Re: San Francisco debate heating up

From: Brian Behlendorf <brian_at_behlendorf_dot_com>
Date: Thu Mar 01 2007 - 21:44:03 CST

On Thu, 1 Mar 2007, Alan Dechert wrote:
> That's a good question. It probably could have been toned-down a notch or
> two. Even Brent Turner remarked about the ! marks. I went back and checked
> .... I counted 4 ! marks ... still, that's only .5 per page!
> I used to collaborate quite a bit with Doug Jones, one of 3 OVC founders.
> Some of his letters had several ! marks per paragraph. So, maybe I can
> blame it on him.

I guess it wasn't so much the !'s as the length and tone. I write as I
would speak to someone in person, and while some of his statements might
be deeply misinformed, a tone like that isn't going to bring him around,

> On the other hand, Arntz delivered a couple of low blows, and this
> absolutely required a response from me. He suggested that I wanted to
> replicate the Sequoia system for my own purposes,

I did see the insinuation that the only reason we are advocating Open
Source is because we want to enable someone close to us to copy Sequoia's
oh-so-precious IP for a competitive offering. But, to be honest, that is
exactly what is legally possible with an Open Source license, and is one
of the big differences between OS and simple disclosure, and until there
is a certified alternative ready to go, there won't be much of a credible
come-back to that claim. The logical links don't hold up - we want voting
systems to be open source, because a true open source community creates
more reliable code - yet none of us see Sequoia's code as forming a
reasonable basis for such an collaborative community, so forcing Sequoia
to open source their code feels like a non-sequitor.

> and he twisted Jim Soper's language and propped it up along side a
> non-existent group ("count as cast") to counter OVC.

That did seem odd, I don't have enough context to judge intent or weight,
but I do suggest that if the reply assumed that it was a mistake rather
than an intentional mischaracterization, it's more likely to be corrected.

> I get no joy from flaming someone. I wish this whole experience in the past
> 6 years had been all love and understanding. It just doesn't work that way.
> I wish we had been able to work with McPherson. I mean, read this letter:

Why go into McPherson at all in the response to Arntz? If you feel that
Arntz hasn't enough knowlege of the history of open source voting in
California, I'd suggest writing a document on that history up on the site
and providing him the URL; including it in the letter, especially if he's
likely to already know, just feels like lecturing. If I received a letter
like that I would be unlikely to reply, let alone contact you in the
future, and Arntz is under no obligation to do either.

> That's all I could do. It wasn't pretty, and it sure wasn't going to win
> him over to our view. We were out to defeat him.

Does the OVC have the time and energy to work against individuals who are
against its goals? Or should it instead focus on supporting those who
favor them? I'm sure it was frustrating to have McPherson saying
supportive things and then turn 180 degrees, but once that's done, what's
the point of making enemies? If your reputation was abused, make a public
clarification via a blog or something, stay on good graces as tides can
shift again, and then route around them or find other better

> Arntz made the outrageous suggestion that I wanted the ability to replicate
> the Sequoia system so I could use it commercially. I showed why this was
> wrong by showing what OVC has done, what OVC is doing, and what OVC is
> trying to do. If the reader gets that, the Arntz suggestion is disproven.
> That was a little windy, I admit. I did what I had to do.

It would not surprise me to hear uninformed or barely-informed people
confusing OVC and OVS, and surmising that since OVS's solution is not yet
certified, that OVS could consider using Sequoia's code commercially. In
addition to correcting that mistaken assumption, let's try to understand
how people arrived at that, so as to avoid it in the future, since we
might need the support or at least consent of uninformed people in the

What we do not want, I don't think, is the cause of Open Source voting to
be cast as an effort to unfairly advantage one specific commercial outfit
over another. Otherwise, OVC's credibility will be shot, and it'll be all
too easy for a politician or beaurocrat to assume the worst and aim to
split the difference or at least take the easy answer.

> Actually, amending the ES&S contract is much better for our cause for a
> variety of reasons. For one thing, there is a better chance of getting
> a new RFP specifying open source if SF gets the ES&S contract extended.

Then I would suggest petitioning Arntz for that outcome, arguing why it's
worth the extra cost over Sequoia to keep that flexibility. Or,
suggesting a shorter time frame for Sequoia, so that as open source voting
solutions become viable, certified, and vendors start responding to RFPs
with such systems, a transition can be made.

>> To be honest, I really don't want to see Sequoia's current code
>> released as open source. Unlike the OVS codebase or other codebases,
>> it was not written with public disclosure in mind, and has no shot of
>> attracting a development community that would help fix bugs, a critical
>> counterbalance to the claim releasing code will just make it easier to
>> hack.
> There is something at work here very different from your experience with
> Apache. Importantly, it is the public's right-to-know factor. That,
> together with the regulatory barriers, makes Open Voting a lot different
> project than your average open source project. The techniques involved to
> get our points across are also necessarily different.

I have a great deal of sympathy for the idea of "right to know", but it's
nowhere near as important IMHO as voter-verified paper ballots, regular
audits, and the other topics that the Holt bill has brought up. IMHO
OVC should be petitioning SF to procure solutions against those criteria
more strongly than petitioning for one vendor's source code.

I would also sooner petition for RFP/procurement language that gives a
preference (rather than a mandate) for open source solutions sooner than I
would petition for full disclosure. Why? Because I will agree with
claims that disclosure of source code, without creation of an active
developer community and ongoing peer review and improvement, and without
auditing/VVPB/etc, will make it much, much easier to steal elections.


OVC-discuss mailing list
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
Received on Sat Mar 31 23:17:03 2007

This archive was generated by hypermail 2.1.8 : Sat Mar 31 2007 - 23:17:09 CDT