Today, I again met with a member (Matthew Cooperrider) of the Web4Roots meetup group.  Our goal was to pare down what a protocol for exchanging information about volunteering opportunities should look like.

First a quick summary of the problem:
1) Organizations need a standard way to syndicate volunteer opportunities they create.
2) Volunteers need an easy way to search for volunteer opportunities that match their location, schedule and interests.

During our first hack at this problem, we tried to treat volunteers and organizations as completely different entities.  This time our thought process was to see if we could model volunteers and organizations in the same way.  So what is fundamentally the same between volunteers and organizations?  Both have something to offer (volunteers: time, money, experience, etc.; organizations: tax right offs, a good time, free food, etc) and both have something they need (volunteers: tax write off, recommendation, satisfaction, etc; organizations: labor, money, supplies, etc).  So in effect, every time a person volunteers, he exchanges what he can give to the organization for what he wants from the organization.  The reverse is true for organizations.  So how can we model this?

Say we have a volunteer name John, who lives in New York City, enjoys cooking, is free on week nights, and wants to volunteer for 5 hours a month. Additionally, he is only willing to volunteer for an organization that can give him a $5000 tax write off. How would the protocol represent this character?

<exchange>
  <export>
    <type>Cooking</type>
    <location>Boston</location>
    <datetime>Weeknights</datetime>
    <quantity>5 hours per month</quantity
  </export>
  <import>
    <type>tax write off</type>
    <location>USA</location>
    <datetime>Now - Jan 1st 2010</datetime>
    <quantity>$5000</quantity>
  </import>
</exchange>

By the same token, say if we have an organization in New York City who is trying to redesign their website by March 1, 2009.  Furthermore, it wants a volunteer who is willing to commit 20 hours a month to this project.  In return, the organization is willing to write a letter or recommendation for the volunteer:

<exchange>
  <import>
    <type>Advanced PHP Web Develop</type>
    <location>New York City</location>
    <datetime>By March 1, 2009</datetime>
    <quantity>20 hours per Month</quantity
  </import>
  <export>
    <type>recommendation</type>
    <location>USA</location>
    <datetime>anytime</datetime>
    <quantity>1</quantity>
  </export>
</exchange>

Each “exchange” package can have 0 or more import and export blocks.  Furthermore, each of these packages would have to name the web service that generated the data (and a unique identifier within that service).  For example, if a client on Survv asked for help for a certain issue, Survv would create a unique URI for the volunteer opportunity before it syndicated the opportunity:

<unique id>http://survv.org/opportunities/19192</unique id>

Each “exchange” package could also have extra optional data such as keywords, tags, categories, and the name of the creator.

A few members of the Web4Roots meetup group and I met up last night to talk about volunteering websites on the web (such as volunteermatch.org, idealist.org,etc).  One problem that jumped out at us was that all of the sites have their own ways of gathering and displaying information about volunteers and volunteer opportunities.  Because of the lack of consistency, it is almost impossible for these sites to interact with each other and share resources.  We tried to imagine a protocol for sharing such information.

We started at a high level.  There are volunteers (or more generally resources) that have the ability to fill needs and volunteering opportunities (more generally projects) that have needs:

In general, there has to be a good match between the services that the resources can provide and the needs of the project.  We all decided, however, that this wasn’t enough.  There had to be an agreement between the goals/mission/purpose of the organization and the values/motivation of the resource.  For example, even though I have web development expertise and the McCain campaign is looking for a web developer, I may not want to volunteer for this position due to my political beliefs.  This thought process led to the following diagram:

So how might the above diagrams play out for a specific example (say cleaning up a park in the neighborhood):

Note: This isn’t meant to be XML, its just notes from the meeting.

<project>
      <needs>
        <need>
            <type>organizer</type>
            <quantity>1</quantity>
            <location>x,y,z</location>
            <datetime>March 11, 2008 2-5pm</datetime>
            <extras>
                <mandatory>
                    False
                </mandatory>
            </extras>
        </need>
        <need>
            <type>physical labor</type>
            <quantity>10</quantity>
            <location>x,y,z</location>
            <datetime>March 11, 2008 2-5pm</datetime>
            <extras>
                <mandatory>
                    True
                </mandatory>
            </extras>
        </need>
        <need>
            <type>Rake</type>
            <quantity>11</quantity>
            <location>x,y,z</location>
            <datetime>March 10, 2008 by 6pm</datetime>
            <extras>
                <mandatory>
                    True
                </mandatory>
            </extras>
        </need>
        <need>
            <type>Trash bags</type>
            <quantity>100</quantity>
            <location>x,y,z</location>
            <datetime>March 10, 2008 by 6pm</datetime>
            <extras>
                <mandatory>
                    True
                </mandatory>
            </extras>
        </need>
    </needs>
</project>

We actually never got to drawing out specifics for anything other than project needs.  Anyway, this was a short summary of the meeting.  We plan to have regular meetings on this topic, so I be sure to blog about any progress.

I tried installing Drupal on Mac OS X Leopard the other night and everything worked right out of the box.  There are the two things I had trouble with

1. The PHP on OS X does NOT come with the GD image library installed.  I found a very good tutorial to get it working here.  (The one problem with the tutorial is that the php.ini file that comes with the download did not seem to work on my machine.  Instead I used the php.ini file that comes with Mac OS X)
2. The .htaccess file that comes with Drupal seems to be missing a line that seems to break “clean urls” on a OS X.  Adding the rule “RewriteBase /path/to/drupal” seemed to fix things.

Great, now that I have Drupal cleanly installed, time to try experimenting …

Last night I went to a Net Squared meet-up at the Oxfam International office in New York City.   Attending the meeting were representatives from a number non-profits around the world (change.org, TheElders.org etc).  In retrospect, there was one theme that tied all of these organizations together: Drupal.  Joshua Wiese, the Online Director for TheElders.org, mentioned how his organization has needs for online mapping software.  When I approached him after the meeting and talked to him about Open Layers, the first question he asked me was if there was a Drupal plugin for it.  After the meeting I also had a long conversation with a board member from Community Mediation Services (CMS).  CMS is looking for volunteers to help revive their website.  What platform are they are planning to create this website on?  Surprise, surprise: Drupal.

After talking to a few co-workers about Drupal earlier this week, I am pretty sure why it is the non-profit framework of choice.  Drupal is trivial to setup and simple to configure, and has an enormous developer community.  Sure PHP takes its shots, but people seem to get a lot of things done with it.   Since I am a member of the non-profit tech community, many of my peers are using Drupal, and thus I think it is worth my while to learn some basics.  As I experiment more with the framework, I will be sure to blog about my experiences.

This weekend, I stumbled upon an interesting document published by the American Bar Association.  The document, titled Standards for Pro Bono Programs, gives a number of criteria that any institution offering pro bono legal care should strive to meet.  A number of points made in the document are relevant to any non-profit that provides community services. (When I say this, I have openplans.org in mind. Openplans.org is produced by my employer The Open Planning Project.)  I have picked some of the most relevant ones:
(more…)

Because I believe that it is much easier to make software with a specific client in mind, I have been trying to contact legal clinics around New York City. Ideally, I would like to go into a legal clinic, create an application for them so that they can better manage their clients and lawyer, and then use this knowledge to make Survv a better product. Law Help, a production of probono.net, has a pretty comprehensive list of legal aid clinics in New York City.  I hope to contact as many of these institutions as possible.

During my phone meeting with Probono.net I was pointed to The Internet Representation Project (IRP), a cool project that doing that some neat things up in Northern Michigan.  They have been successful at doing some of the things that Survv is trying to do: provide a web interface to allow clients ask questions to lawyers.  The premise of IRP is that clients answer a questionnaire to determine their eligibility for the program.  If a potential client is eligible, he or she can fill out a form asking a question to a lawyer.  Presumably, this starts a discussion thread between the two parties.  I was poking around the website and found a couple of interesting things:

1. I was looking through the eligibility requirements and apparently clients who have criminal or abortion issues are not eligible.  The same goes for clients who are not in the country legally.

   I am not sure why this is these requirements are in place.  Maybe the lawyers want to stay away form giving out too controversial advice? Maybe the liability of giving out advice in certain areas is too high.  Most probably, the IRP just doesn’t want people to give out information about criminal activity over the Internet.

2. The Project does not officially verify that participating clients meet the requirements (income, net worth, value of assets, etc) for pro-bono legal council.

   This is one of the major problems I am dealing with.  First, clients might be scared to give a website extremely detailed information.  However, without extremely personal information, it is hard to verify the eligibility of a client.  And second, even if clients are willing to give out detailed information, developing a web application to do this credit check could be tricky.  Relying on the honor system might work well for a website serving a small community in Northern Michigan, but any national system would require a more stringent verification process.

I never actually asked a legal question to any of the lawyers on the website, so I cannot give an analysis on the quality or expedience of the legal advice rendered.  However, I think the people who are behind IRP are doing some pretty revolutionary stuff.  I am going to keep my eyes peeled to see what kind of success they have.

A few weeks ago I went to the Google App Engine Hack-a-thon in New York City.  For those who don’t have the low down, Google App Engine is a framework set up by Google that allows people to host their web applications for free.  Google App Engine excited me for a number of reasons.  One, I don’t like nor am I good at system administration.  Two, I cannot afford monthly web-hosting fees.  Three, other organizations who may be interested in Survv might not be able to afford web-hosting fees.  Fourth, Google App Engine is scalable.  Whether we have one hundred or one million users, Google will host our application for free.  Finally, a Google App Engine application is easy to deploy.  If a legal clinic in Nebraska wanted to use Survv to manage their lawyers and clients, they would be able to set up the website with just a few mouse clicks.

Ian Bicking, one of my co-workers at The Open Planning Project, gave a presentation at work yesterday citing some of the same benefits if one were to use Google App Engine.  However, another co-worker, David Turner, expressed some of the privacy concerns with using something like Google to host a web application.  What would happen if the government issued a subpoena to Google asking for all clients on Survv?  What type of privacy guarantees could be offered to people who use Survv?  In recent years companies such as AT&T and Yahoo have been quick to give the government information about their customers.  Would Google do the same?

I will have come back to this at a later date and look more closely at Google’s history with the government and privacy before I can make any final conclusions.

I had the phone meeting with Probono.net today.  Very shortly into the meeting I realized is that it is extremely hard to present a website over the phone.  However, I tried my best.

The meeting was attended by Mark O’Brien, the director of Probono.net, and Liz Keith, a program manager.  In general the meeting went well.  Mark seemed especially excited about the idea.  These were some of the things that excited him:

1. The possibility of opening a line of communication directly between lawyers and clients is a new concept.  Traditionally, legal clinics are responsible for the matching process.  Unfortunately, these legal clinics also represent a huge bottleneck in the matching process.  Given their resources, they can only process a certain number of cases every year.
2. If there is a natural disaster, such as Katrina, where a large number of people need immediate legal council, a website such as Survv can help do the matching.
3. A website such as Survv might be able to match lawyers in a big city like New York City with clients who live in rural areas in the State of New York.  Often the biggest need for pro bono legal advice is located in small rural locations.

Liz and Mark also voiced a number of concerns:

1. How would the website go about vetting clients?  Often problems that clients think are legal aren’t really legal.  Furthermore, clients might be skeptical to give a website incriminating information about themselves.
2. How would we vet lawyers?  What happens if some quack lawyer signs up and starts giving bad advice to clients?  What happens if a lawyer on the defense side starts giving advice to a client who is filing the charges?  We would have to make sure there are no conflicts of interest.

Despite their concerns, they seemed to think that Survv was a good idea.  However, they didn’t think Probono.net would be a good organization to partner with.  Probono.net’s tentacles expand to over 20 states and it would be hard for them to experiment with something new.  Mark suggested that I partner with a smaller more agile organization.  He promised he would put me in touch with people who might be interested.

All in all, I think I got everything I wanted from the meeting: feedback on my idea and an introduction to more people in the industry.  I hope that Probono.net is able to find a legal clinic or legal aid organization to partner with me.

I am finally done getting the demo online!  If you go to http://survv.appspot.org right now you can see a demo of Survv.  There are many missing features and the design is very rough around the edges, but the entire work flow is there.  So go ahead and check it out.  If you want to check out the client workflow, this presentation can get you started:

Client Experience V0.1.pdf

If you want to check out the lawyer workflow, this presentation can get you started:

Lawyer Experience V0.1.pdf