Apple Patent Blocking PNG Development

Daniel writes: "Apple has a patent (U.S. Patent No. 5,379,129) on compositing a source and destination image using a mask image. This patent appears to read on alpha channel transparency, which the PNG and MNG file formats use. APPLE has declared in their patent statement to the Scalable Vector Graphics Working Group that their patent is only available for RAND Licensing. Since this patent appears to read on the PNG file format, Apple is hampering work on the PNG and MNG file formats. Perhaps Apple would like to clarify this situation by explicitly stating that this patent does not cover the PNG and MNG file formats or by RF Licensing their patent to the PNG and MNG development groups. Alternatively, the PNG and MNG developers are asking people to submit prior art in order to invalidate Apple's patent. SGI in particular appears to have prior art with their 'blendfunction.' Make sure the prior art you submit is older than May 08, 1992, the filing date of Apple's patent."

Comments

Prior art? Yeah, here's some prior art.

[Ryu2]

Try Porter and Duff's paper published in 1984 introducing image composition that started it all!!!!

Obvious and WAY older than 1992

[spitzak]

I have a book printed in 1990 (Foley & van Dam Computer Graphics Principles and Practice, first edition), the references papers that talk about it.

Every program that combines images and works with true color uses this. It is impossible not to, the algorithim is totally obvious. True color images (ie where the numbers represent levels of red, green, blue, rather than be indexes into a color pallette) were in common use in advanced visualization and simulations in 1980, such as Evans & Sutherland flight simulators. I also saw photo touch-up software that could duplicate a portion of the image and put it somewhere else in 1979, and I believe it must have used this, as otherwise the edges of the cut piece would be visible.

PORT84, Porter, T., and T.Duff, "Compositing Digital Images," SIGGRAPH 84, 253-259.

This is the paper most-often cited, however I think it's main addition is the enumeration of compositing operators and the introduction of "premultiplied" images. Before that I believe non-premultiplied was used as that was the more obvious solution. Also PNG does not use premulitplication.

Even if Apple is going to be a pita about this, it does not affect PNG, as the file format itself does not do any compositing, it just stores a 4th "color" called the alpha. So I would not worry about it there. However Photoshop and Gimp and the in-house program I write for Digital Domain uses this, and about six thousand other pieces of commercial and free software.

Obligatory link to the patent in question

[bjk4]

Patent 5,379,129

CLick the "Image" button to see the lovely diagrams. There are a few items this patent does NOT cover, as explicitely mentioned in the text. This patent does not cover additional channel information, like alpha channels. It only covers a seperate, full color, mask image that is used to mask off the source image.

I have a few questions:

• The patent mentions that a black pixel codes for source image. Does this mean that if I decide the opposite, I can avoid royalties?
  
• The patent mentions it uses full color mask images. Is a greyscale or black + white image considered a full color image?

If the patent does not cover greyscale images, then not all sanity is lost!

-B

Prior Art

[corebreech]

From the Second Edition of "Principles of Interactive Computer Graphics", by William M. Newman published by McGraw-Hill in 1979:

We find on page 222 the pseudocode for the WriteColor procedure:

procedure WriteColor(var ColorRaster: raster; x, y: integer);
var i, j: integer;
begin
for j := ColorRaster.ymin to ColorRaster.ymax - 1 do
for i := ColorRaster.xmin to ColorRaster.xmax - 1 do
if GetPixel(ColorRaster, i, j) <> transparent then
SetPixel(FrameBuffer, x+i, y+j, GetPixel(ColorRaster, i, j))
end;

And there was a First Edition published in 1973, for all I know it's in there too.

BTW, there isn't anything Bill Atkinson did for Apple in QuickDraw that isn't spelled out in this book.

Re:Time for an Apple boycot!?

[sporty]

Talk about FUD. Suing over interfaces has been known for a long time. I know 2 companies on the web battling it out right now just because they are competitors but one has a patent. Look at MS vs Apple, Apple vs Xerox (I swear there was a case). Don't make something so general so specific.

Also think of the confussion it causes when two products look identical. It should let the opensource community do this and not allow companies to do it? What's the insentive?

As for the one-click thing, please. They debuted one-click, so what. They are using, or at least were, using a technology. Don't pick on their website just 'cause they licensed a technology from another company that has a patent on it. What's cheaper, lisencing or going to court? Pick on Amazon for creating such a stupid patent.

And as for the cube-NDA contract issue, please. Gimme a friggin' break. So what if Apple likes to hide their new products until they are ready to do it the way they want. The employee signed a contract that he won't leak information from Apple. What if a competitor such as Dell caught wind and released something just as cool, but earlier. Great, Apple would get shafted 'cause of some ass employee who can't play nice.

As for the BSD crack you made, they are using opensource software and giving back the modified version, Darwin for free. Don't be an ingrate and say "Well, a patent (mind you invalid) exists, they get people to stop mimicing their interface verbatim, use OSS and recontribute while making their OS even more stable than before, so apple reeks." statement. Maybe we should step all over Apple and hack QuickTime, which uses a kick ass codec and try our best to drive the company into the ground.

You've just spread extreemist FUD about a company protecting its rights. piss me off...

Now this is just silly

[Sycraft-fu]

As other posters have mentioned:

1) This is obvious.

2) There is prior art.

So even if Apple decided to try and enforce it, it'd get shot down however:

3) Apple has already allowed royalty free use of this patent.

Seriously folks, this story is just silly. Think for a moment how many things out there use multi level alpha transparency. All 3d cards from the Voodoo on do, Windows does in movie file formats and in the UI in XP, a bunch of X WMs do. If Apple tried to enforce this they'd have legal teams from all over after them, fact aside that they've already said they won't.

Just because a company has a stupid patent doesn't mean they will try and enforce it.

Re:FUD?

[mav[LAG]]

There is nothing colossally more wrong with being able to patent software than there is with patenting hardware.

There's a huge difference between software innovation and hardware innovation. Software innovation is sequential and complementary. Software development is not a zero sum game. Developers have always used the work of others to build, improve and enhance functionality. In the open source and Free Software worlds, this works through availability of source and the distribution licenses. In the commercial worlds, it works through user groups, conferences and special interest Web sites (like this one) where people can share ideas and code.

If I make a living by it why should I spend time developing software if I can not protect my self from people ripping me off???

Be my guest. Discover a new algorithm without any access to the work of others (I'd be impressed with that straight away) and then patent it. Oops. Is it too expensive? Darn. It seems only the big companies can afford to patent XORing a bitmap with the background to achieve transparency, something I thought up independently when I was 12. And are they protecting themselves from others ripping off this "innovation". Nope - they use patents for attack, not defence. So smaller developers can't write software even when protected by the patent system.

If a commercial software developer comes up with a clever way of coding something he has a right to patent it like any other inventor.

This is a great idea - in theory. The problem is that there are no "clever ways of coding something" which don't boil down to techniques which have been used for years: linked lists, hash tables, look up tables, mathematical operations, bitwise operations and basic algorithms used on basic data structures. It's easy to check this too - pick any software patent held by say IBM, get it translated into English or pseudocode and it will be a trivial operation. Guaranteed.

Open source organizations will have to live with the fact that if some technology is patented by a commercial organization they can not use it free of charge and without permission.

Do you mean technology or software specifically? If software, then commercial organisations should not be using free or open source software at all. Come to think of it, they shouldn't be using the Internet either.

Pay up or bugger off that is the rule of the game. What Open source organizations can do is either come up with alternatives and/or they can stop whining about patents and try to beat Commercial organizations at their own game by patenting software them selves.

Software innovation doesn't happen when development is hampered by a mass of patents. In the non-software worlds, the inventor needs to recoup his costs and thus I can see the need for a limited time of protection. But in software, all that happens is those that can afford to hold patents use them as a weapon against those that can't. And very few true innovations happen in large patent-holding companies. It's the garage operations, the one or two guys in their back rooms who come up with new stuff all the time.
And why should non-US programmers pay license fees to someone like IBM in the US? What gives them the right to tell me whether I can or can't write software in a country 10 000 miles away?