Dev Chat Summary, November 12th

https://wordpress.slack.com/archives/core/p1415826068000473

Misses

  • Beta 1, 2, 3
  • Feature plugin merge window

Decisions

  • Focus will merge and default to on during beta.
  • Sessions UI will be reduced to a button to log out of all other sessions and be merged.
  • Shiny updates will not be in 4.1.
  • Merge tonight (Wednesday) and aim for beta 1 around lunchtime GMT tomorrow (Thursday)
  • Friday’s bug scrub (starting 4PM GMT) will be flow focused. https://core.trac.wordpress.org/query?status=!closed&keywords=~make-flow

Uncertainties

  • There was debate over whether Focus should be on by default. It will be on during beta and feedback will inform the final decision.
  • Twenty Fifteen has an issue with expanding navigation that currently lacks an elegant solution. Fallbacks discussed. https://core.trac.wordpress.org/ticket/30208

Assignments

  • johnbillion will merge Focus
  • johnbillion will merge sessions UI
  • john billion will aim for beta 1 tomorrow
  • ocean90 will update the patch for 29395
  • iamtakashi will make a patch that unsticks the sidebar if there are child menu for #30208

Links Mentioned

https://make.wordpress.org/core/2014/11/11/focus-v2-demo-video/
https://core.trac.wordpress.org/ticket/30264
https://core.trac.wordpress.org/ticket/29395
https://make.wordpress.org/core/2014/11/12/an-update-on-the-taxonomy-roadmap/
https://make.wordpress.org/
https://core.trac.wordpress.org/query?status=!closed&keywords=~make-flow
https://make.wordpress.org/flow/
http://ioshacker.com/how-to/use-quicktime-record-screen-iphone-ipad-ipod-touch-running-ios-8
https://meta.trac.wordpress.org/ticket/482#comment:12
https://core.trac.wordpress.org/ticket/29979
https://core.trac.wordpress.org/ticket/30208

https://core.trac.wordpress.org/ticket/29979

Beta 1

johnbillion [15:04]
So the main question on everyone’s lips is probably beta 1

johnbillion [15:05]
Unfortunately several members of the core team have been preoccupied with 4.0.1, hence we’ve slipped a bit

johnbillion [15:05]
The current plan is thus

johnbillion [15:38]
On that note, I’m aiming to get some merges made tonight and aim for beta1 around lunchtime GMT tomorrow

johnbillion [15:38]
That may shift a little depending on nacin’s availability as I believe he’s doing some talks tomorrow

Focus

johnbillion [15:06]
I will be merging in the editor focus changes that the focus team have been busy working on. Mark posted a good overview video on make yesterday.

https://make.wordpress.org/core/2014/11/11/focus-v2-demo-video/

So this is *in* for 4.1

eric [15:07]
will it be on by default?

johnbillion [15:07]
It will most likely be on by default with an upcoming admin feature pointer to point it out

ipstenu [15:09]
I’d vote for on for beta and see how people react. If Beta goes “woah awesome!” reconsider leaving it on for realz.

johnbillion [15:09]
Yep this is a simple toggle and we can decide based on feedback during beta

Sessions UI

johnbillion [15:20]
So the other merge item is the user session UI, which, after discussion during last week’s meeting, has been simplified by @jorbin to just a button which displays when a user has more than one active session, which will log them out of other sessions when clicked

johnbillion [15:20]
This removes a whole lot of complexity and UI fluff for what is essentially a power user feature

johnbillion [15:21]
It also mirrors what several other services do, primarily Slack because we all love Slack

johnbillion [15:21]
So that is also *in* for 4.1

johnbillion [15:21]
Like everything else, we can iterate on it at a later date

mark [15:22]
Are we still capturing IP data for later?

johnbillion [15:22]
Yes we will be capturing IP, user agent, timestamps

johnbillion [15:23]
See jorbin’s latest patch on #30264

WordPress Trac [15:23]
#30264: Users should have a UI for managing sessions
https://core.trac.wordpress.org/ticket/30264

Shiny Updates

johnbillion [15:28]
So unfortunately due to the work that some core folks have been doing on 4.0.1, the improvements that were slated for the plugin and theme install (and update) screens has not progressed past mockups, so that has been shelved for 4.1

johnbillion [15:29]
This has a bit of a knock-on effect on #29395
#29395: Site Language: Install translations on the fly
https://core.trac.wordpress.org/ticket/29395

obenland [15:29]
Is this planned to be developed plugin first?

johnbillion [15:29]
@obenland: Yeah possibly, although @pento and @nacin (not here) pointed out that dealing with installs and updates via a plugin is a hassle, and working directly on core via patches could end up being easier

ocean90 [15:31]
For 29395 the plan was to ignore FS credentials for 4.1, is this still okay?

johnbillion [15:31]
29395 now has a need for some form of filesystem credentials interstitial, whether it be a modal window or an interim screen, for users who want to install a new language from the general settings screen.

Yes @ocean90, I think that will end up being the case

mark [15:32]
So, it will fail if FS creds required?

ocean90 [15:32]
@mark There will be no option to install a translation on these systems. (edited)
Would be the same as in <4.0. (edited)

johnbillion [15:34]
@ocean90: When do you think you’ll have an updated patch ready?

We can roll with the current patch for beta1 if necessary

ocean90 [15:37]
I have already started, maybe in the next 2 hours. Otherwise end of tomorrow.

Taxonomy Roadmap

johnbillion [15:41]
So hopefully most of you have seen the awesome work @boone has been doing on the taxonomy roadmap. He posted an update earlier today: https://make.wordpress.org/core/2014/11/12/an-update-on-the-taxonomy-roadmap/

johnbillion [15:42]
I believe this puts us in a good position to think about splitting shared terms in 4.2 and beyond

boone [15:43]
Yeah. 4.1 does a lot of the unsexy work required for more exciting taxonomy development in, say, 4.3+.

johnbillion [15:43]
And as boone pointed out, this actually needs to happen across releases due to the nature of schema changes on multisite

boone [15:43]
Shared terms will be split starting in 4.1, when they’re updated. So the most painful part of shared terms – updating one and having others changed unwittingly – will go away right away.

But we’ll wait until the schema changes have been rolled out (say 4.2) to force all terms to be split (ie, outside of `wp_update_term()`).

Mobile Flow

johnbillion [15:48]
Next up, @boren has expressed some concerns over lack of improvements to the mobile workflow, mostly with regard to media, but also for post editing etc

johnbillion [15:48]
Which I tend to agree with

I’m not sure what the best way to tackle this is. Maybe we should use Friday’s bug scrub time as a chance to talk through some high priority mobile stuff rather than bug scrubbing?

johnbillion [15:49]
Thoughts?

We need a bit of focus here, rather than just “mobile stuff”

boren [15:50]
A flow scrub, perhaps. Pick a flow, clean some stuff up.

eric [15:51]
there’s a lot of good ideation on mobile media going on in the Image Flow group; not sure what easy wins can be picked up for 4.1 for that though.

boren [15:51]
There are also the make-flow tagged tickets.

azaozz [15:51]
Good start would be to go through https://core.trac.wordpress.org/query?status=!closed&keywords=~make-flow

stephdau [15:52]
yeah, the trac tag is probably the focus you’re looking for. The leg work has been done (edited)

boren [15:52]
The image flow stuff is looking good, but picking small things from that for 4.1 would be difficult. Getting rid of the media modal sidebar on phones would be awesome, but that opens a big can.

mark [15:53]
Should we maybe try to direct attention to low hanging fruit elsewhere (in mobile flows)?

boren [15:54]
Those make-flow tickets might be the best bet. Several annoyances in there like admin bar behavior.

netweb [15:56]
Adding /flow to https://make.wordpress.org for greater exposure and/or contributions would also be beneficial (edited)

boren [15:58]
I haven’t done it because it’s been a one person show. But now that @mark has posted. :simple_smile:

johnbillion [15:58]
I don’t think that would increase exposure a whole lot. We just need to talk about it more.

mark [15:58]
Also, if anyone wants to be able to post on /flow, /msg me here.

stephdau [15:58]
johnbillion: it can’t impair it either

johnbillion [15:58]
Agreed

mark [15:59]
The “Ryan makes an impassioned personal plea” approach worked well for me. :laughing:

stephdau [15:59]
maybe it’ll be lucky 13(th box on the page)

boren [15:59]
I have a “Be an Alan Lomax of flow” speech that I won’t go into here. :simple_smile:

mark [16:00]
For anyone with iOS 8 and OS X Yosemite, you can tether and record the screen.

mike [16:00]
It’s also a bit strange, since I’m not sure it’s a “contributor group” in the same way the others listed on the root are.

obenland [16:00]
I was wondering how you did it :simple_smile:

boren [16:00]
That’s another reason its not listed.

mike [16:01]
Not that it’s not important, or doesn’t need more exposure; just feels like a different thing. (edited)

mike [16:01]
:simple_smile:

stephdau [16:01]
@obenland: http://ioshacker.com/how-to/use-quicktime-record-screen-iphone-ipad-ipod-touch-running-ios-8 (edited)

netweb [16:01]
See https://meta.trac.wordpress.org/ticket/482#comment:12, icons already added, /flow just needs adding as a listing 🙂

WordPress.org Meta Trac [16:01]
#482: Add Training to “Get Involved”

johnbillion [16:03]
Ok so Friday’s bug scrub (starting 4PM GMT) will be flow-focused

Twenty Fifteen

johnbillion [16:06]
Twenty Fifteen is moving along nicely, although I must admit I’m not sure of the current status of the separate scrolling sidebar

Ah looks like that’s been resolved in #29979

#29979: Twenty Fifteen: Long sidebar produces double-scrollbars

https://core.trac.wordpress.org/ticket/29979

iamtakashi [16:07]
The scroll bar is gone now

iamtakashi [16:08]
we still have an issue with expanding navigation though

johnbillion [16:08]
Expanding the submenu items?

iamtakashi [16:08]
#30208

#30208: Twenty Fifteen: expanding a sub-menu when the sidebar is fixed needs to conditionally unfix the sidebar

https://core.trac.wordpress.org/ticket/30208

iamtakashi [16:09]
I don’t have an elegant solution yet.

johnbillion [16:09]
So @azaozz and I briefly chatted about this at WCSF

iamtakashi [16:09]
I really don’t want to lose sticky sidebar if it’s short

johnbillion [16:09]
We have a similar situation with the admin menu

azaozz [16:10]
Yeah, no good solution exists :disappointed:

iamtakashi [16:10]
At the very last resort, we have to unstick all the time.

mark [16:11]
Do we need to get some brains together to think it through? Or is it 100% at an impasse?

azaozz [16:11]
The current behavior seems pretty logical: after expanding the menu, the user would click on it to go away or can close it to reach the bottom

iamtakashi [16:13]
But if there are only, say, just three menu items that have twenty child menu items, there is no way to reach some of child menu items.

mark [16:14]
Current behavior on http://twentyfifteendemo.wordpress.com is quite problematic.

Idea…

mark [16:15]
What if, upon expanding down beyond the bottom, it unpins, but it doesn’t re-pin until it is small enough to fit, AND is scrolled to the top?
So that way, when contracting a menu, it wouldn’t suddenly jump.
To the original fixed position.
But once it got to that position naturally, it would stick again.

johnbillion [16:16]
You’d need to scroll back up to fix it again

azaozz [16:16]
If you are scrolled half-way through long page, the menu will “dissapear” when unpinned

johnbillion [16:17]
So as Takashi mentioned, unpinning the menu completely is still an option

mark [16:17]
@ozz I don’t follow

mark [16:17]
Oh, maybe I do. But we could work around that.

mark [16:18]
By not letting it unpin to the top.

mark [16:18]
But letting it unpin in its current location.

azaozz [16:18]
The sidebar is `position: fixed` :simple_smile: RIght, workaround would be to add all the pinTop/pinBottom stuff.

mark [16:18]
I’m just thinking of it acting like two panes.

mark [16:19]
When it fits, it sticks. When it doesn’t, it flows naturally. Until it fits again.

iamtakashi [16:19]
I think what azaozz is talking about this: https://cloudup.com/cCxgCbrWaWv

mark [16:20]
Yeah. it shouldn’t go to the top on unpin.

iamtakashi [16:20]
But as mark says, if we can resolve this issue, that’d be great :simple_smile:

mark [16:23]
Solvable technical issues? Tricky technical issues? Or conceptual issues?

iandstewart [16:23]
All the old patches are in #29979

WordPress Trac [16:23]
#29979: Twenty Fifteen: Long sidebar produces double-scrollbars
https://core.trac.wordpress.org/ticket/29979

mark [16:25]
Is there a partial punt solution? Like not sticking if there are submenus?

iamtakashi [16:27]
The current approach is much simpler than any of patches in the ticket and I kind of like unstick when we have child menus.

johnbillion [16:28]
I think this needs to simmer for a while, see if anyone comes up with an idea

mark [16:29]
I don’t think we can sit with what we have. It creates menu items you can’t access.

iamtakashi [16:29]
we can continue on the ticket #30208

WordPress Trac [16:29]
#30208: Twenty Fifteen: expanding a sub-menu when the sidebar is fixed needs to conditionally unfix the sidebar

johnbillion [16:29]
How about a complete un-sticking of the menu as an interim solution

mark [16:33]
No-pin-if-child is a punt, that allows pinning in some situations. Addresses usability issues, while allowing it to be nice an pinned for some sites.

johnbillion [16:35]
Is someone on the theme team able to spend some time on this in the next couple of days or do we need a volunteer?

iamtakashi [16:35]
I’ll try to make a patch that unstick the sidebar if there are child menu.

iandstewart [16:35]
@mark just to be clear, you’re using “punt” for no-pin-if-child in this case, as in “this’d be a good enough fix”? (edited)

mark [16:36]
iandstewart: it would solve the accessibility issue. So it no longer blocks release for that reason.
@iandstewart: whether it’s a functional dealbreaker is another question. I would say… probably not.

mark [16:37]
It’s less than ideal, because in my mind there is a JS solution that would be perfect. But it’s a good chunk of JS work. And we’re running out of time.