Re: Transparancy

[Christopher J. Osborne]

|   One thing Netscape information providors can do is to make the
|   transparent background colour of their gifs light grey (ie RGB
|   value of 195,195,195). I know this makes it a little more
|   difficult to make sure the gif is actually transparent, but as
|   most other browser use this as a back ground colour, you will
|   greatly increse the chances of you transparent gifs still looking
|   good in other browsers (and I don't just mean Mosaic.

I do this myself, as a remnant from when the current Mosaic version
didn't support transparency.

Unfortunately, most transparent gifs assume that the background is
some shade of grey. You see this most clearly in the edges, where the
black fades to gray. If you for some reason choose, e.g., a dark blue
background, there will be a light halo surrounding the graphic. For
this reason, one should be critical of what graphics one makes
transparent.

Fortunately, the upcoming PNG format will have gradual transparancy on
any colour, and even a full alpha channel if needed.

[Phillip Hallam-Baker]

|   What would be nice would be to have a pixmap type affair such that
|   the drawing was stated as being in the foreground and backround
|   colours of the screen. Thus the text colour would match the
|   picture text colour.
|   
|   Something for the new Gif standard perhaps??

Yes, this would be nice. There already is a "bKGD" which specifies a
suitable background colour for the viewer, if doesn't have one
already.

fRGD  Foreground color
     This chunk specifies that the given color in the image can be
     presented as the overall foreground color in the viewer. [Adapt
     the text for the bKGD chunk for the details.]

We don't need the background corresponding information, since
transparancy handles that nicely. An alternative is to use greyscale
images and let the browser change what black and white means, however
fans of white lettering on black might not like that :-)

Does anyone know the address to the PNG mailing list?


Kjetil T.

Received on Tuesday, 4 April 1995 11:41:10 UTC