| 1 |
-----BEGIN PGP SIGNED MESSAGE----- |
| 2 |
Hash: SHA256 |
| 3 |
|
| 4 |
On 08/08/14 03:56 PM, Kent Fredric wrote: |
| 5 |
> |
| 6 |
> On 9 August 2014 07:34, Peter Stuge <peter@×××××.se |
| 7 |
> <mailto:peter@×××××.se>> wrote: |
| 8 |
> |
| 9 |
> ebuilds often (for me) have artificial dependencies, when the |
| 10 |
> actual version required is too old to be in the tree, but maybe not |
| 11 |
> too old to be installed on an existing system. |
| 12 |
> |
| 13 |
> |
| 14 |
> |
| 15 |
> The inverse is also true, sometimes you see people go: |
| 16 |
> |
| 17 |
> "Well, upstream requires Foo 1.5 at least, but we have 2.0 as the |
| 18 |
> oldest in tree, so we can just say dev-whatever/Foo and be done |
| 19 |
> with it" |
| 20 |
> |
| 21 |
> Which turns out to be horribly wrong if somebody still has Foo 1.4 |
| 22 |
> installed, for whatever reason. |
| 23 |
> |
| 24 |
> And this is just one reason why being excessively lazy about what |
| 25 |
> you upgrade could be secretly detrimental. |
| 26 |
> |
| 27 |
|
| 28 |
Also very true. |
| 29 |
|
| 30 |
I don't think we have any sort of tree-wide policy on this either, do |
| 31 |
we? Although I believe common sense says it's a good idea (and i hope |
| 32 |
most devs do this) to put a minver on a dependency atom if there was |
| 33 |
any ebuild with an older version in the tree within the last year. |
| 34 |
|
| 35 |
-----BEGIN PGP SIGNATURE----- |
| 36 |
Version: GnuPG v2 |
| 37 |
|
| 38 |
iF4EAREIAAYFAlPlMC4ACgkQ2ugaI38ACPDUrgD+OiVN6HQKxNAOusj8PYI1O421 |
| 39 |
Dq2ihfhuQMz2HszG9DoBAJdTZJ9pRM6cFbkN+tcwFc/CAZUiWBe9MsSfoLkqho/C |
| 40 |
=T+NJ |
| 41 |
-----END PGP SIGNATURE----- |
Archives