8 Mar
2011
8 Mar
'11
8:53 p.m.
My proposal about using view mode is not to be in "edit" mode automatically.
My proposal is to have the "edit" button maybe more visible as well as
"Add Widget", and more importantly to have the "edit" button not load
the "inline" mode, but load in place.
Most importantly the "save" buttons don't mean anything as adding a
gadget adds an XWiki object and is permanent. Same with moving and
removing gadgets. That changes the object right away.
So we need to "remove" the "save" actions as they won't mean
anything.
Ludovic
Le 08/03/11 18:07, Sergiu Dumitriu a écrit :
On 03/08/2011 05:57 PM, Marta Girdea wrote:
--
Ludovic Dubost
Blog: http://blog.ludovic.org/
XWiki: http://www.xwiki.com
Skype: ldubost GTalk: ldubost
On Tue, Mar 8, 2011 at 4:01 PM, Luca
Anca<lucaa(a)xwiki.com> wrote:
Same for me. Users tend to just move things around, accidentally click
the mouse, and then "things just broke by themselves".
On Sat, 2011-03-05 at 12:44 +0100, Ludovic Dubost
wrote:
-1 for A)
I'm definitely for B) "Edit mode edit", not only for consistency, but
also for avoiding surprise side effects when interacting with the
dashboard. Hi,
I've tried the dashbord on 3.0 snapshot and could not make the "Add
Gadget" function work on an empty dashboard.
From what I see the issue seems to be in the wysiwyg code. I got the
no,
it's dashboards code, I forgot to handle the case when the column
where the gadget should be added doesn't actually exist.
dialog box from the wysiwyg, when clicking on
Insert Gadget" I got an
exception:
uncaught exception: Class$dF: One or more exceptions caught, see full
set in UmbrellaException#getCauses
at
function nr(){try{null.a()}catch(b){return b}}
This issue seemed to only happen with an empty dashboard macro (with not
gadget yet added).
I think we need to work on two things for each feature that we are
developping:
1/ Having a status of where we are:
In the case of Dashboard, I think that's updating this page:
http://dev.xwiki.org/xwiki/bin/view/Design/Gadget+Scenarios+Prioritized
I updated
the table. There could be new cases, I guess, to comprise the
missing things (like adding a column when editing, or others a like),
but I prefer to use Jira issues for these. WDYT?
2/ Write a test plan and have Sorin include the
testing of new features
in his testing.
For that he needs to know how to test and what is supposed to be
there/not there.
Concerning the development priorities, I think for the next period of
development on the dashboard we need to focus on wiring the UI for
"beginners", making it easy for them to create their first dashboard. In
my view this is in priority:
- Create page with a dashboard macro (In the Add Menu in XE)
- Go to edit more
What's "edit more"?
(I'm still convinced that we need the
dashboard
editing in "view" mode and not in "inline" mode)
What do
others think?
Knowing that the current edit in inline mode has a few limitations:
* remove gadget will be saved immediately, not when the user clicks
"save". If the user says remove gadget, confirms the remove and then
clicks the "cancel" button in the edit form, the remove will still be
saved.
* same for add gadget.
The way I see the two possibilities:
A) "View mode edit" will save on every change
B) "Edit mode edit" (implemented as inline edit, now) will save
everything at the end (current implementation but with the limitations
fixed).
So, WDYT? >> - Add a column in your dashboard
>> - Editing a dashboard
> you mean editing a gadget? as in the parameters of that gadget?
>
> Also, I would add to this list, potentially at the top:
> - finalizing the structure of the dashboard / gadgets objects, so that
> we don't have migration in the next releases (e.g. if we decide to add a
> field in the gadget object to identify which is the dashboard of that
> gadget, in case there are multiple dashboards on one page, we should do
> it now).
>
> Thanks for the feedback,
> Anca
>
>> Also for the 3.1 timeframe I think we need to have some brainstorm to
>> add some more "Dashboard" friendly macros/gadgets to make dashboards
>> more usefull and interesting for users, but that's another story.
>> Right now I think the most important is that we agree on the priorities
>> for the next step and that for 3.0 we have a simple but coherent
>> dashboard feature.
>>
>> Ludovic