[PATCH] Multi-Protocol IO-Slave

nf2 nf2 at scheinwelt.at
Wed Jan 16 17:46:07 GMT 2008

Michael Pyne wrote:
> On Sunday 13 January 2008, nf2 wrote:
>> Also, having a Qt based client library is less liberal in terms of
>> licenses. 
> Well the Qt based client library (including everything that goes with using 
> Qt) is a choice we made.  It seems that just as soon as GNOME got 
> commercialized it became a rallying cry that people can write proprietary 
> apps using GTK+.  Which seems sad given the support provided to GNOME by 
> Stallman and the FSF.
>> Of course, rewriting the core of KIO in plain C (or better
>> GLib + main-loop + C) should have happened many years ago.
> Why should the core of KIO be rewritten in plain C again?  You have convincing 
> arguments about the structure and advantages of the GNOME VFS and then go on 
> talking about how devolving KIO to a C library would somehow be an 
> improvement?

I regret that i made those controversial statements. I just tried to
find reasons why KIO has never been linked outside KDE (never bekame
cross desktop), which from an historical point of view is sad (There
would have been lots of demand for that, as Gnome-VFS never worked that
well). Now, however, i believe the situation is different.

> What I'm interested in is what Kevin Krammer was mentioning?  Is it possible 
> to provide a line-interface compatible KIO-like interface to the GIO mount 
> daemons (like how DBUS has nice language bindings)?  i.e. using the same KIO 
> classes but with GIO job class implementations.

I hope my reply-mail to Kevin answered this question. At least partially.


More information about the kde-core-devel mailing list