[PatternFly] modals and wizards -- should they be movable?

Thu Mar 23 20:47:31 UTC 2017

Hi

The guideline as far as I know, and it is also reflected here is that the
need to move dialogs around is a sign of poor design.

However, if something is missing in a dialog, and making me want to move
it, then perhaps some data need to reside on the dialog itself?

We are also looking into changing the UI design to something similar to
cfme, in which there are no pop up dialogs, but you browse to a new view
and once you finish you go back to the previous one.... Which means you
can't move it neither see anything behind it. It isn't an issue on cfme,
satellite, openshift and more.... Not sure why it is an issue for RHV. Not
any "regression" is a bad thing. Sometimes it fixes a poor decision done in
the past.

If something is missing in any dialog, causing you to move it, then perhaps
it is a bug that should be opened on the relevant team who owns it, so that
they can improve it.

Keep in mind that the reason behind the changes we have made in the last
few months, and also now, is to aligh RHV to patternfly, just like the
other products. I'm not sure whether you had the chance to see how exactly
the dialogs look like now. If you do, I'm sure you'll be impresses, and so
will our customers. I highly doubt moving dialogs is as important to them
as you think.

But perhaps I'm wrong here. If I am, I suggest you open an issue to
patternfly to support that, but as much as I like RHV, I don't think it is
special in this case, and not sure why we need this ability specifically in
RHV.

Unless that patternfly team has another reference to someone requesting
this. Doesn't sound like a priority to me.

Best regards,
Oved Ourfali
Manager, Software Engineering
RHV-M Infrastructure and UX Group

On Mar 23, 2017 12:02 AM, "Marina Kalinin" <mkalinin at redhat.com> wrote:

Hi Michael,

Thank you for your email.
So, it is on 4.2 (I tried seeing how it is on 4.1 and it is still movable).
I totally agree with you from your description, that movable dialogues are
convenient and not-movable are problematic.
Chatting with the support folks on irc, they did not like the idea either,
for the same reasons as Mike didn't like.

I am trying to think if there is another option to solve this problem and
keep the dialogues static.
This will require much more logic probably, on the dialogue itself to avoid
selection of already used items or incorrect placement. Sounds something we
do not want to deal with, if you are asking me.

Some irc comments:
- define "would not be movable"? oh, I see...that'd be frustrating because
right now we can move them around which might be helpful if you need to
check something that's already behind that pane. it'd be more of a problem
if you had to start all over because you had to cancel the dialogue box.
 - I dont like that because sometimes I need to move the windows to see the
info below them.

Cheers,
Marina.

On Tue, Mar 21, 2017 at 11:38 AM, Michael Burman <mburman at redhat.com> wrote:

> I would like to add our PMs and customer support to this discussion as
> well
>
>
> On Tue, Mar 21, 2017 at 5:20 PM, Michael Burman <mburman at redhat.com>
> wrote:
>
>>
>>
>> On Tue, Mar 21, 2017 at 5:19 PM, Michael Burman <mburman at redhat.com>
>> wrote:
>>
>>> Hello All)
>>>
>>> As i mentioned in the bz - https://bugzilla.redhat.com/sh
>>> ow_bug.cgi?id=1434048 , i will add it here as well.
>>>
>>> Some use cases:
>>>
>>> 1) When creating a new VM on a large rhv-m production environment with a
>>> lot of DCs, clusters and VMs, like we have(rhevm-3 for qe and dev).
>>> Almost every time that i need to create a new VM on this environment,
>>> i'm moving the new VM dialog bit a side, because it's hiding a reference to
>>> the desired DC/Cluster which my other VMs are running on.When i move the
>>> dialog, i know exactly on which cluster and DC to create it, without the
>>> need to cancel.
>>>
>>> 2) Create new network with vlan, but vlan is in use, i can drag the
>>> new/edit network window and to look which vlan isn't in use.
>>> This is relevant as well for which network has specific label, QoS and
>>> role. Sometimes you need the info of other networks while creating new
>>> network and you don't always would like to cancel to view this info.
>>>
>>> 3) New data domain, one created with nfs type and one with iscsi, not
>>> always recall what was created first, the iscsi? nfs? can drag the window
>>> and take a look on what i already added.
>>>
>>> 4) New cluster, you need a reference of a Cluster CPU type of a specific
>>> cluster in a large list of clusters.
>>>
>>> Note, that most of the real use cases are in large setups, with a lot of
>>> DCs, clusters, networks, hosts, data domains, VMs and so on..
>>> I'm sure that we can find more use cases if we will ask the whole rhv qe
>>> stuff.
>>>
>>> - The more i ask around the qe stuff , i understand that a lot of us
>>> using this window dragging on a daily/weekly basis. And each team with it's
>>> real need for information.
>>>
>>> - Some of the info that is required, not always could be possible on the
>>> window dialog itself and hidden behind it, but it could be helpful to
>>> complete the task.
>>>
>>> Regards,
>>>
>>> On Tue, Mar 21, 2017 at 3:26 PM, Leslie Hinson <lhinson at redhat.com>
>>> wrote:
>>>
>>>> Greg has pointed to a great resource that highlights some of the
>>>> disadvantages of a moveable dialog. Thanks for passing along! See:
>>>> http://ux.stackexchange.com/questions/81134/should-moda
>>>> l-dialogs-be-movable
>>>>
>>>> In particular, check out the section the highlights the following
>>>> concerns:
>>>>
>>>> * Moving the modal requires a lot of cognitive load.
>>>> * If users are needing to move modals, this is usually the result of
>>>> poor design.
>>>> * User moves dialog partly/mostly offscreen.
>>>> * User moves the dialog, and then resizes the browser window. The
>>>> dialog may now be offscreen, so this case needs to be worked out.
>>>> * Scrolling ambiguity with responsive layouts.
>>>>
>>>> Matt hit on this second point when he said...
>>>>
>>>> If the concern is that it's covering content on the parent screen that
>>>>> the user needs in completing the task, then it seems like a modal is either
>>>>> the wrong solution or that information should be available from within the
>>>>> dialog.
>>>>
>>>>
>>>> The intent of PatternFly is to provide common solutions that adhere to
>>>> ux best practices and standards. If an application determines that a
>>>> moveable dialog is a requirement, the best approach might be to add that
>>>> functionality in their application vs it being introduced to PatternFly
>>>> (given the design concerns and small percentage of need).
>>>>
>>>> Greg: In your use case, is important information being hidden? I'd be
>>>> curious how else the user is being impacted.
>>>>
>>>> Best,
>>>> Leslie
>>>>
>>>>
>>>>
>>>> On Mon, Mar 20, 2017 at 7:03 PM, Liz Clayton <lclayton at redhat.com>
>>>> wrote:
>>>>
>>>>> Hi,
>>>>>
>>>>> On Mon, Mar 20, 2017 at 5:37 PM, Allie Jacobs <ajacobs at redhat.com>
>>>>> wrote:
>>>>>
>>>>>> I'm not sure if there's any harm other than the widget moving around
>>>>>> unexpectedly.
>>>>>>
>>>>>
>>>>> I think there could be some harm if moving the modal around was a
>>>>> distraction that cost the user time and/or task focus.
>>>>>
>>>>>
>>>>>> But to Matt's point the use case for a modal would be one where the
>>>>>> user should not be trying to interact with the base page.
>>>>>>
>>>>>
>>>>> Also echo'ing Matt's point - placing modal dialogs front & center,
>>>>> while blocking interaction with the background, demands attention and helps
>>>>> to convey the importance of the task. Being able to push it aside
>>>>> might minimize that.
>>>>>
>>>>> Maybe the 5% use case could be better served by additional on screen
>>>>> help text or other contextual info.
>>>>>
>>>>> Liz C
>>>>>
>>>>>
>>>>>> I would agree with updating PatternFly's modal documentation and
>>>>>> possibly exploring a separate pattern for a moveable dialog.
>>>>>>
>>>>>> On Mon, Mar 20, 2017 at 4:46 PM, Mike Amburn <mamburn at redhat.com>
>>>>>> wrote:
>>>>>>
>>>>>>> I’m curious… what’s the harm in allowing modals to be moved?
>>>>>>>
>>>>>>> - allowing movement doesn’t alter the state of either the modal
>>>>>>> window or the background
>>>>>>> - no harm for 95% of the time when a user wouldn’t need to see
>>>>>>> information in the background
>>>>>>> - great benefit for the 5% that do
>>>>>>> - keeps the info displayed in the modal focused on the 95% with a
>>>>>>> workaround for the 5%
>>>>>>>
>>>>>>> No strong feelings either way, just thinking it through.
>>>>>>>
>>>>>>>
>>>>>>> On Mon, Mar 20, 2017 at 3:20 PM, Liz Clayton <lclayton at redhat.com>
>>>>>>> wrote:
>>>>>>>
>>>>>>>> Hi,
>>>>>>>>
>>>>>>>> On Mon, Mar 20, 2017 at 1:45 PM, Matt Carrano <mcarrano at redhat.com>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> It's an interesting question, Greg.  IMO, the reason for using a
>>>>>>>>> modal dialog should be to focus the user's attention on completing the task
>>>>>>>>> at hand before doing something else in the UI.  So in working from that
>>>>>>>>> premise, it begs the question of why someone needs the dialog to be
>>>>>>>>> movable.  If the concern is that it's covering content on the parent screen
>>>>>>>>> that the user needs in completing the task, then it seems like a modal is
>>>>>>>>> either the wrong solution or that information should be available from
>>>>>>>>> within the dialog.
>>>>>>>>>
>>>>>>>> +1
>>>>>>>>
>>>>>>>>>
>>>>>>>>> I'm also interested in what others think about this.  I wouldn't
>>>>>>>>> be in favor of making PatternFly modals movable unless there is a use case
>>>>>>>>> justification that we can point to.
>>>>>>>>>
>>>>>>>>
>>>>>>>> I guess if any of the dialogs warranted being modeless
>>>>>>>> <https://en.wikipedia.org/wiki/Dialog_box#Modeless> instead, then
>>>>>>>> allowing them to move would make sense. If not, then I agree that they
>>>>>>>> shouldn't need to move.
>>>>>>>>
>>>>>>>> It might be helpful to have some documentation for the "modal"
>>>>>>>> widget in Patternfly, that offered some usage best practices and clarify
>>>>>>>> terms (modal, modeless, dialog, overlay...) I found this little writeup
>>>>>>>> useful: https://uxplanet.org/5-essenti
>>>>>>>> al-ux-rules-for-dialog-design-4de258c22116#.q1fizexrl .
>>>>>>>>
>>>>>>>> Liz C.
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>> Matt
>>>>>>>>>
>>>>>>>>> On Mon, Mar 20, 2017 at 12:57 PM, Greg Sheremeta <
>>>>>>>>> gshereme at redhat.com> wrote:
>>>>>>>>>
>>>>>>>>>> Hi all,
>>>>>>>>>>
>>>>>>>>>> We recently implemented Patternfly styling on our modal dialogs,
>>>>>>>>>> and instantly received a bug [1] about them no longer being draggable.
>>>>>>>>>> According to a quick search, bootstrap's modals do not support dragging,
>>>>>>>>>> but that functionality could be added with jquery-ui [2]. According to [3],
>>>>>>>>>> it's not a great idea.
>>>>>>>>>>
>>>>>>>>>> oVirt is quite modal-dialog heavy, so I can see why users would
>>>>>>>>>> want this. We are in the planning stages of moving away from having so many
>>>>>>>>>> dialogs, so I'm not terribly worried about it. But I did think it was a
>>>>>>>>>> good idea to ask the list.
>>>>>>>>>>
>>>>>>>>>> So, what do people think about draggable modals?
>>>>>>>>>>
>>>>>>>>>> Best wishes,
>>>>>>>>>> Greg
>>>>>>>>>>
>>>>>>>>>> [1] https://bugzilla.redhat.com/show_bug.cgi?id=1434048
>>>>>>>>>> [2] http://stackoverflow.com/questions/27120579/jquery-dragg
>>>>>>>>>> able-with-bootstrap-modal-scroller-strange-behaviour
>>>>>>>>>> [3] http://ux.stackexchange.com/questions/81134/should-modal
>>>>>>>>>> -dialogs-be-movable
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> --
>>>>>>>>>> Greg Sheremeta, MBA
>>>>>>>>>> Red Hat, Inc.
>>>>>>>>>> Sr. Software Engineer
>>>>>>>>>> gshereme at redhat.com
>>>>>>>>>>
>>>>>>>>>> _______________________________________________
>>>>>>>>>> PatternFly mailing list
>>>>>>>>>> PatternFly at redhat.com
>>>>>>>>>> https://www.redhat.com/mailman/listinfo/patternfly
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> --
>>>>>>>>> Matt Carrano
>>>>>>>>> Sr. Interaction Designer
>>>>>>>>> Red Hat, Inc.
>>>>>>>>> mcarrano at redhat.com
>>>>>>>>>
>>>>>>>>> _______________________________________________
>>>>>>>>> PatternFly mailing list
>>>>>>>>> PatternFly at redhat.com
>>>>>>>>> https://www.redhat.com/mailman/listinfo/patternfly
>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> PatternFly mailing list
>>>>>>>> PatternFly at redhat.com
>>>>>>>> https://www.redhat.com/mailman/listinfo/patternfly
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> Mike Amburn Dixon
>>>>>>> Product Manager, Integrated Solutions BU
>>>>>>> M: 919-818-1201 <%28919%29%20818-1201>
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> PatternFly mailing list
>>>>>>> PatternFly at redhat.com
>>>>>>> https://www.redhat.com/mailman/listinfo/patternfly
>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>> --
>>>>>>
>>>>>> Allie Jacobs
>>>>>> UXD
>>>>>>
>>>>>> calendar
>>>>>> <https://www.google.com/calendar/b/1/embed?src=ajacobs@redhat.com&ctz=America/New_York>
>>>>>>
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> PatternFly mailing list
>>>>> PatternFly at redhat.com
>>>>> https://www.redhat.com/mailman/listinfo/patternfly
>>>>>
>>>>>
>>>>
>>>
>>>
>>> --
>>> Michael Burman
>>> RedHat Israel, RHV-M Network QE
>>>
>>> Mobile: 054-5355725
>>> IRC: mburman
>>>
>>
>>
>>
>> --
>> Michael Burman
>> RedHat Israel, RHV-M Network QE
>>
>> Mobile: 054-5355725
>> IRC: mburman
>>
>
>
>
> --
> Michael Burman
> RedHat Israel, RHV-M Network QE
>
> Mobile: 054-5355725
> IRC: mburman
>

-- 
Thanks,
Marina.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/patternfly/attachments/20170323/eb20c92b/attachment.htm>