[Owncloud] About Bootstrap and ownCloud

Wed Mar 20 15:05:35 UTC 2013

Klaas, Danimo, I think your points are good in general. Please keep in mind
this discussion is about Bootstrap specifically though. Bootstrap might be
the go-to library for website coders who don’t want to invest much into
caring about design. But it’s just some predefined styles, it’s far from
»the web design framework«. We would need to customize and remove so much
that I fail to see the value here.

Klaas, sorry that there’s no text to »read and learn about the "design
direction we're going for" so that we're able to pick the best framework
and design our new route with it?« We’re doing it as we go, and you can
read about it in all the issues and the apps. Be sure that I talked about
it very much with the app developers, much like we did for the desktop
client, writing up the spec and doing mockups and discussions. I have
limited time unfortunately, so if not everything is documented in one place
please excuse that.

What I know is that the design we’re going for – look at the News app or
the Notes app – is so simple that we do not need Bootstrap. We have what –
a header, an app navigation bar and for the app a left navigation bar and a
content area. That’s it.

Stefan:

On Wed, Mar 20, 2013 at 3:31 PM, Stefan Nagtegaal <
development at standoutdesign.nl> wrote:

> Besides the very good point Klaas comes up with here, is that Twitter
> Bootstrap has proven it's value for many, many big projects and websites.
>

Yes, for many _websites_. Bootstrap is primarily for websites and less for
apps. That’s a difference. I used it several times for websites as well as
apps and it indeed proved helpful for websites. For the apps not so much
because things need to be streamlined.

> If you want to have a consistent UI, with usable and proven UI elements
> for navigation, dropdowns, balloons, or whatever this is the way to go imo
> or at least base our core/css/style.css-file upon it.
>

First off, lots of the UI elements in Bootstrap we don’t even need, and
basically all of the others we would need to improve or change looks of
anyway if we want to use them. We don’t need the grid first of all. We
don’t need any balloons, for dropdowns we use jquery.multiselect or Chosen
already, and the left side navigation styles we also already did, see
https://github.com/owncloud/core/pull/2275
We have buttons and button groups and breadcrumbs done already. We don’t
need any pagination, media object, wells, tabs, pills and so much other
stuff. For the Javascript stuff, we don’t need Scrollspy, tabs or popovers.
Our tooltips are handled by tipsy already (which is also used by
Bootstrap), we have alert styles already, we don’t need Accordions nor
Carousels.

What’s left? What we have right now.

Your point here is crap, and doesn't bring any value to the discussion:
> Funfact: No usability test participant ever said »strange, this website
> doesn’t use Bootstrap«. Because people don’t care. Let’s work on real
> issues."
>

Yeah, please don’t attack my arguments just based on one »fun fact« which I
inserted just for laughs.
The point in here was that our most pressing interface issues definitely do
not include the need to use a hyped library which will really not help us
that much.

> People DO care about a UI that is consistent, user-friendly and looks
> good. Bootstrap is bringing that faster to ownCloud than scratching your
> own itch, because a lot of (usability) test participants are already
> familiar with the UI-elements (because of they are used by big companies
> like Twitter, Google, etch etc).
>

That’s simply not true. People are not familiar with the UI elements.
Everyone who uses Bootstrap customizes it. Facebook, Google, Twitter and
Apple all use vastly different elements. None of these big
projects/companies use Bootstrap, because it would hold them back with
strange predefined styles which wouldn’t be used anyway because they have
better designers. I don’t see how Bootstrap will help us here.
Also it’s not about »test participants« but about people, about everyone
who will use the software.
I also don’t get what you mean by »scratching your own itch« in this case.
We constantly listen to feedback and do testing, it’s far from only doing
ownCloud for ourselves. I got feedback by someone who runs ownCloud in
primary schools in France and they love it (they even switched over from
another software) because it’s so easy to use.

Afterall, I'm not saying Bootstrap brings us everything we ever wanted. But
> I AM saying that Bootstrap is a pretty solid base to work from, to bring
> ownCloud a consistant user interface and fix a gap between coders and
> designers.
>

Seeing how we wouldn’t use much of it, and of the remaining stuff heavily
customize anyway, I fail to see that. Maybe I’m wrong.

If you like, you can rework the current interface (ideally also the main
apps) to use Bootstrap, but make sure it looks and works similar to how it
does now. We have reasons for designing it this way and not just using
Bootstrap and its default styles. Then please submit a pull request and
we’ll see if it’s better. If it’s better we can use it.

My main point here is that not everything is magically improved by using a
library like Bootstrap, and it’s not even a good base. Look for instance at
our installation process. It’s so vastly simpler than any other
installation process, and it would look and work way more complicated if we
used Bootstrap. If we want to do proper usable design, we need to invest
more.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/owncloud/attachments/20130320/56d73064/attachment.html>