good job there saving users |
| 19 |
(especially KDE users) from nightmare of handling all package |
| 20 |
refactoring/blocks manually. |
| 21 |
|
| 22 |
-- |
| 23 |
regards |
| 24 |
MM |
| 25 |
|
| 26 |
|
| 27 |
---------------------------------------------------------------------- |
| 28 |
Oryginalne przepisy na grilla. Zaskocz swoich gosci! |
| 29 |
Sprawdz >>> http://link.interia.pl/f217c |
Archives