1 |
On 05.11.2016 00:15, Zac Medico wrote: |
2 |
> On 11/04/2016 03:55 PM, Zac Medico wrote: |
3 |
>> On 11/04/2016 03:47 PM, Brian Dolbec wrote: |
4 |
>>> On Fri, 4 Nov 2016 13:53:02 -0700 |
5 |
>>> Zac Medico <zmedico@g.o> wrote: |
6 |
>>> |
7 |
>>>> On 11/04/2016 01:43 PM, Michał Górny wrote: |
8 |
>>>>> On Fri, 4 Nov 2016 13:19:39 -0700 |
9 |
>>>>> Zac Medico <zmedico@g.o> wrote: |
10 |
>>>>> |
11 |
>>>>>> On 11/04/2016 01:14 PM, Brian Dolbec wrote: |
12 |
>>>>>>> On Thu, 3 Nov 2016 15:55:23 -0700 |
13 |
>>>>>>> Zac Medico <zmedico@g.o> wrote: |
14 |
>>>>>>> |
15 |
>>>>>>>> In about a week, portage-2.3.2 will be eligible for a stable |
16 |
>>>>>>>> request. |
17 |
>>>>>>>> |
18 |
>>>>>>>> The only potential problem that I've noticed is the complaint |
19 |
>>>>>>>> about changes from bug 552814 causing issues for people using |
20 |
>>>>>>>> git sync with overlay filesystems, but setting sync-depth = 0 |
21 |
>>>>>>>> gives those users a workaround. There's also bug 597838, about |
22 |
>>>>>>>> the sync-depth setting being ineffective, but I only know of a |
23 |
>>>>>>>> couple of people that have been able to reproduce that. |
24 |
>>>>>>>> |
25 |
>>>>>>>> So, do we want to do a stable request portage-2.3.2 when the time |
26 |
>>>>>>>> comes? |
27 |
>>>>>>> |
28 |
>>>>>>> I'm not sure. Do we -r1 it adding a patch or two and ask it be |
29 |
>>>>>>> stabled? |
30 |
>>>>>> |
31 |
>>>>>> There are just 4 commits since 2.3.2, and they all look good. |
32 |
>>>>>> Maybe we should just cut a 2.3.3 release and wait another 30 days |
33 |
>>>>>> (we also need to stabilize app-crypt/gkeys since it's needed by |
34 |
>>>>>> emerge-webrsync now). |
35 |
>>>>> |
36 |
>>>>> Wouldn't it be better to have a really working version of gkeys |
37 |
>>>>> before it's stabilized? Like one that could be used without having |
38 |
>>>>> to create custom configuration files and/or run it as root? |
39 |
>>>> |
40 |
>>>> Well, gkeys stabilization is not really mandatory, since |
41 |
>>>> emerge-webrsync has a --insecure option. |
42 |
>>> |
43 |
>>> Why don't I/we work on whatever changes are needed to merge the |
44 |
>>> meta-manifest code to both portage and gkeys. I'll push out another |
45 |
>>> release. I also had some initial code that added gkeys use to verify |
46 |
>>> the pkg Manifest file, but I don't know if that is needed still, the |
47 |
>>> meta-manifest system will need to run a verify at the end of the sync. |
48 |
>>> |
49 |
>>> We'll have to poke Robin some more to get some new infra keys setup. |
50 |
>>> |
51 |
>>> If I have to, maybe I'll create some ansible scripts to run the dev |
52 |
>>> seeds update on vulture, transfer it to my system to push --sign to |
53 |
>>> api.g.o or break down and get Kristian to help me get key forwarding |
54 |
>>> better setup so I can do it from vulture. |
55 |
>> |
56 |
>> Sounds good, but I think we should cut a portage 2.3.3 release before we |
57 |
>> make any more changes. Maybe do a release branch that includes |
58 |
>> everything except the emerge-webrsync change. |
59 |
> |
60 |
> Let's just revert the emerge-webrsync patch, so we can tag a 2.3.3 |
61 |
> release on the master branch. |
62 |
> |
63 |
|
64 |
Will repoman be released with the same tag as well or is the portage and |
65 |
repoman version not going to be syncronized? |
66 |
|
67 |
Cheers, |
68 |
|
69 |
Manuel |