Slashdot Mirror
FreeType Project Cheers TrueType Patent Expiration
FlorianMueller writes "The FreeType project celebrates the expiration of Apple's TrueType bytecode patents. The open source font rendering engine now has the bytecode technology enabled by default. The relevant code existed for some time, but the project felt forced to disable it and advise everyone not to use it due to patent encumbrance. The 20-year maximum of validity of software patents is long, but sometimes the stuff that becomes available is still useful. The Unisys GIF patent was an example. And anything open-sourced 20 years ago would also be patent-free by now (except for the code that has since been added)."
The open source font rendering engine now has the bytecode technology enabled by default. The relevant code existed for some time, but the project felt forced to disable it and advise everyone not to use it due to patent encumbrance.
The statement above makes me worried because it suggests that the Open Source Community could not find their way around these patents for two decades! Think about it....20 years!
That's enough time for an infant born at the time of patent filing to [legally] be a parent at its expiry...and that's a long time folks.
Didn't most distributions that shipped FreeType enable usage of TrueType fonts by default already, or am I misunderstanding this?
"People who think they know everything are very annoying to those of us who do."-Mark Twain
My google-fu must be weak today: is there an example comparing small font sizes with and without the byte code interpreter enabled?
This may become the next drive where software patents are concerned. In much the same way that the H1-B programs get extended and increased while unemployment is hitting new highs, software patents stifle business and innovation and keep the courts systems over-burdened... we need more of it!
Apple apparently uses FreeType in the iPhone. Go to settings->General->Legal and you get the long list of projects uses in the iPhone; the Freetype project is mentioned about a third of the way down (right below the copyright notice for ncurses).
I suppose Apple had no issue recompiling with the flag turned on.
The code would be under copyright, but patent free. Not the same things. That mean you would not be able to COPY the code snippet, but you would be free to make your own implementation.
C. Sagan : A demon haunted world:
http://www.amazon.com/gp/product/0345409469/
visit randi.org
Say Microsoft decided to enforce their double-click patent, how would you find a way around it?
Microsoft cannot be granted a patent on such a thing. What they could patent is the implementation of how events are handled by the system. So double clicking wouldn't be patentable. Under the patent regime today, what they could patent includes the method of effecting a [desired] change after double clicking.
Please correct me if I am wrong.
Here's a side-by-side screenshot of Linux font rendering with and without the now patent-free byte code interpretation: http://avi.alkalay.net/2007/01/freetype-with-bytecode-interpreter.html
Slashdot really needs a moderation score of "-1 Incoherent"
What the fuck are you talking about? How the hell will "some code from Wine" added to a "X11 backend" benefit anything bandwidth wise? You don't make any sense whatsoever.
WTF?!
Not to say whether there's any sense in whatever you're rambling on about or not, but it sure ain't the same thing as the article's talking about, and makes less than no sense in this context.
Copyright needs to be lowered. I find it amazing that huge industries were unable to extend the patent duration while the entertainment industries were able to extend this to over 100 years from the original 14+14 years.
She was like chocolate when she drank... semi-sweet at first and then increasingly bitter.
How is not clear I'm talking about possible optimization and possible trends? Don't expect a complete posted dissertation with all technical details embedded into a Slashdot comment.
No... then no one would be able to read what I wrote. I do the majority of my commenting between 2am and 4am. I know it shows when I later re-read them over the course of my waking hours. Reminds me of a horrible typing mistake when I was ...it didn't cum out well.
TRYING to discuss a "Seagate" "hard diSk" getting inserted into a live RAID.
The article talks about a patent expiration which would enable technology, just like the article mentioned how FreeType having code enabled from what was disabled because of that patent expiration. What is so hard to understand about that? That's all I'm saying besides a possible trend.
Just because the patent expired doesn't make it great. Some of the stuff rendered with BCI almost makes me want to go back to bitmapped fonts. Or maybe it works well, but only with non-free fonts. If that's the case, here's hoping that free operating systems that use FreeType do NOT make this the default immediately.
While I don't know how fonts are rendered in Wine, if Wine actually renders them itself and sends the resulting bitmaps to the X server, then this will certainly consume more bandwidth than if it just sends the text and font info and lets X render it.
The Tao of math: The numbers you can count are not the real numbers.
Does anyone know how to enable with Freetype, the same kind of font appearance as apple does on OS X?
Some say apples font rendering is a bit more blurry, but I find it easier on the eyes than stock font rendering under linux.
Patents used to last 17 years after issue. But certain patent owners would employ dirty tricks to keep a patent in the "pending" state (filed but not issued). Now they last 20 years after filing, with an available 5-year extension for certain health care products to compensate for delays in getting the FDA's marketing approval. This greatly reduced the available techniques to keep a patent below water, though nonpracticing entities have discovered a few that still work.
U.S. copyright is longer because a copyright is far easier to work around than a patent is to invent around. The methods used in a computer program cannot be copyrighted (17 USC 102(b)). Only the expression can, and the key case for that was CA v. Altai. Names of commands were ruled uncopyrightable in Lotus v. Borland, allowing look-alikes and work-alikes of office applications such as OpenOffice.org. But two disturbing copyright interpretations remain: the uncertainty in the scope of copyright in video game scenarios (Atari v. Philips ruled for wide copyright; Capcom v. Data East ruled for narrow), and the possibility of accidentally infringing the copyright in a sequence of musical notes that someone else happened to claim first (Bright Tunes v. Harrisongs).
Easy, use a triple-click with an "accessibility" tolerance of +/- one click.
Support Right To Repair Legislation.
Well, Truetype BCI is useful but today most computer screens are LCD it isn't enough anymore, we also need a good subpixel rendering method, the one included in Freetype isn't so bad but isn't as good as Cleartype either. And the Cleartype code in freetype had been removed, unlike the BCI, which is only disabled, so we who don't live in the US can't use it either. (there are still a way to patch freetype though). I wonder how long will we have to wait until we can have nice looking fonts on Linux desktop.
if Wine actually renders them itself and sends the resulting bitmaps to the X server, then this will certainly consume more bandwidth than if it just sends the text and font info and lets X render it.
Mod parent very insightful.
Some used to pre-render a larger bitmap full of commonly used letters. For example, a 2kx2k bitmap cut into squares for each letter. As letters are needed, an empty square is found and filled with the pre-rendered letter with all anti-aliasing and such applied. Then the program just tells X11 to copy and blend that square to the destination. The bitmap acts like a cache that doesn't have to constantly take up bandwidth to fully send.
That has worked well except when you get into letters combination that shape differently due to different letters in the combination. For example, the tail of the y may extend further under some scripted letters than others. In others languages besides English, this occurs more often.
The enable technology the article speaks about could help in this area if the X11 server was optimized to handle such bytecode interpretation internally. Then there is no need to fill the bitmap cache with every combination of shaped letters.
What some of these anonymous cowards don't realize is the size of the cache needed to store all possible unicode characters combination with all shapes and styles applied, and then you should realize the cache method has become useless.
Dyslexic typing is not trollish, and maybe you should be mindful of those of us with such disabilities.
Notice that Slashdot has never added the "edit" feature to a comment, so that those of us with such disability can come back later and fix the comment. We just have to let live with trollish comments like yours when disabilities have taken over when we try to share the same abilities to post like you.
Useful font stuff here:
http://linuxtweaking.blogspot.com/2010/03/fedora-12-improving-awful-font.html
I've just enabled byte code support on my laptop - makes a big difference.
AT&ROFLMAO
Well, it's simple: It gives good results if the designer of the font put considerable work into it (and the software correctly interprets it, but I assume that's the case). It gives terrible results if the font designer didn't. Now if it looks bad for free fonts, then I can imagine several possible causes:
The Tao of math: The numbers you can count are not the real numbers.
Though you might think so, this is incorrect. In modern fonts far less information is sent by transmitting the bitmap than the font description and this is how all modern font rendering works including on X with XRender.
There are many adjectives or terms that could be used to describe posts on slashdot that are not among the ones normally listed. In addition to the one you recommend, -1 Misinformative, -1 Advertisement, -1 Suspicious link(s), -1 Conspiracy Theory, and -1 Tasteless are only a few that spring to mind.
File under 'M' for 'Manic ranting'
Shouldn't that be +1 Tasteless?
If Freetype was under GPL3, however, this would still not be violation of their patents because they are the owner. Likewise, any GPL product that wants to include the patent would be allowed also, since the GPL3 license says that.
"Improvement". Um... no, the details of what makes the font appear distinctive area destroyed by "BCI". The fonts are not the same -- perhaps in terms of set-width, but not at all for typographic detail.
I guess "BCI" simply replaces the actual character forms with something else at lower pixel resolutions.
For example, look at the 13th line in the example - look at the lower case "c" and "d". Notice that the "BCI" version is considerably more open. Indeed, it has become a much more modern font! Note that kerning would then have to change -- since the bodies of the letters are significantly larger, keeping the kerning rules of the original font would result in a very squished look. (actually, I believe the opposite happens these days -- kerning will simply be ignored, or the kerning rules of a modern font such as Times Roman are applied, and the look of the printed copy allowed to deteriorate.)
At 300dpi resolution, it is very hard to produce proper fonts (I produced a very popular laser printer based proofing system in the 80s -- along with a complement of over 3,000 fonts, for typographers. Hinting at 300dpi did not give accurate enough results for the typographers of the time). At 600dpi it becomes a more reasonable proposition. At typical screen resolutions? Readability and font fidelity are not compatible for classical typefaces. "BCI" may be a solution for you, as long as you realize that what you are seeing on the screen is not actually reflected in print output.
In general, I recommend that screen fonts be chosen for legibility. Until display technology catches up to 600dpi or greater, you just won't get classical fonts to both reflect readability and artistic intent.
Using classic fonts as a primary display format for computer monitors is like viewing the Mona Lisa on a monitor encoded as a jpeg.
Just another "Cubible(sic) Joe" 2 17 3061
Good job getting this gibberish modded "interesting".
I'm still using bitmapped X fonts to this day, even in Firefox.
I much prefer hand-tuned, razor-sharp characters to "hinted", blury ones.
2.4.0 appears to be buggy so they have to re-release it as 2.4.1. Simply amazing ;-)
Lawyers make up the bulk of congress. Of course they just automatically rig the system to benefit their guild, and why this apparent need for millions more laws, and writing them in inane gibberish or "lawspeak". It's a conflict of interest, and it has resulted in a rapidly approaching non functional society, the governmental and also the business side, as in the example right here, nutso software patents, for "products" that already enjoy copyright protection. One or the other, but both is clearly just abuse of society in general and is hindering, not advancing, the useful arts and sciences.
The only reason I can think of that a tasteless comment would deserve a + moderation is if the remark also happened to be particularly amusing (possibly simply because of how tasteless it was). But we already have a mod score for that one: +1 funny.
File under 'M' for 'Manic ranting'
Really? Celebrate over this? If your fonts looked like crap before you would have either enabled the flag and compiled it yourself or moved on to windows by now. Who the hell cares about the patent...
Another person confused on the difference between patent, trademark, and copyright.
Steamboat Willie gets a copyright extension because government is bought and paid for. Mickey Mouse will be a trademark as long as Disney exists. Neither of these are patents.
I think Slashdot needs a "+/-0, WTF" mod.
Property is theft.
Mickey Mouse will be a trademark as long as Disney exists.
True, but at least in the United States, a trademark cannot be used to extend the exclusive rights in a copyrighted work past the copyright's expiration. Dastar v. Fox.
Are you referring to the Circular Transportation Facilitation Device? (google it). Has already been patented, in Australia.
The 'normal' way of rendering text with a modern X server is to use the XRENDER extension. The client renders the glyphs that it will need for a particular font (typeface+size+attributes) to a buffer and sends them to the X server, as a set of images with an alpha channel. It then draws them by telling the X server where to composite them. On a modern GPU, the glyphs will be stored in VRAM and the GPU will handle the compositing. This is exactly the same mechanism that Apple uses with Quartz GL.
A few years go (2003?), some guys at MS Research published a nicer method. TrueType fonts are constructed form Bezier curves. You can approximate a quadratic bezier using two triangles, giving a polygon that contains the line of the curve. The did this and then wrote a shader program that generated the curve from the polygon. The advantage of this approach is that the client only needs to provide a single image for each glyph in the typeface, not one for each size (and, because it's a simple vector image, it's often smaller, so takes less VRAM), and the GPU gets to handle the rasterization (which is basically what it's designed to do) as well as the compositing. I don't know if they actually did this for any shipping products yet.
Oh, and the original post by Dzonatas was just incoherent rambling. I've no idea what he thought he was saying, but what he said was just a string of unconnected concepts.
I am TheRaven on Soylent News
I use Ubuntu, so, according to some here, I already enjoy the benefit of this hinting algorithm. But is this why PDFs viewed in xpdf look so bad, with fuzzy type, but in acroread look fine? Is acroread using this algorithm, but xpdf is not?