| 1 |
On Wed, 28 Mar 2012 16:17:06 +0100, David W Noon wrote: |
| 2 |
|
| 3 |
> > Probably. There is rarely a good reason for having libraries in |
| 4 |
> > world. |
| 5 |
> |
| 6 |
> For us programmers it is often essential that we have one or more |
| 7 |
> library packages in world, since we might be using that library (or |
| 8 |
> those libraries) in projects we are developing. |
| 9 |
|
| 10 |
Which is exactly what I was thinking of when I wrote "rarely" and not |
| 11 |
"never". It is also the case if you are installing some out of tree |
| 12 |
software that has dependencies in the tree, but I prefer to handle that |
| 13 |
by creating a set as I can then unmerge the set if I remove the |
| 14 |
software, instead of trying to remember wheat I added to world and why. |
| 15 |
|
| 16 |
> The question I think Todd Goodman is trying to ask is why a package in |
| 17 |
> world should be a candidate for depclean. |
| 18 |
|
| 19 |
Because the other slot satisfies the requirements of world, which |
| 20 |
contains an unslotted version. But then --update always tries to |
| 21 |
installed the newest suitable version. In other words, his system is |
| 22 |
broken. |
| 23 |
|
| 24 |
|
| 25 |
-- |
| 26 |
Neil Bothwick |
| 27 |
|
| 28 |
I can't walk on water, but I can stagger on alcohol. |
Archives