>
"
list of configurations grows faster than the number of facets"
> I would
put this down to abuse of configurations than anything else.
>
Personally to me, it is much better anyday to select a single item in
>
configurations than having to go and select individual
The problem is that “abuse” is in the eye of the
beholder. What I have seen is that it is natural for the facet creator to view
their facet as having extra importance and therefore deserving of broader exposure
via a preset. As there is no impartial central arbitrator of what constitutes a
“deserving” preset nor could you define a criteria for such central
arbitration, the list grows quickly due to the cross product effect. As people’s
usecases and perceptions differ widely, the argument over the merits of a
particular preset is not one that anyone can ever hope win.
If you need an example of this effect, take a look at the
XDoclet facet and it associated presets.
>
Personally to me, it is much better anyday to select a single item in
>
configurations than having to go and select individual facets.
That is true in principle, but not in reality. Once you have a
large list of presets named like “x with y and z”, it takes more
time to scan the list to find the one that you are looking for than it would to
select the facets directly. The problem is even worse for novice users. The
present names may not be obvious. The descriptions often don’t help much.
At the same time, the concrete definition of what you are selecting (the actual
facets) is hidden from view. In my observation of both novice and experienced
users, I have seen presets as being more of a hindrance than help. The only
people who I have seen thrive on presets are tooling QA engineers who have to create
the same configurations over and over again.
>
I just hope we address the right problem.
Indeed.
- Konstantin
From:
wtp-dev-bounces@xxxxxxxxxxx [mailto:wtp-dev-bounces@xxxxxxxxxxx] On Behalf
Of Jaspreet Singh
Sent: Wednesday, December 23, 2009 10:25 PM
To: General discussion of project-wide or architectural issues.
Cc: wtp-dev@xxxxxxxxxxx; wtp-dev-bounces@xxxxxxxxxxx
Subject: Re: [wtp-dev] Improving discoverability of facets in WTP
wizards
Hi,
" list of
configurations grows faster than the number of facets"
I would put
this down to abuse of configurations than anything else. Personally to me, it
is much better anyday to select a single item in configurations than having to
go and select individual facets. I just hope we address the right problem.
We might as
well just rename the existing "Modify" button to either "Modify
Technologies..." or just "Technologies...".
"Konstantin
Komissarchik" <konstantin.komissarchik@xxxxxxxxxx>
Sent by:
wtp-dev-bounces@xxxxxxxxxxx
12/24/2009
02:53 AM
Please respond to
"General discussion of project-wide or architectural issues."
<wtp-dev@xxxxxxxxxxx>
|
|
To
|
<wtp-dev@xxxxxxxxxxx>
|
cc
|
|
Subject
|
[wtp-dev]
Improving discoverability of facets in WTP wizards
|
|
I’d
like to solicit feedback about some proposed changes. The changes are tracked
by the following bug, but I am also repeating them in this e-mail for
convenience.
https://bugs.eclipse.org/bugs/show_bug.cgi?id=298483
Facet
selection page used to be the second page in all WTP project creation wizards.
A few releases ago it was removed out of the wizard flow and placed into a
separate dialog accessible via the "Edit" button next to the
"Configurations" combo box on the first page of the wizard.
Since then,
we have started hearing numerous feedback from novice users that indicates that
we have hidden facets too much. The most recent feedback was a usability study
conducted by a team at Oracle where users who have never used WTP were tasked
with writing a basic application. One of the shocking discoveries is that they
couldn't figure out how to enable JPA. These weren't novice programmers either
(the team was composed of devs from our WebLogic Server group), but they were
new to our tooling.
Time is right
to take another look at this problem. I've had a discussion with a UI designer
at Oracle about this and together we have came up with a mockup (attached) that
proposes the following changes that should in principle make facets more
discoverable without doing anything too drastic:
1. Create a
"Technologies" group that merges the module spec version frame and
the facet configurations frame.
2. Eliminate
the facet configurations drop-down. Configurations (aka presets) is one of
those ideas that I largely consider a failure. They sound good in principle,
but do not actually improve usability as the list of configurations grows
faster than the number of facets due to the cross-product effect. Once the list
is long, picking through it is next to impossible, especially for novice users
who this feature is trying to help in the first place. This change doesn't
propose that configurations are eliminated completely (although that's something
that we should seriously consider in the future). They would still be available
in the facet selection panel.
3. Convert
the edit button into a more obvious "Choose other technologies to use in
the project..." hyperlink.
- Konstantin_______________________________________________
wtp-dev mailing list
wtp-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/wtp-dev