Graeme Geldenhuys ha scritto:
On 29/01/2008, Lee Jenkins <[EMAIL PROTECTED]> wrote:
Will fpGUI/LCL still support theming later when theming is implemented?
To be honest, I'm not sure how it's going to work. I still need to do
a lot of work on theming support in fpGUI and I don't know what the
other LCL widgetsets do in such a case? I would guess fpGUI would
loose some of it's features when used as LCL-fpGUI widgetset - or
maybe I just don't understand the LCL backend fully. Could others
comment here?
I wouldn't be much in favor of an LCL-fpGUI widgetset, considered the
amount of work it requires.
LCL is aimed to provide native support for ready made widgeset, which
were not originally intended to be handled that way.
This, besides being a core developers decision, fulfills a requirement
for a number of applications, but also creates a number of constraints.
It also makes dealing with LCL backend much more difficult than strictly
necessary, because LCL interfaces with different high level functions
each one conceived it's way, and is forced to includes workarounds for
different philosophies of widgeset designers, widgetset whims, and
widgeset bugs.
If the aim isn't that of supporting existing widgesets, but to have a
Lazarus native widgetset, I'd see more efficient, both in terms of
development time, and in terms of final result, to think of fpGUI as an
*alternative* to LCL.
IOW:
a) You want/need GTK, WIN, Qt, Carbon, etc. widgetsets providing
platform dependent native look'n feel? Then use LCL.
b) You want/need native fpc widgetset providing identical look and
behavior on different platforms? Then use fpGUI, fpCL, or whatever you
want to call it, *in place of LCL*.
It would be a higher level choice: LCL or fpGUI. Then next choice: in
case of LCL the widgeset choice, in case of fpGUI the graphic interface
choice (nobody forbids to use e.g. X in Windows or Mac environment. A
bit harder to use GDI in Linux environment ;-) )
I've started my own experiment to verify if this road is viable, by
attempting a port of Delphi clx to Lazarus. I.e replacing LCL with
something else. The first results are quite promising. Just some
compatibility problems between Delphi and fpc implementation, which
demand many days to be located, and a couple of hours to be fixed.
In my test case I'm using something which is very roughly LCL
compatible, while the native approach could not only be much more LCL
compatible, but also profit of a great amount of LCL code, without being
hampered by the constraints LCL developers must live with.
This would also make the fpGUI development line more integrated with
Lazarus development line, so that each branch could profit from whatever
comes from the other branch.
Just my 2 cents.
Giuliano
--
Giuliano Colla
Whenever people agree with me, I always feel I must be wrong (O. Wilde)
_________________________________________________________________
To unsubscribe: mail [EMAIL PROTECTED] with
"unsubscribe" as the Subject
archives at http://www.lazarus.freepascal.org/mailarchives
