Task Centric Interface

[Thomas Pfeiffer]

On Monday 04 February 2013 15:54:39 Wolfgang Romey wrote:
> Hello
> 
> I follow the discussion about the task-centric-interface with high interest
> and think, i will be important for PA to make real progress in that field.
> Together with activities PA  could become something very special.

Glad you agree with us here :)

> But as it will take a while, to get there, i think, you should give the user
> ways to work i the old way, as long as the new way is not there. An
> example: as long as tagging is not really working in files and existing
> tags in media files are not read, the possibility to browse folders in
> Plasma-Mediacenter should be back. Without it, one can't really test and
> use it.

Well... sorry, but we won't do that. If you choose a paradigm shift as big as 
the one from a folder hierarchy to semantic information only, you have to go 
all the way.
If we reactivate file system browsing in any application because it doesn't 
fully support Nepomuk, users will start organizing their files in a folder 
hierarchy again, something which other applications in Plasma Active don't 
support. Thus the whole concept falls flat on its face.
Plus, if we reactivated the file system browsing, you'd have a workaround for 
any Nepomuk-related problems and might simply forget about them, thus 
depriving us of your feedback. So as mean as it may sound: You as an early 
adopter will have to suffer a bit for the benefit of all the users who can enjoy 
a polished product in the end ;)

And specifically: What do you mean by "tagging is not really working in files"? 
Could you more precisely tell us what doesn't work? If there are bugs, please 
report them in Bugzilla. If there are features missing or usability problems 
present, tell us either on Bugzilla or via email.
That tags from media files are not read by PMC is a big issue, because a media 
center _has_ to work with metadata, not a folder hierarchy. Bangarang has 
shown how that works, it's just lacking a more touch-friendly UI.
If PMC isn't usable at the moment with Nepomuk data, that has to be fixed, 
instead of just reactivating the workaround aka the file system browser.

> And you should give the users a choice of more software. When you offer
> Marbel in a version which is not optimized for a touch screen, why don't
> you offer for example kstars? I know from the slx-image, that it is usable
> in PA; or for example some educational software? Or is there a way to
> install from another repro?

To be honest: Though I'm not the one doing the packaging, I am pretty sure 
that the non-availability of applications is simply a matter of prioritization 
of scarce resources. Marco Martin seems to be almost running a one-man show at 
the moment, at the same time fixing bugs, polishing features _and_ doing the 
whole user-facing part of a distribution all by himself.
That's why he doesn't - and in fact should not - invest his time into 
packaging applications that aren't touch-optimized anyway.
The fact that Marble packages exist is mainly due to the fact that Basyskom is 
involved in Marble development and since they wanted to demo Marble in PA, 
they packaged it while they were still actively involved in PA development.
If you know somebody interested in packaging other desktop applications - 
which work at least okay on a touch device - for PA, though, I'm pretty sure 
that Marco won't object to it :)
 
> I think it is right, that PA should not compete in number of applications,
> but i think, that it could (should) have a more complete offer.

Yes, I agree to that. But since currently there are not so many projects 
working on touch-friendly versions of their applications, there is not so much 
we can do at the moment to fix that. And we do not want to create bad user 
experiences by offering applications in PA which just don't work well enough on 
a tablet.

> I use PA-devel on a weTab and make updates nearly every day. Up to now,
> everything is working well and i see progress. My thanks for that.

Great! Your feedback is very valuable to us. We need early adopters like you 
who are willing to try out new things and report back to us what they like 
about it and where they still see problems. And unless your problem is somehow 
unique to your specific case, we will try to find a solution to it, which may or 
may not be the one you suggest.

So thanks again, and keep the feedback coming!