systray specification...

Carsten Haitzler (The Rasterman) raster at rasterman.com
Fri Feb 3 08:03:20 EET 2006


On Thu, 02 Feb 2006 23:27:08 -0500 Matthias Clasen <mclasen at redhat.com> babbled:

> On Fri, 2006-02-03 at 12:38 +0900, Carsten Haitzler wrote:
> 
> > 
> > true. the other suggestion here is to use xrender pictures. it is a point.
> > did think about it, but then decidced against for 2 reasons.
> > 
> > 1. it allows the spec to be implemented on ANY XSERVER - even that ancient
> > labtam monochrome 1bit terminal sitting on a 10mbit coax network. it will
> > work.
> 
> I don't think RGBA tray icons are very relevant on that kind of
> hardware, so I don't think this is a valid argument against using
> Render.

why not? the alpha channel can just be thresholded at 128 and a mask generated.
not to mention as an example - e17 runs on that hw just fine.. with alpha
channeled compositied icons and graphics all over. alpha blendind did exist
well before xrender - and doing it in software client-side is still by far the
fastest way to do it in general - modern xrender extension or not. so i think
it still stands to reason that providing the information that way is valid. HOw
you chose to display such data is still an open descision on the part of the
toolkit/tray manager etc.

> [...]
> 
> > but do we want a spec like systray to RELY on currently "new" 
> > extensions
> > that are in fact still mostly experimental even as of today (xrender
> > acceleration is iffy - on a god day still).? i would dearly like such
> > a spec to
> > work even on old xservers embedded into x terminal thin clients in
> labs or on
> > old sun workstations etc. ?
> 
> I dont see the quality of current Render implementations have much
> relevance to the question of using Render Pictures for holding image
> data either.

specifying a netwm specification thet REQUIRES an extension that may or may not
be supported and if it is supported still is od dubious quality i think is bad.
if the specification does not rely on an extension and leaves that up to a
descision of the toolkit used for display - then you haven't created any extra
problems.

> Matthias
> 
> 
> Matthias
> 


-- 
------------- Codito, ergo sum - "I code, therefore I am" --------------
The Rasterman (Carsten Haitzler)    raster at rasterman.com
裸好多
Tokyo, Japan (東京 日本)



More information about the xdg mailing list