Peer To Patent

The UK Patent Office has an online site for patent commentary and prior art research. This should be checked for functionality that might make sense to me.

This effort is going rather slowly; I've been busy!

Patent database

Here's a "world patent database" in Chinese: [list] [patent] Interesting cross-referencing going on there. Makes me want to scrape the whole thing. (They might notice, of course.)

Progress thus far

CSS and HTML work perfectly in the Decl context now (if somewhat primitively) and so the building blocks for Websites are in place. I can even emit PHP in the right places, and collect Javascript and CSS into centrally organized files.

What I'm working on now is the macro/template system. This is the meat of the thing, and arguably the second Really Big Part of the Decl system. (The first being its very nature as an easily parsed, concise language for the construction of complex data structures.)

The macro system will essentially be the way that Decl can impose structure on any textual language, and it's going to be the way we build abstractions of site features (like "user signon system" or "forum") and page features (like "newsbox" or "navigation menu").

In other words, the project itself has now started.

Got restarted on CSS today

I know, I know, most of the action isn't even Depatenting-related; I'm still working on infrastructure.

FreePatentsOnline: literature search tools

FreePatentsOnline has added journals, many with full text, to its search tool.

Article One Partners

Article One Partners has the crowdsourcing aspect already in place. I don't see an open community, which is more what I hope to do (provide a place for the hacker community to see and make news about patents), but they're definitely competition. (No reason I couldn't index them, of course.)

Progress

I've been spending a lot of time with WWW::Publisher and HTML::Declarative. The new target is (was) June 15, but I don't see that actually happening at that point, because three of my action points are still Macro/boilerplate facility, DB management code, and Write/debug site.

Granted, my initial specification of the site is actually written. But when that hits viable code status, it isn't going to work. I'm down to two weeks, during which time I'll be flying to Budapest for my niece's wedding. So maybe I'll have to wait for my next Google coupon.

That said - the community is really buzzing with patent furor lately, thanks to Lodsys and their in-app upgrade patent. Wow. This is going to be well-timed if I can just get it out the door.

Linkdump

• A Groklaw post on patents
• HNN thread: court rules Linux kernel infriges

Polyphasic sleep

Just in case you were wondering, starting polyphasic sleep in order to catch up with something doesn't work.

Marketing

I have a $100 coupon from AdWords good until the end of the month - so that, my friends, is the date I want to have something live. It's ... looking less probable. Sigh. I was sick for four days during a busy week, and I've been trying polyphasic sleep in an attempt to catch up since then.

I've been haphazardly working on database definition code in the framework. It might honestly be less stupid to just develop the damned site, but the whole point is the framework, so ...

Anyway, once I have something operational, I should find out who's doing patent stuff online and thus might be interested in this. To that end, I've found one: http://fosspatents.blogspot.com/ - so once things are that far, I'll drop him a line. This post is just so I don't forget.

Here's another site: http://news.swpat.org/2010/03/transcript-tridgell-patents/ - that's actually a pretty good transcript of a talk about patent work, indicating that non-infringement is often what you really want to go for. A depatenting community should probably support workarounds as the speaker suggests at the end of the talk.

I should try to find out who's who in the field.

Rough estimate of complexity

OK, so I count 18 pages plus a generic administration tool for the database, plus the database. According to Start Small Stay Small, that gives me something like 240 hours of work ahead of me. I'm not sure how true that is, but I'm damn well going to find out.

Rough feature list

So. Depatentizing should feature:

• A home page for each patent in question
• News searches related to the patent (i.e. what impact the patent is having)
• A description of the claims of the patent
• A way to post prior art in a way that's not just the forum
• A way to post bounties, with escrow
• A way to submit bounty claims
• A single PDF retrieval for a report on a given bounty and all activity related to it at any time, in whatever format is useful for attorneys (like I know!)
• Specific abuse cases posted in a way that's not just the forum
• Forum/commentary on any primary object (patents, news items, prior art submissions, bounties, abuse cases, and bounty claims that have specifically been posted as open)

That's all I can think of tonight. More will probably occur later as I flesh this out.

Analysis of BountyQuest (I)

This is just going to scratch the surface here, but the Wayback Machine has pretty good records for BountyQuest from 2000 to 2003 when it closed. (In 2006 the domain was parked, with the same image it has now. Kinda weird.) Here's the index.

The central focus is the bounty. Since a bounty seems to correspond 1-1 with a patent, I'm just going to call my central table the "patent". Whether a bounty gets posted or not is immaterial.

The "how to read this page" instructions for a specific bounty basically tell us the table schema that BountyQuest was using. (Incidentally, look at the amounts of money in play here. Staggering.)

BQ used Coolboard.com for a forum provider. The titles and structure of the boards are stored, but not the individual content, sadly.

The submission form is also not archived (it's behind a post button, for some reason - perhaps because it contained the bounty number?)

But at least we know the structure of a bounty at BQ!

History

Back in 2000, Jeff Bezos and Tim O'Reilly put together a neat-o site called BountyQuest. I remember it quite well (although I was shocked to realize how long ago it had been), and when a couple of patent outrages took place this year, I was shocked again to realize BountyQuest had gone away (Tim O'Reilly explained why in 2003).

Depatentizing is meant to be kind of a revival of the concept. BountyQuest was the attempt to start a business on this; Depatentizing is just going to tick along on its $20/month Linode instance with a hat out.

Oh, I'll put in a bounty system, just in case. But I'm not focused on a business model; I want something to serve as a focus for patent abuse cases.

Welcome

This is a stream-of-consciousness blog regarding my development of Depatenting.com. It's going to overlap to a certain extent with the semantic programming blog because this site is intended to be a showcase of the techniques I'm evolving there, but this will focus more on the minutiae of the Depatenting site itself and my thoughts as I wrestle with actually producing something.

The target architecture for Depatenting is standard LAMP (P=PHP here) and there may be some fancy Javascript if I can manage to make it make sense. It'll probably host on Linode; during testing I'll self-host it on my development machine, but once it hits usability it's getting its own digs.