On Sun, Apr 21, 2002 at 10:33:52AM +0200, Edwin Bos wrote:
> On Sat, 2002-04-20 at 22:30, Adam McDaniel wrote:
> 
> > I'm probably about 50% finished the code on the viewer side, but I have
> > no idea on how to make the parser smart enough to handle this. Hopefully
> > that'll come with time :)
> 
> Can't you just check the filesize, and if it's larger than 64k, convert
> it again with a lower bitdepth or size?

That would be on the parser side of things. I'm working on the viewer's
side. The idea is that when the parser is working and seems an image
over 64k, it will basically chop the data in half (or more) to fit
within hotsync protocols. At the end of each chop (or perhaps by using
an image index record) there will be a redirect to the next chunkett
of data.

The viewer sees this and re-builds the image accordingly. Unfortunatly, I
found out that this technique doesn't work with OS 3.5 as I said before,
but a minimum of 4.0 is required. Not a horrible issue, but downward
compatibility will be tricky.

I am still trying to figure out how this'll be handled on the viewer,
once thats done and stable.. I guess I'd have to learn python.

-- 
Adam McDaniel
Array Networks
Calgary, AB, Canada

Reply via email to