When people leave the project, please remove them from OWNERS

[Chris Palmer]

Hello,

I'm finding a lot of OWNERS files containing people who have left Chromium. In some cases, that leaves directories effectively unowned (!). This is bad, especially for security-sensitive directories (e.g. stuff in third_party!).

Please let's go through and remove people who have left, and replace them with people who are still here. :) Thanks! I'll send a few starter CLs and CC this thread to get the party started.

[Chris Palmer]

Here are some starter CLs for your entertainment! :)

• libxslt
• dominicc
• scottmg

dcheng noted that XML and related things are perhaps used only by web platform team? They should at least be co-OWNERS of all XML-related things, even if they are not the exclusive users of XML. Any volunteers or suggestions? As of those CLs, dcheng and I are the XML OWNERS, which is far less than ideal.

[Mark Mentovai]

How do you want to define "left the project?" Googlers being assigned to different management chains doesn't necessarily say anything about activity within Chromium.

--
--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-dev
---
You received this message because you are subscribed to the Google Groups "Chromium-dev" group.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/CAOuvq2122pM6mTkm-ttk%2B9E5F7d1gyeDUhW0v7gY0hRjVBvC7Q%40mail.gmail.com.

[Dirk Pranke]

Generally speaking, from an account management standpoint, we assume nothing about someone's involvement with Chromium (or Chrome) regardless of whether they switch teams in a company or leave the company they're at. Someone can ask to have everything disabled, though, of course.

Separately, we do have a vague practice of disabling accounts after they've been inactive for some extended period of time, and we need to get stricter and cleaner about that.

-- Dirk

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/CAHk1GDwNgiAQzaf7kwXY9qSrN7jxx71Bp9W_O%2B3r8osN_WApCg%40mail.gmail.com.

[Chris Palmer]

Sure, we may have to take it on a case-by-case basis. But that's not a reason to let the status quo, in which some OWNERS are definitely not owning, linger. XML is a key example.

If, for whatever reason, a listed OWNER is not really acting as an OWNER of that directory, then we need to replace them with 1 or more people who really are. We (sheriffs, CL authors, et al.) depend on OWNERS files to be accurate. When they're not, it wastes everyone's time, and sometimes bugs get mis-assigned and go un-handled. I'm running into this just about every time I am security sheriff, and not uncommonly when trying to get CLs reviewed.

[Dirk Pranke]

Yes, I didn't mean to confuse things; code ownership (even in the most mundane sense of getting up-to-date OWNERS files) is definitely still an area where we could improve a bunch.

-- Dirk

[Mark Mentovai]

It's definitely desirable to purge inactive owners and to keep OWNERS files up-to-date. But whatever policy might lead to a conclusion that Scott is not an active, competent owner, at least in the parts of the tree I'm familiar with, is an overreach. You framed this as a blanket thing, but I don't agree that "left the project" alone, at least how you seem to have interpreted it here, is a reasonable place to set the bar.

[Chris Palmer]

On Tue, Dec 4, 2018 at 4:03 PM Mark Mentovai <ma...@chromium.org> wrote:

It's definitely desirable to purge inactive owners and to keep OWNERS files up-to-date. But whatever policy might lead to a conclusion that Scott is not an active, competent owner, at least in the parts of the tree I'm familiar with, is an overreach. You framed this as a blanket thing, but I don't agree that "left the project" alone, at least how you seem to have interpreted it here, is a reasonable place to set the bar.

Understood.

And I would never impugn Scott's (or anyone's!) competence. I don't think I did.

But indeed I am seeing areas of the code that are close to un-owned. There's no blame; this is natural. But we need to fix it, so I am trying.

[bruce...@chromium.org]

I think that the person who is being removed should be the first reviewer on the CL. If they have no comment or if they are okay with it then it can be reviewed by others. That gives the owners a chance to say whether they still want to participate.

[Václav Brožek]

I second Bruce's suggestion that it's good to involve the owner being removed first (as long as they are still responding).

And I share Chris' concern that stale information about people's affiliation with the code is adding unnecessary churn, and can even be dangerous.

Speaking from a very recent experience, there is a surprising amount of things to do when a person is leaving the project:

• Updating OWNERS
• Updating <owner> tags in histograms.xml
• Updating/fixing TODOs
• Updating WATCHLIST

I might even have forgotten something in the above list (suggestions welcome).

Perhaps we could write a short "leaving guide" (I'm happy to volunteer) and try to get into the habit of going through it for every person which leaves. Even if the person who leaves forgets to check the above steps, they usually do have managers and team-mates who could help bringing this topic up.

Cheers,
Vaclav

--

--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-dev
---
You received this message because you are subscribed to the Google Groups "Chromium-dev" group.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/1f9d4c8a-1fac-4ad3-8012-864a4c425dd4%40chromium.org.

--

If you have anonymous feedback for vabr@: go/vabr-feedback

[Václav Brožek]

On Wed, 5 Dec 2018 at 10:26 Václav Brožek <va...@chromium.org> wrote:

I second Bruce's suggestion that it's good to involve the owner being removed first (as long as they are still responding).

And I share Chris' concern that stale information about people's affiliation with the code is adding unnecessary churn, and can even be dangerous.

Speaking from a very recent experience, there is a surprising amount of things to do when a person is leaving the project:

• Updating OWNERS
• Updating <owner> tags in histograms.xml
• Updating/fixing TODOs
• Updating WATCHLIST

One more point: removing the leaving person from the sheriffing rotation, to avoid having unexpected "channel is the sheriff" outages.

 

I might even have forgotten something in the above list (suggestions welcome).

Perhaps we could write a short "leaving guide" (I'm happy to volunteer) and try to get into the habit of going through it for every person which leaves. Even if the person who leaves forgets to check the above steps, they usually do have managers and team-mates who could help bringing this topic up.

Cheers,
Vaclav

On Wed, 5 Dec 2018 at 02:42 <bruce...@chromium.org> wrote:

I think that the person who is being removed should be the first reviewer on the CL. If they have no comment or if they are okay with it then it can be reviewed by others. That gives the owners a chance to say whether they still want to participate.

On Tuesday, December 4, 2018 at 4:26:38 PM UTC-8, Chris Palmer wrote:

On Tue, Dec 4, 2018 at 4:03 PM Mark Mentovai <ma...@chromium.org> wrote:

It's definitely desirable to purge inactive owners and to keep OWNERS files up-to-date. But whatever policy might lead to a conclusion that Scott is not an active, competent owner, at least in the parts of the tree I'm familiar with, is an overreach. You framed this as a blanket thing, but I don't agree that "left the project" alone, at least how you seem to have interpreted it here, is a reasonable place to set the bar.

Understood.

And I would never impugn Scott's (or anyone's!) competence. I don't think I did.

But indeed I am seeing areas of the code that are close to un-owned. There's no blame; this is natural. But we need to fix it, so I am trying.

--
--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-dev
---
You received this message because you are subscribed to the Google Groups "Chromium-dev" group.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/1f9d4c8a-1fac-4ad3-8012-864a4c425dd4%40chromium.org.

--

If you have anonymous feedback for vabr@: go/vabr-feedback

[Dirk Pranke]

Inside Google, we do actually have an exit checklist, but I'm not sure how widely it is used.

-- Dirk

[Daniel Bratell]

There is a fair number of OWNERs that have comments in the code review system that discourages contacting them for reviews. It's not supposed to be like that and just removing very busy owners might not make things better. At least that is better than those who are inactive with no comment. There it can take a long time (days) of experimenting until you find a reviewer.

/Daniel

To unsubscribe from this group and stop receiving emails from it, send an email to chromium-dev...@chromium.org.
To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/CAEoffTBnZYJh%2BurJ3A1odFsUzhGWcxyUBxDmwESU74e%3D8Oomwg%40mail.gmail.com.

--

/* Opera Software, Linköping, Sweden: CET (UTC+1) */

[dan...@chromium.org]

On Wed, Dec 5, 2018 at 3:18 PM Daniel Bratell <bra...@opera.com> wrote:

There is a fair number of OWNERs that have comments in the code review system that discourages contacting them for reviews. It's not supposed to be like that and just removing very busy owners might not make things better. At least that is better than those who are inactive with no comment. There it can take a long time (days) of experimenting until you find a reviewer.

This is exacerbated by the fact that we have been over time trending towards OWNERS in smaller and smaller pieces of the subtree, and I wonder if we should reexamine this practice in favor of larger sets of more central (or a single set) of OWNERS. Some areas have resisted this trend, but it is something I see at large.

 

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/op.ztkf0ei3rbppqq%40cicero2.linkoping.osa.

[Peter Kasting]

On Wed, Dec 5, 2018 at 12:36 PM <dan...@chromium.org> wrote:

On Wed, Dec 5, 2018 at 3:18 PM Daniel Bratell <bra...@opera.com> wrote:

There is a fair number of OWNERs that have comments in the code review system that discourages contacting them for reviews. It's not supposed to be like that and just removing very busy owners might not make things better. At least that is better than those who are inactive with no comment. There it can take a long time (days) of experimenting until you find a reviewer.

This is exacerbated by the fact that we have been over time trending towards OWNERS in smaller and smaller pieces of the subtree, and I wonder if we should reexamine this practice in favor of larger sets of more central (or a single set) of OWNERS.

Given that Chrome largely doesn't use "set noparent", it's already possible today to escalate to higher-level OWNERS if absolutely necessary, so I don't think discouraging fine-grained OWNERS would do anything to aid responsiveness.  If there aren't multiple, responsive OWNERS for areas of the code, we should aim to fix that by adding more people willing to take on greater OWNERship in those areas over time.

OTOH, coarse-grained OWNERS increase "forum-shopping" reviews to the people you already know well, make it more difficult for reviewers to figure out which people will have the most context on the code they're changing, and tend to encourage setting one OWNER as the reviewer of an entire large CL (which is hard on the reviewer).

For those reasons, I would see "move away from fine-grained OWNERS" as a change that incurs several significant costs without providing any benefit, at least until convinced otherwise :).

PK

[dan...@chromium.org]

On Tue, Dec 4, 2018 at 5:39 PM Chris Palmer <pal...@chromium.org> wrote:

Here are some starter CLs for your entertainment! :)

• libxslt
• dominicc
• scottmg

Here's another:

• cc and viz

 

dcheng noted that XML and related things are perhaps used only by web platform team? They should at least be co-OWNERS of all XML-related things, even if they are not the exclusive users of XML. Any volunteers or suggestions? As of those CLs, dcheng and I are the XML OWNERS, which is far less than ideal.

--

--
Chromium Developers mailing list: chromi...@chromium.org
View archives, change email options, or unsubscribe:
http://groups.google.com/a/chromium.org/group/chromium-dev
---
You received this message because you are subscribed to the Google Groups "Chromium-dev" group.

To view this discussion on the web visit https://groups.google.com/a/chromium.org/d/msgid/chromium-dev/CAOuvq20%2BV2RZys%2B3uP%2B%2BUcefvHk%2B_Xe%2BYJgdy%3DWzJ9g9ASRUfQ%40mail.gmail.com.

[Chris Palmer]

Hi all,

I apologize for poorly communicating about this OWNERS cleanup idea. In a distributed team like this, communication is key, and I did not do a great job here. I apologize and I accept the responsibility for how my communications appear. In the future I'll be more careful about how my emails sound.