Add an account setting to opt-in to new UI

Add an account setting to opt-in to new UI features.

[rch, 2013-06-03 16:53:44]

How does this look?

[iannucci, 2013-06-03 19:03:40]

I'd add the codereview-discuss@googlegroups.com to the CC list and M-A to
reviewers since this will land on the default branch :)

https://codereview.appspot.com/9971043/diff/1/templates/user.html
File templates/user.html (right):

https://codereview.appspot.com/9971043/diff/1/templates/user.html#newcode27
templates/user.html:27: {%if use_experimental_ui%}
would it make sense to break this whole section out into a
user_include_experimental.html, and then a user_include.html and simply switch
between two {% include %} directives?

[rch, 2013-06-06 19:38:58]

I've addressed your comment and have not seen other comments.  When would you
like to land?

https://codereview.appspot.com/9971043/diff/1/templates/user.html
File templates/user.html (right):

https://codereview.appspot.com/9971043/diff/1/templates/user.html#newcode27
templates/user.html:27: {%if use_experimental_ui%}
On 2013/06/03 19:03:40, iannucci wrote:
> would it make sense to break this whole section out into a
> user_include_experimental.html, and then a user_include.html and simply switch
> between two {% include %} directives?

Done.

[iannucci, 2013-06-07 03:19:56]

+codereview-discuss@googlegroups.com

I didn't see the discuss group on the CC?

Anyhow, this LGTM. I'll land it, but there's a change blocking the chromium
branch at the moment. I'll land as soon as that clears up.

https://codereview.appspot.com/9971043/diff/16001/templates/user_experimental...
File templates/user_experimental_include.html (right):

https://codereview.appspot.com/9971043/diff/16001/templates/user_experimental...
templates/user_experimental_include.html:1: <tr>
I think this and the other include file can de de-indented by 2 spaces?

[iannucci, 2013-06-07 17:40:42]

+maruel I'd like to get an LGTM on this from an owner :)

[M-A Ruel, 2013-06-07 17:42:42]

On 2013/06/07 17:40:42, iannucci wrote:
> +maruel I'd like to get an LGTM on this from an owner :)

Isn't that overkill? I mean, that's why AppEngine supports multiple concurrent
simultaneous versions.

[iannucci, 2013-06-07 17:49:19]

On 2013/06/07 17:42:42, M-A Ruel wrote:
> On 2013/06/07 17:40:42, iannucci wrote:
> > +maruel I'd like to get an LGTM on this from an owner :)
> 
> Isn't that overkill? I mean, that's why AppEngine supports multiple concurrent
> simultaneous versions.

The purpose is so that we can improve and iterate on the UI without incurring
the wrath of people who are very sensitive to UI changes, especially when those
UI changes are WIP :).

It also lets us surreptitiously gather data about who actually is using the new
stuff v. the old stuff and lets us transition over to the new UI after
'perfecting' it (e.g. sorting CLs v. bolding them, adding a button to un-bold an
issue explicitly, etc.).

[M-A Ruel, 2013-06-07 18:02:54]

On 2013/06/07 17:49:19, iannucci wrote:
> On 2013/06/07 17:42:42, M-A Ruel wrote:
> > On 2013/06/07 17:40:42, iannucci wrote:
> > > +maruel I'd like to get an LGTM on this from an owner :)
> > 
> > Isn't that overkill? I mean, that's why AppEngine supports multiple
concurrent
> > simultaneous versions.
> 
> The purpose is so that we can improve and iterate on the UI without incurring
> the wrath of people who are very sensitive to UI changes, especially when
those
> UI changes are WIP :).
> 
> It also lets us surreptitiously gather data about who actually is using the
new
> stuff v. the old stuff and lets us transition over to the new UI after
> 'perfecting' it (e.g. sorting CLs v. bolding them, adding a button to un-bold
an
> issue explicitly, etc.).

That's to be committed in the chromium branch, right?

[iannucci, 2013-06-07 18:33:20]

On 2013/06/07 18:02:54, M-A Ruel wrote:
> On 2013/06/07 17:49:19, iannucci wrote:
> > On 2013/06/07 17:42:42, M-A Ruel wrote:
> > > On 2013/06/07 17:40:42, iannucci wrote:
> > > > +maruel I'd like to get an LGTM on this from an owner :)
> > > 
> > > Isn't that overkill? I mean, that's why AppEngine supports multiple
> concurrent
> > > simultaneous versions.
> > 
> > The purpose is so that we can improve and iterate on the UI without
incurring
> > the wrath of people who are very sensitive to UI changes, especially when
> those
> > UI changes are WIP :).
> > 
> > It also lets us surreptitiously gather data about who actually is using the
> new
> > stuff v. the old stuff and lets us transition over to the new UI after
> > 'perfecting' it (e.g. sorting CLs v. bolding them, adding a button to
un-bold
> an
> > issue explicitly, etc.).
> 
> That's to be committed in the chromium branch, right?

Well... It was to be committed to both. But it would make sense to change the
default to 'on' for default and 'off' for chromium.

Otherwise the merge conflicts will be crazy.

[anagy4698, 2013-06-07 19:15:19]

2013/6/7  <maruel@google.com>:
> On 2013/06/07 17:49:19, iannucci wrote:
>>
>> On 2013/06/07 17:42:42, M-A Ruel wrote:
>> > On 2013/06/07 17:40:42, iannucci wrote:
>> > > +maruel I'd like to get an LGTM on this from an owner :)
>> >
>> > Isn't that overkill? I mean, that's why AppEngine supports multiple
>
> concurrent
>>
>> > simultaneous versions.
>
>
>> The purpose is so that we can improve and iterate on the UI without
>
> incurring
>>
>> the wrath of people who are very sensitive to UI changes, especially
>
> when those
>>
>> UI changes are WIP :).
>
>
>> It also lets us surreptitiously gather data about who actually is
>
> using the new
>>
>> stuff v. the old stuff and lets us transition over to the new UI after
>> 'perfecting' it (e.g. sorting CLs v. bolding them, adding a button to
>
> un-bold an
>>
>> issue explicitly, etc.).
>
>
> That's to be committed in the chromium branch, right?
>
> https://codereview.appspot.com/9971043/
>
> --
> You received this message because you are subscribed to the Google Groups
> "codereview-discuss" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to codereview-discuss+unsubscribe@googlegroups.com.
> To post to this group, send email to codereview-discuss@googlegroups.com.
> Visit this group at http://groups.google.com/group/codereview-discuss?hl=en.
> For more options, visit https://groups.google.com/groups/opt_out.
>
>

[rch, 2013-06-17 23:49:25]

On 2013/06/07 18:33:20, iannucci wrote:
> On 2013/06/07 18:02:54, M-A Ruel wrote:
> > On 2013/06/07 17:49:19, iannucci wrote:
> > > On 2013/06/07 17:42:42, M-A Ruel wrote:
> > > > On 2013/06/07 17:40:42, iannucci wrote:
> > > > > +maruel I'd like to get an LGTM on this from an owner :)
> > > > 
> > > > Isn't that overkill? I mean, that's why AppEngine supports multiple
> > concurrent
> > > > simultaneous versions.
> > > 
> > > The purpose is so that we can improve and iterate on the UI without
> incurring
> > > the wrath of people who are very sensitive to UI changes, especially when
> > those
> > > UI changes are WIP :).
> > > 
> > > It also lets us surreptitiously gather data about who actually is using
the
> > new
> > > stuff v. the old stuff and lets us transition over to the new UI after
> > > 'perfecting' it (e.g. sorting CLs v. bolding them, adding a button to
> un-bold
> > an
> > > issue explicitly, etc.).
> > 
> > That's to be committed in the chromium branch, right?
> 
> Well... It was to be committed to both. But it would make sense to change the
> default to 'on' for default and 'off' for chromium.
> 
> Otherwise the merge conflicts will be crazy.

M-A: *ping*

[M-A, 2013-06-17 23:54:44]

On 2013/06/07 18:33:20, iannucci wrote:
> On 2013/06/07 18:02:54, M-A Ruel wrote:
> > That's to be committed in the chromium branch, right?
> 
> Well... It was to be committed to both. But it would make sense to change the
> default to 'on' for default and 'off' for chromium.
> 
> Otherwise the merge conflicts will be crazy.

Wait, you mean the default branch gets all the experimental stuff enabled by
default and the chromium branch doesn't get these?

[iannucci, 2013-06-17 23:57:47]

On 2013/06/17 23:54:44, M-A wrote:
> On 2013/06/07 18:33:20, iannucci wrote:
> > On 2013/06/07 18:02:54, M-A Ruel wrote:
> > > That's to be committed in the chromium branch, right?
> > 
> > Well... It was to be committed to both. But it would make sense to change
the
> > default to 'on' for default and 'off' for chromium.
> > 
> > Otherwise the merge conflicts will be crazy.
> 
> Wait, you mean the default branch gets all the experimental stuff enabled by
> default and the chromium branch doesn't get these?

The 'experimental' stuff is the interface which has already been deployed (i.e.
what you see when you go to your own issues page). Some ppl on chromium-dev
would like it to be better before deploying.

I was assuming it would be on for default since the people on default seem to
like it.... I have no preference either way.

[anagy4698, 2013-06-18 09:46:40]

2013/6/18 <rch@chromium.org>

> On 2013/06/07 18:33:20, iannucci wrote:
>
>> On 2013/06/07 18:02:54, M-A Ruel wrote:
>> > On 2013/06/07 17:49:19, iannucci wrote:
>> > > On 2013/06/07 17:42:42, M-A Ruel wrote:
>> > > > On 2013/06/07 17:40:42, iannucci wrote:
>> > > > > +maruel I'd like to get an LGTM on this from an owner :)
>> > > >
>> > > > Isn't that overkill? I mean, that's why AppEngine supports
>>
> multiple
>
>> > concurrent
>> > > > simultaneous versions.
>> > >
>> > > The purpose is so that we can improve and iterate on the UI
>>
> without
>
>> incurring
>> > > the wrath of people who are very sensitive to UI changes,
>>
> especially when
>
>> > those
>> > > UI changes are WIP :).
>> > >
>> > > It also lets us surreptitiously gather data about who actually is
>>
> using the
>
>> > new
>> > > stuff v. the old stuff and lets us transition over to the new UI
>>
> after
>
>> > > 'perfecting' it (e.g. sorting CLs v. bolding them, adding a button
>>
> to
>
>> un-bold
>> > an
>> > > issue explicitly, etc.).
>> >
>> > That's to be committed in the chromium branch, right?
>>
>
>  Well... It was to be committed to both. But it would make sense to
>>
> change the
>
>> default to 'on' for default and 'off' for chromium.
>>
>
>  Otherwise the merge conflicts will be crazy.
>>
>
> M-A: *ping*
>
>
https://codereview.appspot.**com/9971043/<https://codereview.appspot.com/9971...
>

[Andi, 2013-06-18 19:13:57]

Sorry, I didn't followed the whole discussion. What's the long-term purpose of
this patch? For now it only seems to make the CSS class "issue updated"
toggleable. I suppose there's more in the long run?

[iannucci, 2013-06-18 19:18:47]

On 2013/06/18 19:13:57, Andi wrote:
> Sorry, I didn't followed the whole discussion. What's the long-term purpose of
> this patch? For now it only seems to make the CSS class "issue updated"
> toggleable. I suppose there's more in the long run?

Essentially, there's some bikeshedding on chromium-dev about the recent rietveld
interface changes. The purpose of this patch is to hide new UI changes (and
future developments thereof) behind an opt-in boolean so that changes don't
upset people while we perfect them. It also allows us to inspect the DB and get
a 'real' poll of who is using the new interface v. those who aren't.

The only reason this would need to go into default is so that we can continue to
develop the new UI on both branches without having nightmare merging conflicts.

[Andi, 2013-06-18 19:34:34]

On Tue, Jun 18, 2013 at 9:18 PM,  <iannucci@chromium.org> wrote:
> On 2013/06/18 19:13:57, Andi wrote:
>>
>> Sorry, I didn't followed the whole discussion. What's the long-term
>
> purpose of
>>
>> this patch? For now it only seems to make the CSS class "issue
>
> updated"
>>
>> toggleable. I suppose there's more in the long run?
>
>
> Essentially, there's some bikeshedding on chromium-dev about the recent
> rietveld interface changes. The purpose of this patch is to hide new UI
> changes (and future developments thereof) behind an opt-in boolean so
> that changes don't upset people while we perfect them. It also allows us
> to inspect the DB and get a 'real' poll of who is using the new
> interface v. those who aren't.

Interesting! What are the key arguments (I'm not on chromium-dev)? At
first I was a bit confused too when I first saw the UI changes in
production. And I'm still not totally convinced that they work as I'd
expected them to work. However they point in the right direction and I
like your phrase "while we perfect them" :)
oh, and my "not totally convinced" may be the same reason as for the
bikeshedding on chromium-dev: I have no concrete point. It's just the
issues show up as updated where I've already read the comment. What we
really need to do is to track a "read/unread" status and I'd like to
use labels for that.

>
> The only reason this would need to go into default is so that we can
> continue to develop the new UI on both branches without having nightmare
> merging conflicts.

It's not thought-out, but what about marking different topics with
different tags then? For example the update/unread/new improvements to
the issue list could have a tag like "experimental:updatemark"
(instead just "experimental"). That would make it much easier to
officially move a feature from experimental to stable. Mainly because
we can just grep the code for such a tag to remove it. The downside is
that we would have to include additional CSS files then or work out a
naming scheme that could easily be refactored if a experimental
feature goes live.

--
Andi

>
> https://codereview.appspot.com/9971043/

[M-A Ruel, 2013-06-20 14:34:26]

On my side, I think it's more a naming issue; it's not experimental if it's on
by default on the default branch. It's just "new vs old".

[sec.test.dragnet, 2025-07-07 19:31:57]

irrelevant