Quis custodiet ipsos custodies?
Solid State Builders Group
Huh? Is Zoof handing out free Zingers? I used to love those things when I was in High School!
Adam
DUDE Zingers are for queers, real men eat Zebra Cakes
Quis custodiet ipsos custodies?
Solid State Builders Group
The company that stocked the snack machine at the school didn't have a Little Debbie franchise license. But they did have both Dolly Madison and Hostes. So we had Zingers and Twinkies, but no Zebra Cakes...
Adam
Please realize that it NEVER was "my version".
I think there is a very basic idea here:
If an international standard is first published one way and then changed to something different, that change should have only been done if there was an absolutely necessary reason, like it doesn't work ... now it does.
Since you (carmangary) have now worked with ALL version of the ILDA format 0, 1, 2, 3, busted 3, 4 & 5, PLEASE explain ANY benefits you see in busted 3.
That is ALL I have ever asked of ANYONE !!!!
WHY?
You're absolutely right about the fact that something like this can totally mess up other people's software. At the time I added 24-bit color to LaserBoy there was NO OTHER SOFTWARE that did it.
James.
I agree that the way you have the header makes more sense. But, I don't think it makes the format 3 data any more or less good. My objection to format 3 is that it requires a seperate frame of colors. Format color tables make sense because they apply to many frames. You can do the same with Format 3 but it wouldn't make sense. I think format 4 & 5 are superior. I understand your argument you made awhile back about why Format 3 is good but it isn't compelling enough to make me like it.
Well, thanks for at least agreeing with me in part.
You are right about the difference between a palette (2) and a color table (3) in that a palette has a one-to-many relationship with frames that follow and a color table has ONLY a one-to-one. But other than that, reading them from a file is equally complicated. In both cases you have to store all of the colors before you get the vectors.
I have already suggested another variation of the color table that WOULD have a one-to-many relationship with frames that follow. This would be IDEAL for 24-bit color math generated abstract animations that have a large set of frames that all have exactly the same number of vertices.
At least you didn't say "Because ILDA said so."
Thank God for that.
Also, 4 & 5 are NOT extensions to the original 0, 1, & 2. They REPLACE them entirely. That is a totally new file format.
James.
Last edited by James Lehman; 01-12-2009 at 11:42.
xperts.[/SIZE]
BTW, this is the same group that now wants to be in control of your credentials as a laserist, whether you're a member or not.
No, its not, the committee has changed in membership.
Steve
4 & 5 aren't extensions in the way that 2 & 3 are but I wouldn't go so far as to call them a different file format. It isn't like you have to totally rewrite code to read them. You only have to slightly change existing code. (We're talking only a few lines of code). So, for that reason, I would say they are very much extensions of the current formats. The only real difference being that 0 & 1 have an index for a color while 4 & 5 explicitly define the color. 0 & 1 seem like a natural thing to do when you have only a small fixed number of colors that can be used. 4 & 5 seem natural when the colors are expected to be unique and within a very large range. 3 doesn't seem natural at all. I consider it nothing more than a hack and based on it never coming to light I think others felt that way as well.
I'll give you credit in that your idea of using a combination of formats 3, 2 & 0 or 1 to be able to specify frames with > 256 colors while still being backwards compatible is clever. But, it's a bloated approach and not likely to actually work since the legacy apps that were written before format 3 will more than likely not handle format 3 and not load the files at all. I know a well written app should skip unrecognized format headers but I am not even entirely certain that will work with format 3 because if the color bytes happened to spell "ILDA" then the software could be fooled into thinking it was at the beginning of a frame.
Format 3 was fixed when they decided it wasn't going to be a format.