Slashdot Mirror
WebSense Patents Censorware System
Matthew Skala writes "As reported in SiliconValley.internet.com, filtering-software vendor Websense has received US Patent 6,606,659 on a "System and method for controlling access to internet sites". The new features in the patented system seem to revolve around using time limits instead of filtering sites out entirely; offering users a choice of viewing a site and having it logged, or not viewing it; and a scheme for automatically categorizing sites that looks very much like the "Bayesian filters" we've heard so much about in recent weeks. You may be interested in the filtering company's press release about their patent, or my own view."
If internet filters are going to cost money, then maybe schools and libraries will stop using them.
My local library blocks out anything to do with pregnancy (like the council run pregnancy advice service), anything with chat in the domain name (like the casual chat web forum) but doesn't block goatse.cx. Go figure.
The only news that could be better is that someone had patented spam emailling and was taking every spammer in the world to court.
Beep beep.
I don't want to use a censorware application anyway. Hopefully they price things high so that other people won't use them, an in particular, so that the government won't use them (in libraries, etc).
Now, if you could find where someone had software that used Bysien (sp) filters for categorizing web sites, and also used time of day as an input into the access control algorithm, from my quick read over this, that would be prior art.
I, for one, have never seen this approach used before, so the patent may be valid. Remember, he didn't patent filtering, or time of day; the patent is for a specific application of multiple algorithms.
Patent or no, I like this approach. The rules-based approach just doesn't work well. Maybe this will have some sanity. This company deserves kudos for thinking.
Different topic (one I've always wondered): If there are any lawyers reading this: My understanding of patents is that if I profit off someone else's patenteded invention, I owe them money. Legally speaking, if I implement their patented technique completly in the open, and give it to the public domain, do I (or anyone) have any liability, as long as no one makes money off of it?
Sarcasm and hyperbole are the final refuges for weak minds
The patent is GB2366891. The crux of it is that programs that use more than a certain percentage of the CPU (eg: 50%) are incrementally slowed down by quickly pausing/unpausing their threads at short intervals until their CPU usage is reduced below the threshold.
How does this qualify for a patent? It's self evident! Things like this have been done in real time control systems (software and physical) for decades. It is nothing more than a high-frequency 1-bit DAC controller. Just because instead of controlling chemical reaction rates, the system is used for controlling processor usage, suddenly this method is worthy of a patent? Take a look at one of their diagrams. Is that the standard for new and inventive developments in the software industry? A flow chart with four, count them, four steps?
The patent system needs an overhaul, and fast.
In his article, Skala tars Bayesian spam filters with the same brush as website filtering. He touches on the differences but seems to miss the main one - one has to actively seek out material on the web, while spam is pushed onto you by third parties. If my child happens to type in a web address that shows pr0n, I may not like it, but short of typosquatting, the site is not in the wrong. However, if a spam merchant sends email to my child's account that includes images of sex acts, the child has no choice to not view it. So Bayesian filters are great for spam, because they are freeing me from seeing crap I don't like, and would waste my time to delete. Web filters prevent me from accessing stuff (good or bad) that I know may be out there, and must actively seek.
Does this particular story add anything to the debate or is it just a troll?
It most certainly *does* add something. If you say something once, people will be very unlikely to remember it. If you say it twice, a few will remember it.
If it's repeated every week or so for a year, most everybody will have gotten the point. It's called "repetition".
I'll paraphrase Hitler: "Repeat a lie often enough and people will believe it to be true".
Except, in this case, there's no lie, except maybe at the patent office.
I have no problem with your religion until you decide it's reason to deprive others of the truth.
So parents trying to keep their young children off porn sites are the same people defending software patents? What does good parenting have to do with software patents?
So not allowing my young children to see porn makes me evil? I'm sorry, there are substantial, good uses for this type of software. There are also bad uses. It is like any tool: the tool is not good or bad, the use is good or bad.
Sarcasm and hyperbole are the final refuges for weak minds
No, not really. See, you can always give your password to someone else (or rsa key if that may be the case). The major "concern" here, if you want to call it that, is the GNU folks don't like giving the sysadmin "too much control." Something like logon restrictions is against their thinking, but then again, the sysadmin can just change your password or delete the user.
Of course, since you can do that, it makes the 'wheel' rant a bit silly, now doesn't it? In theory your statement should hold true, but it doesn't, so the whole "freedom" thing falls apart. If it were truely free, then you'd be right -- no one would have passwords. So disabling 'wheel' support in the name of freedom, in my opinion, is just dumb. Root is privelged access and *should* be restricted.