KDE is not an OS platform... (And neither is Gnome)

[nf2]

On Sun, Nov 1, 2009 at 3:27 PM, Dario Freddi <drf54321 at gmail.com> wrote:
> On Sunday 01 November 2009 15:08:49 nf2 wrote:
>> I agree. GUI applications should always use async APIs like KIO or
>> GIO. But: At the moment many of them won't, because that implies
>> deciding for a certain desktop environment. That's the problem i would
>> like to solve.
>
> As we discussed on IRC some time ago, this is where a shared interface should
> kick in and be implemented on both sides.
>
> Then you can provide it in various forms (DBUS API as the main entry point,
> and eventually qt/gobject wrapper libraries to ease the dbus handling) to any
> 3rd party application (such as firefox), and write bridges from/to KIO to/from
> GIO (as you are currently doing right now, if I understood correctly) to allow
> native KDE/GNOME applications to appear native on the other side as well.
>

Hmm... I gathered a list of solutions to "the problem" (which perhaps
isn't that important):

1) Putting a different label on an already existing system, like I
would suggest: "This ist the standard system for the major
file-management protocols, thanks, problem solved". Technically easy,
but won't work because there are understandably different opinions on
which one to pick.

2) Completely aligning the behavior of protocol-handlers in KIO and
GVFS in every detail. Sounds like lot's of work which realisticly is
never gonna happen, because that's a very non exciting task. It
completely opposes the way developers feel motivated. Also has the
disadvantage that you can't move forward or backward, change behavior,
introduce new protocols... Plus this will never be really charming,
because all the little differences in "the systems" remain (Like GVFS
mounts appearing in GTK Filecoosers, but not in KDE dialogs...).

3) Your proposal: Waiting for "the day the cows come home" that people
come together desiging a shared interface. Technically hard + years of
work.

4) Wait for the kernel to provide all that *and* wait for applications
to use different kernel APIs.

5) Do nothing and continue dreaming about one desktop winning the
competition (which would also define a platform for 3rd party apps,
but in a rather cruel way).

>
>>
>> The free desktop desperately needs those 3rd party applications, and
>> it's definitely not in the market position to tell application
>> developers: "Then you just write your application twice! One for KDE
>> and one for Gnome".
>
> Let me remind you that, even if GNOME and KDE have a huge portion of the
> market, there are some other DE out there, that might (one day, if not
> already) have their own IO subsystem.

Oh no! Hopefully they won't. And if they are GTK based, they will
probably use GVFS.

Norbert