| 1 |
On 2012-05-23 22:42, Michael Weber wrote: |
| 2 |
> -----BEGIN PGP SIGNED MESSAGE----- |
| 3 |
> Hash: SHA256 |
| 4 |
> |
| 5 |
> Hi, |
| 6 |
> |
| 7 |
> i've looked at the blockers of "[TRACKER] portage migration to git" |
| 8 |
> [1] and want to discuss "testing git-cvsserver" [2]. |
| 9 |
> |
| 10 |
> There are two proposed scenarios how to migrate the developers write |
| 11 |
> access to the portage tree. |
| 12 |
> |
| 13 |
> "Clean cut" turns of cvs access on a given and announced timestamp, |
| 14 |
> rsync-generation/updates is suspended (no input -> no changes), some |
| 15 |
> magic scripts prepare the git repo (according to [3], some hours |
| 16 |
> duration) and we all checkout the tree (might be some funny massive load). |
| 17 |
> |
| 18 |
> "testing git-cvsserver" proses "Clean cut" with the additional ability |
| 19 |
> to continue using cvs update/commit, - in best case - on the old |
| 20 |
> checkout w/o alteration on the developers side. |
| 21 |
> |
| 22 |
> "Clean cut" forces us to clean up out dirty checkouts (I have some |
| 23 |
> added directories, added ebuilds i hesitated to `repoman commit`). |
| 24 |
> Plus we have to alter all our hot-wired portage mangling scripts from |
| 25 |
> cvs'ish to git'ish (I use my read/write checkout as portage tree (cvs |
| 26 |
> checkout + egencache for checkout) and have an automated google-chrome |
| 27 |
> bump script). But this can be accomplished on a per developer basis, |
| 28 |
> and slackers don't stall the process. |
| 29 |
> |
| 30 |
> "testing git-cvsserver" forces us all to test these cvs'ish scripts |
| 31 |
> and behaviours against a git-cvsserver and report. |
| 32 |
> We all know that this test-runs will never happen, stalling this bug |
| 33 |
> till infinity. |
| 34 |
> Plus infra/"subset of devs marshalling the migration" get stuck |
| 35 |
> between fixing git issues and git-cvsserver. |
| 36 |
> |
| 37 |
> *if you still read this* *wow* |
| 38 |
> |
| 39 |
> Please discuss my arguments and come to the conclusions to |
| 40 |
> RESO/WONT-FIX "testing git-cvsserver", make a "clean cut" and remove |
| 41 |
> this bug from the blockers of "[TRACKER] portage migration to git". |
| 42 |
> |
| 43 |
> My lengthy 2 cents. |
| 44 |
> |
| 45 |
> [1] https://bugs.gentoo.org/333531 |
| 46 |
> [2] https://bugs.gentoo.org/333699 |
| 47 |
> [3] https://bugs.gentoo.org/333705#c2 |
| 48 |
> - -- |
| 49 |
> Gentoo Dev |
| 50 |
> http://xmw.de/ |
| 51 |
> -----BEGIN PGP SIGNATURE----- |
| 52 |
> Version: GnuPG v2.0.17 (GNU/Linux) |
| 53 |
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ |
| 54 |
> |
| 55 |
> iF4EAREIAAYFAk+82z0ACgkQknrdDGLu8JBUWAD/dmuqyES/mYDrMam+/txnHmgd |
| 56 |
> VaQaqwHMlwzzqQwbpY4A/0h+5Vp8sLbOE78k4SCaGE2dCQtmeOz0jd1YxkDzP+YW |
| 57 |
> =jXLQ |
| 58 |
> -----END PGP SIGNATURE----- |
| 59 |
> |
| 60 |
> |
| 61 |
Another vote for clean cut. |
Archives