* Re: GNU GRUB maintenance
2015-10-08 14:52 ` Andrei Borzenkov
@ 2015-10-08 19:34 ` Konrad Rzeszutek Wilk
2015-10-09 12:14 ` Vladimir 'φ-coder/phcoder' Serbinenko
2015-10-09 4:41 ` Fajar A. Nugraha
2015-10-09 12:10 ` Vladimir 'φ-coder/phcoder' Serbinenko
2 siblings, 1 reply; 16+ messages in thread
From: Konrad Rzeszutek Wilk @ 2015-10-08 19:34 UTC (permalink / raw)
To: The development of GNU GRUB, Andrei Borzenkov
On October 8, 2015 10:52:25 AM EDT, Andrei Borzenkov <arvidjaar@gmail.com> wrote:
>On Thu, Oct 8, 2015 at 12:14 AM, Vladimir 'φ-coder/phcoder' Serbinenko
><phcoder@gmail.com> wrote:
>> Hello, all. I'm sorry for not being available to do enough
>maintenance
>> for GRUB in last time but I was overbooked. Yet there is a good news.
>At
>> Google there is a 20% project and GRUB has been approved as 20%
>project
>> for me. The goal is to have 2.02 released before the end of this
>year.
>> Other than the raw lack of time there is another issue which makes
>> maintenance difficult: inefficient VCS.
>
>VCS is actually OK. The project of size Linux kernel seems to work
>well using pull request e-mails. The disadvantages are
>
>- contributors must have repository available via Internet
That is quite easy nowadays. And you can always ask for signed tags if you are worried about repos being subverted.
>- contributors are trusted to actually submit pull request for branch
>that was reviewed
<blinks>
It is a disadvantage to trust people!?
>- it needs to be done locally and pushed
Or you can have different maintainers pushing the patches in if they are Acked or Reviewed.
Meaning the committee does not have to be the same person who reviews/acks it.
>
>> It requires me
>or someone with
>> privileges manually copy the patch. What other systems would be ok?
>It
>> obviously has to be a free software and hosted on free
>software-friendly
>> hosting. It also has to have an efficient 1-click merge (so that
>someone
>> with privileges can get any patch submitted to the system merged in
>> couple of clicks).
>>
>>
Clicks? That sounds like a GUI thing. And it sounds like you need to have an admin to set it up, patch it occasionally, deal with spammers, etc.
What is wrong with the old mechanism of emails.
>
>It does not like like we have much choice. If we speak about free
>external hosting, this is probably github, gerrithub, gitlab. I do not
>know if any of them is considered friendly enough by FSF.
>
>If we speak about self hosting, then it is probably gerrit and
>reviewboard (I wish we could join KDE reviewboard, but grub hardly can
>be called KDE application ... :) )
>
>I am not thrilled by github workflows. From what I could gather
>gerrithub looks more appealing, but would love to hear from someone
>who actually used both.
>
>One problem is that none of them apparently allows reviewing by
>E-Mail. This worked (and probably works, just I'm no more involved)
>quite well in KDE reviewboard. This means all review must be done via
>web. For me it is rather disadvantage. Also merged requests are
>removed, which means history and past discussions are no more present.
>Which again is better using e-mail review.
Aye!
>
>_______________________________________________
>Grub-devel mailing list
>Grub-devel@gnu.org
>https://lists.gnu.org/mailman/listinfo/grub-devel
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: GNU GRUB maintenance
2015-10-08 19:34 ` Konrad Rzeszutek Wilk
@ 2015-10-09 12:14 ` Vladimir 'φ-coder/phcoder' Serbinenko
2015-10-09 12:18 ` Andrei Borzenkov
2015-10-10 0:30 ` Konrad Rzeszutek Wilk
0 siblings, 2 replies; 16+ messages in thread
From: Vladimir 'φ-coder/phcoder' Serbinenko @ 2015-10-09 12:14 UTC (permalink / raw)
To: The development of GNU GRUB, Andrei Borzenkov
[-- Attachment #1: Type: text/plain, Size: 2370 bytes --]
On 08.10.2015 21:34, Konrad Rzeszutek Wilk wrote:
> On October 8, 2015 10:52:25 AM EDT, Andrei Borzenkov <arvidjaar@gmail.com> wrote:
>> On Thu, Oct 8, 2015 at 12:14 AM, Vladimir 'φ-coder/phcoder' Serbinenko
>> <phcoder@gmail.com> wrote:
>>> Hello, all. I'm sorry for not being available to do enough
>> maintenance
>>> for GRUB in last time but I was overbooked. Yet there is a good news.
>> At
>>> Google there is a 20% project and GRUB has been approved as 20%
>> project
>>> for me. The goal is to have 2.02 released before the end of this
>> year.
>>> Other than the raw lack of time there is another issue which makes
>>> maintenance difficult: inefficient VCS.
>>
>> VCS is actually OK. The project of size Linux kernel seems to work
>> well using pull request e-mails. The disadvantages are
>>
>> - contributors must have repository available via Internet
>
>
> That is quite easy nowadays. And you can always ask for signed tags if you are worried about repos being subverted.
>
>> - contributors are trusted to actually submit pull request for branch
>> that was reviewed
>
>
> <blinks>
>
> It is a disadvantage to trust people!?
>
>
>> - it needs to be done locally and pushed
>
>
> Or you can have different maintainers pushing the patches in if they are Acked or Reviewed.
>
> Meaning the committee does not have to be the same person who reviews/acks it.
>
>>
>>> It requires me
>> or someone with
>>> privileges manually copy the patch. What other systems would be ok?
>> It
>>> obviously has to be a free software and hosted on free
>> software-friendly
>>> hosting. It also has to have an efficient 1-click merge (so that
>> someone
>>> with privileges can get any patch submitted to the system merged in
>>> couple of clicks).
>>>
>>>
>
> Clicks? That sounds like a GUI thing. And it sounds like you need to have an admin to set it up, patch it occasionally, deal with spammers, etc.
>
> What is wrong with the old mechanism of emails.
>
It takes too much effort to:
a) Track if there are any unresolved issues
b) It takes non-trivial amount of effort to commit once it's reviewed:
you need to copy patch from mail client to git, do commit, copy
description and so on
c) No integration with continous testing systems
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 213 bytes --]
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: GNU GRUB maintenance
2015-10-09 12:14 ` Vladimir 'φ-coder/phcoder' Serbinenko
@ 2015-10-09 12:18 ` Andrei Borzenkov
2015-10-10 0:30 ` Konrad Rzeszutek Wilk
1 sibling, 0 replies; 16+ messages in thread
From: Andrei Borzenkov @ 2015-10-09 12:18 UTC (permalink / raw)
To: The development of GNU GRUB
On Fri, Oct 9, 2015 at 3:14 PM, Vladimir 'φ-coder/phcoder' Serbinenko
<phcoder@gmail.com> wrote:
> It takes too much effort to:
> a) Track if there are any unresolved issues
> b) It takes non-trivial amount of effort to commit once it's reviewed:
> you need to copy patch from mail client to git, do commit, copy
> description and so on
Which is why I usually ask to do 'git send-mail" which can be applied
using "git am". Of course all other points are still valid.
> c) No integration with continous testing systems
>
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: GNU GRUB maintenance
2015-10-09 12:14 ` Vladimir 'φ-coder/phcoder' Serbinenko
2015-10-09 12:18 ` Andrei Borzenkov
@ 2015-10-10 0:30 ` Konrad Rzeszutek Wilk
2015-10-10 1:24 ` Konrad Rzeszutek Wilk
1 sibling, 1 reply; 16+ messages in thread
From: Konrad Rzeszutek Wilk @ 2015-10-10 0:30 UTC (permalink / raw)
To: The development of GNU GRUB,
Vladimir 'φ-coder/phcoder' Serbinenko,
Andrei Borzenkov
On October 9, 2015 8:14:39 AM EDT, "Vladimir 'φ-coder/phcoder' Serbinenko" <phcoder@gmail.com> wrote:
>On 08.10.2015 21:34, Konrad Rzeszutek Wilk wrote:
>> On October 8, 2015 10:52:25 AM EDT, Andrei Borzenkov
><arvidjaar@gmail.com> wrote:
>>> On Thu, Oct 8, 2015 at 12:14 AM, Vladimir 'φ-coder/phcoder'
>Serbinenko
>>> <phcoder@gmail.com> wrote:
>>>> Hello, all. I'm sorry for not being available to do enough
>>> maintenance
>>>> for GRUB in last time but I was overbooked. Yet there is a good
>news.
>>> At
>>>> Google there is a 20% project and GRUB has been approved as 20%
>>> project
>>>> for me. The goal is to have 2.02 released before the end of this
>>> year.
>>>> Other than the raw lack of time there is another issue which makes
>>>> maintenance difficult: inefficient VCS.
>>>
>>> VCS is actually OK. The project of size Linux kernel seems to work
>>> well using pull request e-mails. The disadvantages are
>>>
>>> - contributors must have repository available via Internet
>>
>>
>> That is quite easy nowadays. And you can always ask for signed tags
>if you are worried about repos being subverted.
>>
>>> - contributors are trusted to actually submit pull request for
>branch
>>> that was reviewed
>>
>>
>> <blinks>
>>
>> It is a disadvantage to trust people!?
>>
>>
>>> - it needs to be done locally and pushed
>>
>>
>> Or you can have different maintainers pushing the patches in if they
>are Acked or Reviewed.
>>
>> Meaning the committee does not have to be the same person who
>reviews/acks it.
>>
>>>
>>>> It requires
>me
>>> or someone with
>>>> privileges manually copy the patch. What other systems would be ok?
>>> It
>>>> obviously has to be a free software and hosted on free
>>> software-friendly
>>>> hosting. It also has to have an efficient 1-click merge (so that
>>> someone
>>>> with privileges can get any patch submitted to the system merged in
>>>> couple of clicks).
>>>>
>>>>
>>
>> Clicks? That sounds like a GUI thing. And it sounds like you need to
>have an admin to set it up, patch it occasionally, deal with spammers,
>etc.
>>
>> What is wrong with the old mechanism of emails.
>>
>It takes too much effort to:
>a) Track if there are any unresolved issues
Isn't that the job of the folks submitting the patches?
>b) It takes non-trivial amount of effort to commit once it's reviewed:
>you need to copy patch from mail client to git, do commit, copy
>description and so on
Huh? 'git am' takes your patches in mbox format and commits them in. With description and all.
I just save the emails from the mail client and then apply them all in one go with 'git am -s'.
>c) No integration with continuous testing systems
There is no continuous testing at all now.
But if you want - the 0-day build system picks up emails posted on LKML and compiles them to at least test that are compilable.
>
>
>
>------------------------------------------------------------------------
>
>_______________________________________________
>Grub-devel mailing list
>Grub-devel@gnu.org
>https://lists.gnu.org/mailman/listinfo/grub-devel
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: GNU GRUB maintenance
2015-10-10 0:30 ` Konrad Rzeszutek Wilk
@ 2015-10-10 1:24 ` Konrad Rzeszutek Wilk
0 siblings, 0 replies; 16+ messages in thread
From: Konrad Rzeszutek Wilk @ 2015-10-10 1:24 UTC (permalink / raw)
To: The development of GNU GRUB,
Vladimir 'φ-coder/phcoder' Serbinenko,
Andrei Borzenkov
On Fri, Oct 09, 2015 at 08:30:01PM -0400, Konrad Rzeszutek Wilk wrote:
> On October 9, 2015 8:14:39 AM EDT, "Vladimir 'φ-coder/phcoder' Serbinenko" <phcoder@gmail.com> wrote:
> >On 08.10.2015 21:34, Konrad Rzeszutek Wilk wrote:
> >> On October 8, 2015 10:52:25 AM EDT, Andrei Borzenkov
> ><arvidjaar@gmail.com> wrote:
> >>> On Thu, Oct 8, 2015 at 12:14 AM, Vladimir 'φ-coder/phcoder'
> >Serbinenko
> >>> <phcoder@gmail.com> wrote:
> >>>> Hello, all. I'm sorry for not being available to do enough
> >>> maintenance
> >>>> for GRUB in last time but I was overbooked. Yet there is a good
> >news.
> >>> At
> >>>> Google there is a 20% project and GRUB has been approved as 20%
> >>> project
> >>>> for me. The goal is to have 2.02 released before the end of this
> >>> year.
> >>>> Other than the raw lack of time there is another issue which makes
> >>>> maintenance difficult: inefficient VCS.
> >>>
> >>> VCS is actually OK. The project of size Linux kernel seems to work
Looking at the grub2 mailing list there hasn't been a ton of patches.
And even if there is a pickup of it - is it going to be around 1K a
month or more like in 100s?
If it is going to be in 100~ or so per month - I would think that
the existing mechanism of:
1) save email in mbox file
2) at the end of day, 'git am -s' < mbox file.
3) git push
Would work - it is a five minute job. (This is assuming that the patches
don't break a build and so on - if they do, then you can break this
up and complain to the folks that post patches that it does not
build,etc).
If you are thinking there is going to 1K or so then some automated
system makes sense... but if you are trying to get a release out before
the end of this year and only have 10 business days to do this
(20% of 40 week = 8 hours, and there are 10 weeks left), then I would
concentrate on the existing patches + review.
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: GNU GRUB maintenance
2015-10-08 14:52 ` Andrei Borzenkov
2015-10-08 19:34 ` Konrad Rzeszutek Wilk
@ 2015-10-09 4:41 ` Fajar A. Nugraha
2015-10-09 12:10 ` Vladimir 'φ-coder/phcoder' Serbinenko
2 siblings, 0 replies; 16+ messages in thread
From: Fajar A. Nugraha @ 2015-10-09 4:41 UTC (permalink / raw)
To: The development of GNU GRUB
On Thu, Oct 8, 2015 at 9:52 PM, Andrei Borzenkov <arvidjaar@gmail.com> wrote:
> One problem is that none of them apparently allows reviewing by
> E-Mail.
Github partially support that. If you're watching a repo or has commit
access to it, you should get notification emails for every issue. And
you can review by replying to it
https://github.com/blog/811-reply-to-comments-from-email
Of course it DOESN'T support email-review from users with
non-github-registered email.
> Also merged requests are
> removed, which means history and past discussions are no more present.
What do you mean removed? I believe it's preserved, just not the
default view on issue page.
Here's a link to merged request from another project on github:
https://github.com/FreeRADIUS/freeradius-server/pull/1270
... and here's a link to list of closed pull requests, both merged
(marked with a tick mark) and rejected (marked with an x):
https://github.com/FreeRADIUS/freeradius-server/pulls?page=2&q=is%3Apr+is%3Aclosed
--
Fajar
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: GNU GRUB maintenance
2015-10-08 14:52 ` Andrei Borzenkov
2015-10-08 19:34 ` Konrad Rzeszutek Wilk
2015-10-09 4:41 ` Fajar A. Nugraha
@ 2015-10-09 12:10 ` Vladimir 'φ-coder/phcoder' Serbinenko
2 siblings, 0 replies; 16+ messages in thread
From: Vladimir 'φ-coder/phcoder' Serbinenko @ 2015-10-09 12:10 UTC (permalink / raw)
To: The development of GNU GRUB
[-- Attachment #1: Type: text/plain, Size: 2906 bytes --]
On 08.10.2015 16:52, Andrei Borzenkov wrote:
> On Thu, Oct 8, 2015 at 12:14 AM, Vladimir 'φ-coder/phcoder' Serbinenko
> <phcoder@gmail.com> wrote:
>> Hello, all. I'm sorry for not being available to do enough maintenance
>> for GRUB in last time but I was overbooked. Yet there is a good news. At
>> Google there is a 20% project and GRUB has been approved as 20% project
>> for me. The goal is to have 2.02 released before the end of this year.
>> Other than the raw lack of time there is another issue which makes
>> maintenance difficult: inefficient VCS.
>
> VCS is actually OK. The project of size Linux kernel seems to work
> well using pull request e-mails. The disadvantages are
>
> - contributors must have repository available via Internet
> - contributors are trusted to actually submit pull request for branch
> that was reviewed
> - it needs to be done locally and pushed
>
>> It requires me or someone with
>> privileges manually copy the patch. What other systems would be ok? It
>> obviously has to be a free software and hosted on free software-friendly
>> hosting. It also has to have an efficient 1-click merge (so that someone
>> with privileges can get any patch submitted to the system merged in
>> couple of clicks).
>>
>>
>
> It does not like like we have much choice. If we speak about free
> external hosting, this is probably github, gerrithub, gitlab. I do not
> know if any of them is considered friendly enough by FSF.
>
> If we speak about self hosting, then it is probably gerrit and
> reviewboard (I wish we could join KDE reviewboard, but grub hardly can
> be called KDE application ... :) )
>
> I am not thrilled by github workflows. From what I could gather
> gerrithub looks more appealing, but would love to hear from someone
> who actually used both.
>
I spoke with Stefan Reinauer and he proposed to host us at
review.coreboot.org if we don't generate too much traffic. I had
positive experiences with their gerrit except that some functions are
broken on mobile. I'd like to be able to review from phone but it's not
a hard requirement.
> One problem is that none of them apparently allows reviewing by
> E-Mail. This worked (and probably works, just I'm no more involved)
> quite well in KDE reviewboard. This means all review must be done via
> web. For me it is rather disadvantage.
That's a disadvantage but I believe that being able to get changes
merged quickly outweights this disadvantage.
> Also merged requests are
> removed, which means history and past discussions are no more present.
They're kept on review.coreboot.org case
> Which again is better using e-mail review.
>
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> https://lists.gnu.org/mailman/listinfo/grub-devel
>
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 213 bytes --]
^ permalink raw reply [flat|nested] 16+ messages in thread