finally I have some time to work through all the stuff we discussed on the Tagging Workshop.
Topics I would like to discuss:
Well, actually my impression was, that tagging won't help the USPTO very much right now ..
Anyway it could still be a very good thing for the community in the sense of finding code for use or reuse more quickly.
The 'aspect oriented tagging' is still the only idea I heard so far that goes beyond flat tagging or simple wiki documentation.
The good news is, that sourceforge said, that they will implement tagging one way or the other. If we can come up with something *cool*, they'll adopt it. The bad news is, that I'm not a semantic web expert, and that I doubt that I'm really able to figure out what is good and what is not..
Timestamping is a very clear subject, and we should just start a project to set up a server and DO this! I showed you what I already do on my site, but the problems are:
So all we would need, is a company that want's to support this project. We should also try to find magazines that will publish our hashes.
The benefits are:
No *research* is needed here, let's go and do this now!
I really forgot to show this project to the USPTO during the workshop. With TeamFound you can setup your own project, create categories and add websites to those categories (e.g. USPTO 707 category). So whenever an examiner would find a project during his examination work, he could just add that site into the appropriate category, even if it isn't useful for the current patent at all. The cool thing is, that once some websites are added, you can *full text search* those sites, and thus you'll find for your search-words only programs, that already some examiner categorised into that category in which you are currently interested AND you will also search google at the same time. (google results will be displayed on the same site, right after the TeamFound results :)
I did set up a simple demo project 'USPTO' with 3 categorys: 700, 701 and 707 Just install this Firefox Extensionhttp://prdownload.berlios.de/teamfound/teamfound-0.9-fx+fl.xpi restart your Firefox, go to menu Tools/TeamFound/Settings and set the Project-ID to '15' (projectid=15) and restart Firefox again.
General informations on this project can be found at:http://teamfound.berlios.de/
It turned out that the USPTO doesn't do any kind of community patent review internally at all .. They don't even do pair-examination .. They don't have much time for internet searches at all ..
Is it their fault? Probably not, it's more the politics and that the patent office finances itself, and and ..
So I think, we actually do help the examiners, whenever we put more political pressure and write bad press about bad patents. This gives the examiners big help for argueing with their bosses and their bosses' bosses that there is really *much more time needed* to actually check if a patent is new or not. Also I really doubt that a single person can verify that something is new at all!
One possible way would be some kind of campaign, where we would oppose EACH SINGLE PATENT in category G06F, that will be granted during a whole month (e.g. January 2007). I'm pretty sure, that we could show that about 80% of all granted patents in this category are actually not new. (In germany an opposition costs only 200,- Euro, no matter if you loose or win!) The count of patent grants in G06F was in germany in June 2005: 50 (fifty granted patents).