On 2/9/2017 8:48 AM, Joseph Rushton Wakeling wrote:
> Contrast this with the experience I had the one time I submitted a (tiny,
> trivial) patch to rust: immediately after submitting the PR I got a message from
> their 'highfive' robot that included:
>
>   * a friendly thank you for the PR;
>
>   * the GitHub ID of a contact who I could expect to be taking responsibility
>     for the PR, who was also assigned as a reviewer;
>
>   * some helpful notes on how to add changes to the PR if requested;
>
>   * a link to the contributor guidelines.
>
> By contrast with a Phobos PR it's not clear who to contact if review or
> decision-making is not forthcoming.

Good idea! Please investigate how to get github to generate such emails. In the meantime, the PR guidelines are here:

  http://wiki.dlang.org/Starting_as_a_Contributor#Create_a_pull_request

The teams can be found here:

  https://github.com/orgs/dlang/teams

Please help improve the guidelines.

On Thursday, 9 February 2017 at 19:36:52 UTC, Walter Bright wrote:
> Good idea! Please investigate how to get github to generate such emails. In the meantime, the PR guidelines are here:

This is already somewhat done with the PR bot we have. The DlangBot notifies reviewers on the DMD repo, but not Phobos for some reason. All it does on Phobos is auto close bugzilla issues when a bug fix PR is pulled into master.

I say enable the bot for Phobos. It would be a small task to also have the bot post the PR guidelines.

I think Martin is the maintainer of the bot.

On 2/9/2017 8:55 AM, Joseph Rushton Wakeling wrote:
> On Thursday, 9 February 2017 at 09:49:53 UTC, Walter Bright wrote:
>> In any case, shouldn't it be an uphill battle to merge things? There are a lot
>> of things that need to be satisfied to merge something. Being too hasty leads
>> to legacy code that we come to regret, angry people whose code was broken, and
>> technical debt.
>
> There's a difference between it being an uphill battle because review and
> feedback are careful, cautious, in-depth and strict (as they should be!), versus
> it being an uphill battle because no feedback or interest is being offered and
> PRs are left to bitrot. :-(

That certainly does happen, but it isn't quite the case with Nick's PRs.


> I accept that there are a lot of things that need to be satisfied to merge
> something.  Personally speaking, I'm willing to endure any number of rebases and
> conflict-fixes, so long as I'm getting feedback and engagement that allows my PR
> to become better code.  It's when I'm _not_ getting any indicators as to what
> needs to be satisfied that things become problematic.

There's a lot going on needing attention, and sometimes a bit of championing is needed by their proponents.

Also, please keep in mind that the smaller and more focussed a PR is, the more likely it'll have a quick resolution.

On Thursday, 9 February 2017 at 19:42:03 UTC, Jack Stouffer wrote:
> On Thursday, 9 February 2017 at 19:36:52 UTC, Walter Bright wrote:
>> Good idea! Please investigate how to get github to generate such emails. In the meantime, the PR guidelines are here:

We gave this a try a couple of months ago with Facebook's mention-bot:

Example: https://github.com/dlang/phobos/pull/4318#issuecomment-241817191
Repo: https://github.com/dlang-bots/mention-bot

Eventually I disabled it because people complained about the added noise.

> This is already somewhat done with the PR bot we have. The DlangBot notifies reviewers on the DMD repo, but not Phobos for some reason. All it does on Phobos is auto close bugzilla issues when a bug fix PR is pulled into master.

It does a bit more (e.g. since a couple of weeks it can also automatically merge PR once all CI status checks pass)

> I say enable the bot for Phobos. It would be a small task to also have the bot post the PR guidelines.

As mentioned above, the bot is already enabled and it receives all interesting hook events from Phobos, see e.g.:

https://github.com/dlang-bots/dlang-bot/blob/master/source/dlangbot/app.d#L124

> I think Martin is the maintainer of the bot.

You can find the code here:

https://github.com/dlang-bots/dlang-bot

On Thursday, 9 February 2017 at 19:58:57 UTC, Seb wrote:
> We gave this a try a couple of months ago with Facebook's mention-bot:
>
> Example: https://github.com/dlang/phobos/pull/4318#issuecomment-241817191
> Repo: https://github.com/dlang-bots/mention-bot
>
> Eventually I disabled it because people complained about the added noise.

It's great that you gave this a go.  Seriously, people considered one small informative post in the PR to be "added noise"? :-(

Out of curiosity: is it typical that it would not post until some way into the discussion (as in that example)?  I could see why it would be irritating if it popped up once discussion and review had already started happening.

Assuming that the bot can be relied on to be 'first responder' to any PR, I'm happy to try to draft an alternative text for it to post (and maybe also look at what texts it can link to).

On Thursday, 9 February 2017 at 19:53:37 UTC, Walter Bright wrote:
> There's a lot going on needing attention, and sometimes a bit of championing is needed by their proponents.

Yes, but it could be good to examine what can be done to more pro-actively look at open PRs that have had no recent follow-up.  It's not really going to work that well if attention gets given substantially on the basis of who's most eagerly seeking it.

People like that metaphor "It's the squeaky wheel that gets the grease", but as a first aid trainer once pointed out to me, the quieter patient may be the one more in need of immediate attention.

> Also, please keep in mind that the smaller and more focussed a PR is, the more likely it'll have a quick resolution.

Unfortunately my current one is large with good reason.  But I think you'll find I also provided very detailed commit messages to explain and justify all my changes ;-)

On 2/9/2017 12:29 PM, Joseph Rushton Wakeling wrote:
> Yes, but it could be good to examine what can be done to more pro-actively look
> at open PRs that have had no recent follow-up.

*Anyone* in this community can step up and do that.

Dne 9.2.2017 v 21:43 Walter Bright via Digitalmars-d-announce napsal(a):

> On 2/9/2017 12:29 PM, Joseph Rushton Wakeling wrote:
>> Yes, but it could be good to examine what can be done to more pro-actively look
>> at open PRs that have had no recent follow-up.
>
> *Anyone* in this community can step up and do that.
This obviously does not work :(

On Thursday, 9 February 2017 at 16:48:16 UTC, Joseph Rushton Wakeling wrote:
> which is that after some initial interest and feedback, the PR just got left alone with no decision to accept or reject it, and no indication of why.

This is why I only contribute to Phobos to be quite honest.

I count ten active reviewers with pull rights on Phobos.

The other repos,

tools: one
dmd: three
druntime: three
dlang.org: two

On Thursday, 9 February 2017 at 20:43:00 UTC, Walter Bright wrote:
> *Anyone* in this community can step up and do that.

Anyone can make observations and proposals, but not everyone has the authority to effect change.

I appreciate how frustrating it must be to have people saying, 'Hey, do this!  Do that!' without necessarily volunteering their own efforts in support, but organizational improvements so very often fail unless they are eagerly pursued at a leadership level.