| 1 |
As previously discussed I would like to suggest that the number of |
| 2 |
tolerated EAPIs be reduced. There's been some discussion |
| 3 |
over the last 2+ years, with a weak consensus towards deprecating |
| 4 |
some EAPIs. What, when and how still isn't decided. |
| 5 |
So let's add some data so we have a better idea: |
| 6 |
|
| 7 |
EAPI Statistics: |
| 8 |
|
| 9 |
The daily updated stats [2] show a slow general trend down. |
| 10 |
There's historical data (well, just a few days right now at [3] |
| 11 |
|
| 12 |
The current sum of all ebuilds in tree adds up to ~10% EAPI 1+2 |
| 13 |
EAPI0 is still surprisingly big around ~15% |
| 14 |
EAPI3 is about as big as EAPI2 at around ~7% |
| 15 |
|
| 16 |
According to [1] EAPI 1 and 2 are deprecated. |
| 17 |
At the time EAPI 0 was in limbo as toolchain required it |
| 18 |
(What's the current status of toolchain on that?) |
| 19 |
|
| 20 |
I suggest the following change: |
| 21 |
|
| 22 |
[ EAPI 0 depends on toolchain's acceptance. |
| 23 |
Should toolchain agree then adding EAPI0 ebuilds becomes forbidden] |
| 24 |
Adding EAPI 1 and 2 ebuilds is forbidden. (repoman-fatal) |
| 25 |
EAPI 3 becomes deprecated (like 1 and 2 are now) |
| 26 |
Adding EAPI 3 ebuilds is forbidden in now +3 months |
| 27 |
|
| 28 |
EAPI 4 becomes deprecated when/if there's a new EAPI allowed in-tree |
| 29 |
(EAPI 6, most likely) |
| 30 |
|
| 31 |
EAPI 5 is the recommended default. |
| 32 |
|
| 33 |
|
| 34 |
Rationale: |
| 35 |
More than two supported EAPIs is an unneeded burden on developers. |
| 36 |
Thus 4 will be deprecated when post-5 becomes added. |
| 37 |
|
| 38 |
There's no immediate need to forbid the antepenultimate EAPI immediately. |
| 39 |
|
| 40 |
The goal of this upgrade policy is that we can accelerate the expiration |
| 41 |
of old EAPIs - EAPI 2 happened |
| 42 |
at some point in 2008, I think (or 2009?) and we still carry |
| 43 |
five-year-old cruft around. |
| 44 |
|
| 45 |
Given the percentages - |
| 46 |
|
| 47 |
EAPI 1 can be "cleaned out" by a single motivated individual. It's |
| 48 |
almost gone. |
| 49 |
|
| 50 |
EAPI 2 and 3 will take a few months at least, even if there's a |
| 51 |
coordinated effort to migrate. |
| 52 |
EAPI 0 is as big as 2 and 3 together, and depends on toolchain herd's |
| 53 |
actions. |
| 54 |
It'll still be around for a looong time. (Given the current rate of |
| 55 |
change, plus repoman support, plus people actively working on pruning |
| 56 |
it, I would assume it'll take at least a year, more likely two) |
| 57 |
|
| 58 |
In other words, even if we go for the "fastest" option you won't see any |
| 59 |
revolutionary change :) |
| 60 |
|
| 61 |
Please no bikeshedding, |
| 62 |
|
| 63 |
Patrick |
| 64 |
|
| 65 |
|
| 66 |
[1] https://www.gentoo.org/proj/en/council/meeting-logs/20130409.txt |
| 67 |
[2] http://packages.gentooexperimental.org/eapi-stats.txt |
| 68 |
[3] http://packages.gentooexperimental.org/eapi-history/ |
Archives