On Mon, Oct 07, 2002, Bill Janssen wrote:
> Sure, I can put that in.  Though I'd rather do it with a new image
> record,

Well, the idea was to *not* access the image record, i.e. putting
the data in a new record will not help (it will help when supporting
large images, though) ;-)

I'm talking about inlined images, not the (optional) large version
of an inlined image. The "problem" with inlined images is that there
can be a lot of them and if the viewer must find, open, uncompress,
and get the image dimensions for every image both when "creating"
the layout and also when rendering the images it will slow down the
viewer.

> which could also handle an image composed of multiple parts.

That's a completely different issue.

> May I suggest that we make it function code 0x1F, and the parameters
> ID, width, height, and alternate-text-length?

Then we have a problem, since I would like to the use the last byte
for the bit depth ;-) 

Still, maybe we could use the last (data) byte to indicate the "extra"
data bytes that can be found after the first batch of data, i.e. today
we are limited to use at most 7 bytes of data,

<function><data 1><data 2>...<data 7>,

If we use the last byte to indicate the number of "extended" data 
items, we could use much more data,

<function><data 1><data 2>...<data 6><# extended data items>
<extended data1><extended data2>...<extended data N>

For example, the new image function code would look like this,

<0x1F><image record id><width><height><0x2><bit depth><alt len (N)>
<alt text char1>...<alt text char N>, where N <= 255

That would make it possible to use an "unlimited" amount of data...

> That is, we add a byte that gives a number of additional bytes
> (0-255) which, if present, contain the "alternate text" for the
> image tag. 

I think it is a great idea to include the text the viewer should
use when not displaying the image instead of just showing [img].

/Mike

_______________________________________________
plucker-dev mailing list
[EMAIL PROTECTED]
http://lists.rubberchicken.org/mailman/listinfo/plucker-dev

Reply via email to