RE: Global snapshots - Mailing list pgsql-hackers

Hi Andrey san, all,

From: tsunakawa.takay@fujitsu.com <tsunakawa.takay@fujitsu.com>
> And please wait.  As below, the patent holder just says that Clock-SI is not
> based on the patent and an independent development.  He doesn't say
> Clock-SI does not overlap with the patent or implementing Clock-SI does not
> infringe on the patent.  Rather, he suggests that Clock-SI has many
> similarities and thus those may match the claims of the patent
> (unintentionally?)  I felt this is a sign of risking infringement.
> 
> "The answer to your question is: No, Clock-SI is not based on the patent - it
> was an entirely independent development. The two approaches are similar in
> the sense that there is no global clock, the commit time of a distributed
> transaction is the same in every partition where it modified data, and a
> transaction gets it snapshot timestamp from a local clock. The difference is
> whether a distributed transaction gets its commit timestamp before or after the
> prepare phase in 2PC."
> 
> The timeline of events also worries me.  It seems unnatural to consider that
> Clock-SI and the patent are independent.
> 
>     2010/6 - 2010/9  One Clock-SI author worked for Microsoft Research as
> an research intern
>     2010/10  Microsoft filed the patent
>     2011/9 - 2011/12  The same Clock-SI author worked for Microsoft
> Research as an research intern
>     2013  The same author moved to EPFL and published the Clock-SI paper
> with another author who has worked for Microsoft Research since then.
> 
> So, could you give your opinion whether we can use Clock-SI without
> overlapping with the patent claims?  I also will try to check and see, so that I
> can understand your technical analysis.
> 
> And I've just noticed that I got in touch with another author of Clock-SI via SNS,
> and sent an inquiry to him.  I'll report again when I have a reply.

I got a reply from the main author of the Clock-SI paper:

[Reply from the Clock-SI author Jiaqing Du]
--------------------------------------------------
Thanks for reaching out.

I actually did not know that Microsoft wrote a patent which is  similar to the ideas in my paper. I worked there as an
intern.My Clock-SI paper was done at my school (EPFL) after my internships at Microsoft. The paper was very loosely
relatedto my internship project at Microsoft. In a sense, the internship project at Microsoft inspired me to work on
Clock-SIafter I finished the internship. As you see in the paper, my coauthor, who is my internship host, is also from
Microsoft,but interestingly he is not on the patent :)
 

Cheers,
Jiaqing
--------------------------------------------------


Unfortunately, he also did not assert that Clock-SI does not infringe on the patent.  Rather, worrying words are mixed:
"similarto my ideas", "loosely related", "inspired".
 

Also, his internship host is the co-author of the Clock-SI paper.  That person should be Sameh Elnikety, who has been
workingfor Microsoft Research.  I also asked him about the same question, but he has been silent for about 10 days.
 

When I had a quick look, the patent appeared to be broader than Clock-SI, and Clock-SI is a concrete application of the
patent. This is just my guess, but Sameh Elnikety had known the patent and set an internship theme at Microsoft or the
researchsubject at EPFL based on it, whether he was aware or not.
 

As of now, it seems that the Clock-SI needs to be evaluated against the patent claims by two or more persons -- one
fromsomeone who knows Clock-SI well and implemented it for Postgres (Andrey-san?), and someone else who shares little
benefitwith the former person and can see it objectively.
 


Regards
Takayuki Tsunakawa