On 1/2/06, Kent Sandvik <[EMAIL PROTECTED]> wrote:
> On 1/2/06, Mike Emmel <[EMAIL PROTECTED]> wrote:
>
> > Saving 2-3 megs is good but with a default 56 meg footprint
> > it's like a few percent of the total. And as I said before I can't see
> > LiTe not growing to close to a 1 meg if it fully supports mozilla this
> > may be in mozilla specific widgets but still the total widget support
> > will probably approach 1 meg.  So your really looking at a 1-2 meg
> > footprint difference between what I call compelete LiTE and a stripped
> > GTK. The cause of this is  is really just the unremoved bloat plus all
> > the gobject cruft.
>
> LiTE will never become 1-2 Mb as long as I live :-).  Anyway, any
> additional widgets and support will be placed outside LiTE itself.
> LiTE is just a toolkit enabler, any specific code needed for ports
> should be moved to the abstraction level on top of LiTE and below the
> target to be ported.
>
>
My point is your total support code would approach 1 MB if not more
getting close to the same size as the minimal gtk support code.
Gtk is not that poorly written :)

> > If your serious about firefox/mozilla in a small environment your far
> > better off keeping gtk around for a while until you remove the huge
> > amount of bloat thats in mozilla then consider going to a smaller
> > widget library.
>
> Just curious, how did the BeOS and Photon ports then work out?
>
Dunno for sure I've not used them just looking there about the same
size most of the ports are. Mozilla has plenty of abstractions layers
the ports are fairly similar. I've looked at the code of all the
ports.

> > Junk code percentages for my analysis
> > Full gtk has about a 30% junk code level
> > Browser targeted gtk about 10%
> > Intrisic over architected hard to remove glib/gobject junk 20%
> > WebKit  10-30% junk from keeping the gt binding plush some over 
> > architecting.
> > Mozilla  50-70% junk code  from xpcom cruft massive over architecting
> > extra api's etc etc.
> Thats's my worry about Mozilla, the xpcom support...
>
They are working to reduce remove xpcom but its slow work.
Right now there is no huge drive to support mozilla  on embedded
platforms a good quick directfb/gtk port might actually be the key to
encourage work to reduce mozilla.  I'm sure there is a good chance it
would be used in some commercial offerings that use directfb.

>
> > I think the fact that minimo is 12 megs and WebKit 6 megs supports my
> > junk code analysis of Mozilla.
>
> Minimo's RSS needs are also on the range of 50Mb or more, requiring an
> embedded system to have most likely 128Mb of RAM, of which most is
> dedicated just to run one instance of Minimo...
>
> > Mozilla ranks second only to open office as the junk code king.
>
> Yep, I agree. WebKit is not yet that messy... But alas it seems the
> industry really likes the Mozilla/Firefox story.
>
Agian I'm not agianst using/porting it to directfb. There are
platforms with that much ram that want to use directfb over X.
DirectFB can stand on its own merits :)
I'm just saying that for Mozilla removing gtk does not help much and
your better spending time cleaning up mozilla itself.
Nokia chose WebKit for series 60 btw so its the only embbeded full
featured open source browser out ther ( Excluding Opera/IE ).
Mozilla has never made it too any reasonbly small device.
WebKit is also getting pretty standard since the core is part of
Safari I'm pretty confident its viable.  I'd love to see both browser
avialable with a reasonable footprint.

> > And just to finish this long diatribe today I think its better to
> > invest the time effort and energy into doing a good implementation
> > based on the following stack
> >
> > llvm->directfb->cairo->javascript->svg->xul->html
>
> I'm still worried about cairo itself. Do they yet have a version with
> zero floating point support needed? svg is another big huge system,
> too. Also, the more code, the less chance to have decent support of
> the tiny L1 and L2 caches in embedded systems, resulting in a lot of
> work going over a (usually) slow data bus.
>
First Cairo is fixed point with a floating point api.
I know of no floating point ops off the top of my head used inside
cairo. Certainly not in any main pathways. SVG is not that huge :)
My point is that its obvious today that the browsers are going to
include vector support right now its flash uuugh hopefully in the
future it will move more to SVG. I think it will esp with AJAX.
Unless you want to be in the postion of paying for Flash support
forever your better off getting good SVG support in asap.
I think the fact that SVG support can be embedded deeply into the
browsers since it open while Flash is a plugin will allow it to beat
flash in the not to distant future.


> Maybe five years from now when NAND is cheap and most systems ship
> with 1Gb RAM all this is a moot discussion. Then again embedded
> systems seems to always be lean, of some interesting COGS price
> reason.
>
Yeah and the fact that hardware guys rarely even ask software guys
what they would like. I.e a 32 bit bus enough vram for double
buffering.  I think though as the platforms get more powerful there
will be some work to support better graphics.
With that said I've successfully
implemented java2d on constrained devices ( cell phones) so I know
most of the higher end feature phones are up to the task. 32mb
ram/100mhz plus cpu.
Btw freetype is almost a full blown 2d engine under the covers for
truetype fonts.

All I can say is trust me on this one :)


> Anyway, would be worth trying out *at least something*, and see what happens.
>
Yep I agree and I think we should support Mozilla on directfb I just
don't think that jumping right into a lite port right away is the
right answer. Get it out there on directfb/gtk if the mozilla
community/directfb developers can get mozilla cut down to a sane size
then at that point LiTE makes sense. Buy concentrating on getting the
directfb/gtk port into the hands of people who can and are willing to
work on the mozilla fat you can then move to LiTE.
I just don't think directfb has enough developers to support multiple
ports of Mozilla and the LiTE will take quite a lot longer to get out
the door.


Mike


> --Kent
>
> _______________________________________________
> directfb-dev mailing list
> [email protected]
> http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
>

_______________________________________________
directfb-dev mailing list
[email protected]
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev

Reply via email to