...many of us don't want to read your blog posts on Obama and Clinton. Get bent.

EVT 2008 is on!

elections, news, friends, research, policy

The 2008 USENIX/ACCURATE Workshop on Electronic Voting Technology is open for registration with the full technical program posted here:

uite popular to the extent that the organizers and PC chairs this year, David Dill and Tadayoshi Kohno, worked to expand it to two days! I'll be moderating a panel and presenting the last chapter of my thesis, so be sure to come early if you plan on attending USENIX Security or if you're a die-hard elections geek!

Questions about the Voting Information Project

elections, standards, copyright, news, open source, friends, research, policy, usability, legal

Doug Chapin and Chris Backert pointed me to the new Voting Information Project (Chris entitled his post, ?Just About The Coolest Thing?). My initial reaction was much like the title of Chris? post: it did seem very cool. Not only that, but it?s funded/coordinated by Google, Inc., Pew?s Center on the States and the JEHT Foundation. This will be a wonderful resource when finished and will start to regularize structured data publication by local election jurisdictions. Many thanks and my gratitude to those behind VIP for their efforts and investment.

However, after looking a bit more deeply into the current content of the VIP site, I?m left with some serious questions. I hope we can work through some of these comments to best position VIP to fulfill its goals.

What?s the vision?

It?s hard to find a coherent vision for the project. They have a FAQ and a longer description of the project, but no mission, vision or roadmap and no indication of who (as in people, individuals) are participating (other than the developers listed on their Google Code page).

Right now the VIP appears to be a data standard (a markup language specification), a funding program for states to collect data in this format and a method of uploading such data to the VIP site. (In the longer description of the project, they say they?ll soon provide data feeds.)

It would be nice to know what else they have planned. One concern during an election year is a data collection overload; the VIP shouldn?t develop in a vacuum, which it appears to have over the past eight months. That is, much of this information is regularly made available by NGOs. Ballot information can be had at the state and local level via LWV?s smartvoter, for example. The Verified Voting Foundation has invested quite a bit of time and effort in producing a mapping of the types of voting technologies used at the local level (check out their Verifier tool). I can understand how this is a ?push? effort rather than a ?pull? effort---that is, the idea is for government entities to start providing this data instead of NGOs to have to collect it---but a more likely outcome for this project is that a few states provide feeds in 2008 and then stop maintaining this information when the attention dissipates.

(NB: The specification doesn?t have a name other than the ?Voting Information Project?s open format? (or, in various places, ?EVP XML? or ?EIP XML?). I?ll call it the ?VIP XML format?.)

Where?s the EML?

The VIP XML format specification document (via this very ugly URL), doesn?t mention the OASIS? Election Markup Language (EML). EML is an international standard markup language for election data that is currently in the process of becoming an ISO standard. It would be good if the VIP folks could say why they decided to make their own standard instead of reusing pieces of the EML standard. I suspect this is because the EML standard is general to more than just US elections and that it might not include elements that they need for their goals. However, it would make sense to see how EML could be changed to accommodate these needs rather than eschewing it altogether.

This is an important point: There is no mention on the VIP site that they?re working with voting system vendors. Vendors are planning on supporting data interchange in EML format and a number of us have made the case that the new US VVSG standard should require vendors to produce data in EML (see the text of ACCURATE?s VVSG comments). The VIP project would be much more useful if they also invested time in XSL transforms that would allow EML interoperability with the VIP XML format.

Where are the data entry tools, etc.?

Related to this last point, if the idea is to provide regularly-updated ?feeds? of elections-related information, there needs to be mechanisms for getting this information out of the tools that jurisdictions currently use, or providing data-entry ?wizards? that will walk local election officials---where most of this information will reside---through the process of creating a data feed.

Obviously, data ?feeds? are best and most useful when automatically created rather than manually updated. There isn?t any attention on the VIP site to the integration needed with existing elections information management tools and election management systems to best support data publication and dissemination such that the end-to-end connection is made from tools that contain this information to publication of the information via the VIP feeds mechanism.

What will encourage jurisdictions to enter/provide data?

In all seriousness, why would a jurisdiction want to provide this information? Unless election officials see a clear benefit from providing the data and that the process for doing so is easy (and hard to provide incorrect, incomplete, etc. data), I can?t imagine many jurisdictions will take the time to do this.

The VIP project seems to think that this is a problem of initial costs. To remedy this, they?re offering $20,000 per state to help with these fixed costs. I?m not convinced that?s adequate. It might make more sense if there was some sort of incentive offered for providing data, similar to a non-profit equivalent to Dan Tokaji and Thad Hall?s proposal for providing federal funding contingent upon completing a detailed election data survey instrument.

I don?t think financial incentives are absolutely necessary to get comprehensive, high-quality data. Instead, election officials need to see there is some sort of network effect that compels them to provide data. For example, if there was a social network-component or if, through using it, they got access to some tools that made their jobs easier (such as voter-oriented elections information portals).

What?s with the draconian terms?

Being not-quite-a-lawyer, I was really interested to see the VIP project?s Terms of Service. There are a few interesting, and possibly troubling, features of this ToS:

  • CC licensed specification: The specification document is licensed under a Creative Commons Attribution-ShareAlike 2.5 license. It appears that the text of recent Google and Microsoft specifications are so licensed but, I?d feel more comfortable with a license that recognizes that a specification is more than just text (see Creative Common?s Mike Linksvayer?s post ?What good is a CC licensed specification??).
  • Changes could be arbitrary: For a site that aims to provide data feeds, it?s a bit troubling that the ToS says anything can happen at any time without any notice. Some notion of a commitment to provide data that people are relying on would be a good thing.
  • It?s a contract: Unlike most ToSs, it explicitly says that the ToS is a ?contract? that we, as visitors, enter into by using the website. In the world of terms of use, terms of service, etc., it?s rare to see the word ?contract? used; usually ToS writers prefer less-specific language. I?m not sure if this is necessarily a problem and would be interested if any legal minds out there have an opinion on such adhesion contracts on the internet. (I just might not be aware of recent developments that advise using the more specific ?contract? language.)

My point here is that these terms seem somewhat draconian and reflect the desire of the backers to avoid legal messes and liability. But will these terms serve their goals of creating a interactive data community? I?m doubtful. The question of what terms should apply to ?open services? will have to wait for a future article.

What kind of standard is their spec?

Finally, the big kicker for me: This doesn?t appear to be an open standard. I would call it a disclosed standard or a limited open standard (after my paper on source code disclosure in voting systems... that I?m currently updating for my thesis publication): Only a few possible constituencies are able to provide input into the VIP XML standard and it can change arbitrarily at any point in time. If one of the goals is to support data publication in an end-to-end manner, it would be best to include users of the data as well as election management system developers (vendors) in development of the standard. As it stands now, it appears to be an effort entirely contained within Google and a set of unidentified state-level election officials.

I don?t want to sound too harsh; I think this is exactly the kind of endeavor we need more of. However, the overall feeling I get from the VIP is not as promising as it could be.

UPDATE [2008-05-28T12:26:47PDT]: Added some sugar to my criticism.

Some graduation pics...

SIMS, berkeley, family, friends, education, iSchool

Here are some pics taken at our commencement... a superb pic taken by Andrew Fiore (catch his album on Facebook):

Image of Joe speaking at commencement

A picture taken by Jens of me messin' around (link is to my flickr set):

Joe flashing a westside/westcoast sign at graduation

A picture taken by me with Yuri's camera... this is Yuri, Fredrik, danah and Jens (the other four PhD graduates):

Yuri Takhteyev, Fredrik Wallenberg, danah boyd and Jens Grossklags

There must be a god...

an image of the upcoming weather showing a very nice, cool, day on Saturday for graduation
Contact / Help. (cc) 2021 by Joseph Hall. blog software / hosting.
Design & icons by N.Design Studio. Skin by Tender Feelings / Evo Factory.
And a few words about the structure of the eye . Everyone " retina ". Especially often we hear it buy clomid online in the phrase " retinal detachment ." So what is the retina ? This - the front edge of the brain, the most distant from the brain part of the visual analyzer. The retina receives light first , processes and transforms light energy into irritation - a signal that encodes all the information about what the eye sees . The retina is very complex and in their structure and function . Its structure resembles the structure of the cerebral cortex. The shell of the retina is very thin - about 0.14 mm.