Hacker News new | past | comments | ask | show | jobs | submit login
White House nixes Patent Office pick after tech-sector outcry (arstechnica.com)
213 points by aaronharnly on July 11, 2014 | hide | past | web | favorite | 20 comments



And in other news, today the USPTO gave Lockheed Martin a $268,910.52 competition-free contract because it is apparently the "only" company in the United States writing enterprise software.

https://www.fbo.gov/index?s=opportunity&mode=form&id=c6652fc...

It's good that the White House backed off here, but the USPTO is still an ungodly mess.

I have been politely asking both USPTO and the White House to fix the problem where courts report IP assets involved in litigation via a form called AO 120 for some time (months)--to no avail. AO stands for Administrative Office, which means the judicial branch controls the form. To view all of these forms on PACER it's $0.10 per page. Fortunately, they're shipped (really, shipped) to USPTO where they're made available on an automated FOIA site. That's great, except that they're printed and scanned, and huge numbers of them are filled out incorrectly; stamped with various crazy stamps; entirely blank in fields, etc. Also, different courts report case numbers completely differently.

This should be handled with a web-based system that would take most people on this site a few days to write and test. But here we are, in 2014, trying to convince the people who decide what is and is not an invention in the United States of America that they should start using the web.


No you couldn't.

You could build something half-baked that didn't work in the majority of cases.

If you think you could build a fully-operational EDMS system capable of handling a variety of documents coming in from different source, provide workflow management, change control, auditing, accurate search, reporting tools, analytics, third-party integrations (CRM, email,etc.), FOIA support, etc. in a couple of days then I'd suggest you try it.

There's a reason USPTO wanted an off-the-shelf solution rather than building their own, it's a lot cheaper and faster to buy something pre-made than to spend millions building your own.


I think the problem is the "competition free" part.


Read the linked document, they looked at other vendors, all the other vendors that could meet the requirements needed to make significant (charged) customization to their product to bring up to spec. LM had the only product that could be used off-the-shelf.


While I generally agree that the government should buy something off-the-shelf that's already built -- and at a quarter-million dollars, this doesn't even come close to boondoggle territory -- sometimes the requirements are set explicitly to grant the purchase to a specific buyer the government had in mind, instead of the requirements really being generated in a neutral manner and then looking at the marketplace.


If you think you could build...

Oddly enough, The Internet Archive did a substantial portion of just that: https://openlibrary.org/dev/docs/bookreader

It's used by numerous other organizations as well:

https://openlibrary.org/dev/docs/bookreader#users


I don't think you understand what USPTO need, it's not just a scanning system, it's an entire workflow system.


I didn't claim TIA created everything the USPTO requires.

I did point out that they created precisely the element you indicated was challenging: accomodating different document types from different sources.

Given that the USPTO's system is likely based on electronic filings at this point, even the scan-and-present aspects of the BookReader solution are likely more than is required.

Another salient point is that by breaking the problem into discrete components, you're more likely to come up with a viable solution, or at least significant portions of one.

And I've worked on a number of systems (FDA clinical information systems) which address a great deal of your other feature points. Suffice to say: such systems do in fact exist.

They do, in fairness, take more than a couple of days to assemble.


But workflow systems can't do shit with scanned documents other than route them via human instructions.


> This should be handled with a web-based system that would take most people on this site a few days to write and test.

Ah yes, the mythical "system that [should] take . . . a few days to write and test." Gives me flashbacks to running code in Visual C++'s debugger during a live demo so I could edit and continue if it crashed.


Hmm, they're parsing the reports a bit stronger than I am. The nomination is held back for now, but it hasn't quite shuffled off the mortal coil yet.

PS - If folks are interested, the new satellite USPTO office in San Jose would love to hear from you! Come to Hacker Dojo and talk to the Hackers & Founders team if you want help setting something up with the USPTO staff. Or talk to Silicon Valley Leadership Group, who also has good relationships with them (probably better, to be fair!).


> Hmm, they're parsing the reports a bit stronger than I am. The nomination is held back for now, but it hasn't quite shuffled off the mortal coil yet.

Its not even clear that the planned nomination was ever even real -- it was claimed in an "e-mail newsletter" by an author citing anonymous "reliable sources" and then picked up by outlets with broader readership (like Ars) and reported (and headlined) as fact; whether the origin was a result of real knowledgeable people leaking information, or a trial balloon, or people wanting to assure a contrary result pretending to knowledge to create political pressure in the opposite direction or, the author of an email newsletter making stuff up to get attention, or any of a million other ways Washington rumors get manufactured is far from clear.


> Its not even clear that the planned nomination was ever even real -- it was claimed in an "e-mail newsletter" by an author citing anonymous "reliable sources" and then picked up by outlets with broader readership (like Ars) and reported (and headlined) as fact; whether the origin was a result of real knowledgeable people leaking information, or a trial balloon, or people wanting to assure a contrary result pretending to knowledge to create political pressure in the opposite direction or, the author of an email newsletter making stuff up to get attention, or any of a million other ways Washington rumors get manufactured is far from clear.

It's a fair point, as we don't and likely won't know the why. Hell, I see the event chain as a not-unlikely consolation prize that "we" defeated an opponent of patent reform at least, even if the patent reform legislation failed.

But also, we were so badly defeated on the legislative front, the White House could have wanted to test if they needed to switch positions based on relative political strength, if we couldn't stop the nomination.

Having just gotten back from DC to talk startup visa implementation, it's been underlined to me how much goes on under the surface in that town...


"floated" is the term you are looking for.


If Senators were commenting on it then it was real.


> If Senators were commenting on it then it was real.

Senators comment on things that aren't real (either because of erroneous beliefs about what is real or because they see political advantage in commenting even though they correctly believe it to be unreal) all the time.

Senators are, as a class, neither omniscient nor omnibenevolent.


As they say in DC, though, "perception is reality."


Considering the administration's public statements in the past, this would have been a pretty surprising pick for them. How sure were people about the nominee. Also, anyone know anything about the acting director from google?


Michelle Lee basically came to the USPTO straight from Google, and has a strong silicon valley background. This is much more so than most other senior officials from USPTO in DC area.


Michelle Lee is smart, but the bureaucracy under her is massive and it will take the White House and Congress to really set any changes in motion I'm afraid.




Registration is open for Startup School 2019. Classes start July 22nd.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: