klug@mit-amt.MEDIA.MIT.EDU (Michael Klug) (07/26/89)
I'm looking for routines that do graphics. It would be most helpful to me if these are written in a computer language. Does anyone out there have anything like this? Thanks in advance. :-) Michael Klug
joe@hanauma.stanford.edu (Joe Dellinger) (07/26/89)
In article <354@mit-amt.MEDIA.MIT.EDU> klug@media-lab.media.mit.edu.UUCP (Michael Klug) writes: > >I'm looking for routines that do graphics. It would be most >helpful to me if these are written in a computer language. Does >anyone out there have anything like this? Thanks in advance. :-) Every parody needs a serious article to go with it, or it doesn't work! OK, here's a serious article to go with this one: I am the primary designer of the "Vplot" graphics driver system. (I hope people out there have found it useful.) The main advantage of Vplot (and why we keep using it here) is that it allows you to support lots of different graphical primitives (like vectors, filled areas, raster, text) on lots of different kinds of devices (vector, raster, smart, dumb, color, monochrome, screen, hardcopy, large, small, etc). Starting a few months from now (got to get a PhD first), I think I would like to make a "succesor to Vplot" graphics driver system. As much as possible, I would like to merely ASSEMBLE other code that already exists and not WRITE new code. To this purpose, I am interested in collecting useful generic graphical subroutines. Many of the routines I need I'll be able to steal from Vplot, but I'd like to be able to support a more postscript-compatible set of graphical primitives than Vplot currently uses. Here is a shopping list of routines I already know I'd like: 1) INPUT: a set of N RGB triples M, the number of color table entries to fill OUTPUT: a set of M RGB triples which come closest to covering all the colors present in the input set of N RGB triples 2) INPUT: a set of M RGB triples specifying a color table a single RGB triple OUTPUT: the color table entry which is closest in color to the input single RGB triple 3) INPUT: a set of bezier curve definitions some sort of "resolution required" specification OUTPUT: a set of points that trace out the bezier curves using straight vectors to the required resolution 4) INPUT: a clipping window defined by an arbitrary path ("A") an arbitrary path ("B") OUTPUT: a set of paths that result when input path "B" is clipped to the clipping path "A" 5) Same as (4), but clip a FILLED PATH to another PATH. (Harder!) Don't bother sending any routines that can't be redistributed with another package. All contributors will be acknowledged in the source code and documentation, if you so wish. If you don't have code to give me but know a good reference in a book I can get, that would be good to know too. Thanks! With a little group effort, we could put together something _very_ useful for all of us! If I can pull this off, we'll be able to plot postscript on anything! \ /\ /\ /\/\/\/\/\/\/\.-.-.-.-.......___________ \ / \ / \ /Dept of Geophysics, Stanford University \/\/\.-.-....___ \/ \/ \/Joe Dellinger joe@hanauma.stanford.edu apple!hanauma!joe\/\.-._
rokicki@polya.Stanford.EDU (Tomas G. Rokicki) (07/26/89)
klug@mit-amt.MEDIA.MIT.EDU (Michael Klug) writes: > I'm looking for routines that do graphics. It would be most > helpful to me if these are written in a computer language. Does > anyone out there have anything like this? Thanks in advance. :-) Thanks, Michael, for the biggest laugh today. It has been depressing in this conference of late, completely devoid of any graphics discussions. Okay, I'll start the ball rolling. Dithering seems to be a hot topic, so let's talk. We've got the various error diffusion algorithms that seem to be doing real well for the screens, and the classic clustered dot algorithms for printers that have troubles with isolated dots. Has anyone tried to combine the two, so you get a clustered dot F-S, for instance? This might be done with hysteresis (bloody word isn't in my dictionary; sorry for the spelling); for instance, perhaps something like an error function of -1 3 1 1 -1 -3 2 1 . . . . . or some such (I have no idea what would be reasonable numbers) but use a *negative* value for the feedback, so you get clustering. I'm not even sure if this would work. This might help reduce some of the aliasing/big dots problems with clustered dots, while not depending on isolated pixels printing well . . . -tom
markc@Solbourne.COM (Mark Connell) (07/27/89)
In article <354@mit-amt.MEDIA.MIT.EDU>, klug@mit-amt.MEDIA.MIT.EDU (Michael Klug) writes: > > I'm looking for routines that do graphics. I have a routine that will turn on a single pixel on a memory mapped 8-plane dumb frame buffer. All you need to supply is the address of the pixel and the 8 bit value you want stored in the pixel. Send $10,000 to me at the address below and I will send it to you, complete with source code in C (a computer language). > Thanks in advance. :-) You're welcome. :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) Mark A. Connell Solbourne Computer, Inc. 1900 Pike Road Longmont, Co 80501 (303) 772-3400 markc@Solbourne.COM ...!uunet!stan!markc
rwg@Solbourne.COM (Rick Gillespie) (07/28/89)
In article <1796@stan.Solbourne.COM> markc@Solbourne.COM (Mark Connell) writes: >> I'm looking for routines that do graphics. > >I have a routine that will turn on a single pixel on a memory mapped >8-plane dumb frame buffer. All you need to supply is the address of the >pixel and the 8 bit value you want stored in the pixel. Send $10,000 to me >at the address below and I will send it to you, complete with source code >in C (a computer language). >:-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) :-) Of course, what Mark has neglected is a routine to initialize the lookup table. So, if you want a truly useful set of graphics routines (set a pixel, and set a lookup table entry) then send me your first born male child. For twins I'll throw in a routine to set a pixel in a 24-plane system. :-)
NU113738@NDSUVM1.BITNET (07/28/89)
Any preferable Language, Computer system, graphics system? It would help to be a little more specific. Also, what type of routines? General line, ellipse, fill algorithms or programming routines for monitor/graphics card autodetect and initialize? I may be able to help, but please, give me some more info. Jeff Bakke NU113738@NDSUVM1.BITNET