[ih] Why the six month draft expiration ?

[Karl Auerbach]

You (and Scott) are both correct that the USPTO has improved from the 
days when Bruce Lehman used to call those submitting patent applications 
and paying USPTO fees "our customers" and promised to focus on servicing 
those paying customers.

(Lehman was so deeply in the pocket of patent applicants that he was 
actually booed after a presentation to a meeting of IP lawyers in San 
Jose.  And Carl Malamud's buying - and then publishing - USPTO's 
income-generating, but public domain, DVDs of patent materials - drove 
Lehman to turn apoplectic shades of red at the mere mention of Carl's 
name, a fact that I have personally tested and verified.)

To provide an example, although not IETF related:  One has to wonder at 
the institutional blindness when the USPTO issued to IBM a patent on 
token ring well after Farber's work on DCS at UC Irvine.  (Eventually, 
if memory serves, the IBM patent was cancelled after issuance because of 
DCS.)

Today the community (in part thanks to EFF) has become more vigilant 
about recent era applications and raising awareness of prior thinking 
before issuance.

The key word, as you note, is "before".

Once issued a patent gains a rebuttable presumption of validity; that 
presumption changes the calculus of the weight of evidence that is 
needed to reverse the granted patent.  It's the usual case of "easy" (or 
"easier") and less expensive before but "harder" and more expensive 
afterwords.

Long ago I suggested that the IETF follow the path of entities such as 
the IEEE, by which I mean, the IETF should create a printed (yes, paper 
[perhaps only printed on demand] as well as online) quarterly journal, 
call it "Proceedings of the IETF", which would publish every ID and 
minutes of every working group (including email interactions).

(Ole Jacobson's Internet Protocol Journal is already a great start in 
the right direction.)

While this level of formality may now be unnecessary in the US (and I 
emphasize the word "may") the world of patents is broader than the US 
and not all patent offices around the world may have become as willing 
as the USPTO to go beyond classic professional journals.  For the IETF 
and Interenet, in this particular area, it's best to look and act like 
an old time professional engineering or scientific body.

(I wonder, could one of these new LLC generative AI systems be trained 
on IETF materials, including e-mail of working groups, and be provided 
as a service to the community?  It would be nice if other bodies such as 
3GPP, W3C, and Chinese Next Gen IP, and other networking materials could 
be included, but that could raise copyright/licensing issues.)

     --karl--

On 2/3/24 11:12 AM, John Levine wrote:
> It appears that Karl Auerbach via Internet-history <karl at iwl.com> said:
>> Back in the day (whenever that was) the US Patent office only considered
>> filed (and perhaps only issued) patents and items published in a small
>> set of professional technical journals as representing "prior art".  (I
>> don't know whether that self-inflicted blindness still exists, I
>> certainly hope it does not.)
> I think you're confusing what the examiners use to search and what's prior art.
>
> You're right, the searches for software patents used to be terrible so
> they allowed patents on stuff that was widely used and not at all
> novel. I believe they use more sources now, what with Google and all.
>
> It's always been possible for a third party to challenge a patent,
> formerly via reexamination, now by inter partes review or post grant
> review. Anything published more than a year before the application was
> filed is prior art, and as Scott noted, I-D's count.
>
> They also allow Preissuance Submissions for applications that have
> been published but not allowed, to alert the examiner to prior art.
> IPR and PGR are very expensive, tens of thousands of dollars to apply,
> while PS are cheap, free for up to three documents and $180 for each
> 10 thereafter.
>
> R's,
> John
>
>
>