Eric Leland's blog

Is Amazon the next Salesforce?

Recently, Amazon released its SimpleDB project into "unlimited public beta". Several months I read about this intriguing bleeding edge database effort by Amazon on Larry Keyes blog here, who thought this was an innovation to keep an eye on. As Frank Bell puts it in his review, SimpleDB is a "web service that simulates a database". Since Salesforce is a very successful web service that merely uses a database, I became more intrigued about what this might mean for databases of the future.

I like SimpleDB's departure from the relational database model. Typically, relational databases have a spreadsheet model, with rows representing John Smith's whole contact record, columns representing his various contact fields (title, street, city, etc), and cells holding his actual contact information. SimpleDB provides the ability to put attributes at the cell level - in other words, more than one piece of data in each box. So John's professional title "Director" might have the available attributes "Development", "Finance" and "IT". Other titles might have different attributes. SimpleDB is a web service, so we developers will build databases, then "call" them using one of many programming languages, such as Ruby, PHP and several others, from wherever we decided to host our applications. Nice! A new way for developers to build complex applications and not get bogged down in optimizing table joins and other mind-numbing tasks. Is this web 2.0 for databases?

The cost model is intriguing too. Pricing is similar to data storage services, which makes sense, since unlike Salesforce, you are not really purchasing an application, just a database with web services. Costs are per "machine hour" used, data transfer, and data storage. The free plan offers 25 machine hours and 1 GB of data transfer and storage. With the free plan, your developer could build a system that calls the database 2 million times a month. For many of us, this would be plenty. Add 10 cents per additional gig of transfer and 25 cents per gig of storage per month. Add 14 cents per additional machine hour per month. Cheap as chips, assuming your developer does not charge an arm and a leg for the application that actually makes the database usable by us regular folk. And so many possibilities for integration!

Amazon built it, but will developers come? Salesforce offers a wealth of resources, including most critically an Application Exchange where we can benefit from the labors of others to expand our database features. Salesforce enjoys the contributions of many developers targeted at features in demand from the nonprofit sector. It will be interesting to see who climbs on board, and how this might play out for the nonprofit sector.

Can Software Help You Trust Your Volunteers?

In the wake of the world's natural and politically motivated disasters over the past several years, a lot of attention has been put on nonprofits and accounting for their actions. 9/11 brought us the Patriot Act, forcing many funders to make decisions about who may be a terrorist based on often wrong, confusing and generally inaccessible lists of names from multiple Federal agencies. Hurricane Katrina brought flashy stories of scam groups masquerading as well-meaning charities, taking donations for personal gain.

The Philanthropy Journal published a brief article recently about the online tool MyBackgroundCheck.com to help nonprofits screen their volunteers. With healthy skepticism, I started prowling around the site. In a past project, I spend time helping foundations use a background check software called Bridger Insight from Choicepoint. It promises to make sense of all the competing lists of bad people by compiling the information in one secure central system for clients. Similar to MyBackgroundCheck.com, basic information about people is entered into the system and matched to various records. In the case of Bridger, the sysetm churns out matches to evildooers based on complicated pattern matching algorithms to obsure lists of people and organizaitons, while the MyBackgroundCheck.com system looks for more standard public records of evildoing.

Although it was humorous to demonstrate the false positives of the software that would link, say, a local nonprofit fire station to the terrorist of the German Red Army Faction, what was more alarming are the implications of the system. There was some vague notion that foundations should do their diligence to screen people they were granting money to, but no real direction on how best to determine if someone is a terrorist or not. Some foundations took a safe, conservative approach. If someone was matched with a bad group in the software, then done - no money for you. After all, who knows exactly how to figure out if that person is or is not the real terrorist? If I set the system to find more matches, I can also be safe to get the bad guys, even if some good ones slip in there.

Then the news broke, ChoicePoint's systems were breached, and over 100,000 records were compromised. The evildooers in this case became ChoicePoint customers, and proceeded to compromise the system. So much for the secure central system, really glad their screening worked! Now there was the possibility that all those real positive and false positive matches were out there, not to mention a variety of personally identifiable information. I did not feel too great about training folks to use this software anymore.

I took a scant five minutes to start the process of checking my background through the MyBackgroundCheck system as employed by the Red Cross, so I could read the fine print about what I would agree to do. Its interesting - again all the things I agree to do is printed in legalese, small type, and is difficult to understand. I get the chance to agree before reading the disclosures - something we should not encourage as nonprofits. I grant permission for a variety of checks on my public records, which they specifically list, including "etc", to accommodate anything else they feel like checking. Maybe they would check by credit record? I wonder if my late rent payments might factor into my suitability to be a volunteer? I might not go through the whole process if it were more transparent up front what they were looking for. What exactly am I agreeing to provide?

The big problem with this software is the thorny issue of how you use it. Building trust is a two-way street. These tools are never magic bullets, and come with tremendous responsibilities to be good stewards of the information learned. While it is valuable, in many cases essential to screen the background of who we work with, we need to work extra hard to make the information about these systems and how they are used accessible to everyone involved.

Quality Time with Your Technology

When you find yourself frustrated that your technology tools just don't work, remember that its not only you that needs comforting and support. The folks with the Pew Internet and American Life Project released a study looking at why technology fails, and produced some interesting findings, including:
  1. Nearly half surveyed rely on someone to help them work their cellphones or internet.
  2. Over a third reported their laptops or computers did not work sometime in the past 12 months.
  3. Over a third contacted user support to try to fix their technology problems.
  4. Nearly half were discouraged over the amount of effort required to fix the problem.
I know I expect my technology to "just work". And sometimes its so beautifully perfect that it really does just work. But for the most part, tools are imperfect... We can blame the makers, and it might even be their fault, but we are still stuck with them. Most every tool I use requires my investment to get to know it a bit better, warts and all. Knowing in advanced what to do if it breaks really helps me cut down on frustration when it finally does happen. Some strategies I follow include:
  1. Court your technology: Get to know your new tool in smaller bits, to keep the excitement high, energy positive and project successful. Learn all the resources available to you to help use the tool. Try before you buy. Ask friend and peers what they think. After buying, apply the tool in stages to avoid being overwhelmed.
  2. Save the documentation: Whether your technology is hand built by you, arrives in a box or is just there when the internet is on, find whatever written documentation and contact information there is, print it and minimally throw it in a box under your desk. Luckily boxes do not require rebooting.
  3. Try using support in good times: When you just finished a really great day with your technology, take a moment to call support. Find out what it is like to get (or not get) service at a time less dangerous for your stress levels. Practice your questions before you call or email - see what kind of responses you get and what works best.
It's lucky if I follow all of these for every tool. But it is so rewarding when, during some kind of tech failure (such as my desktop) at the perfectly wrong time (such as yesterday during a web call), I realize that I wrote the website and various service agreement information on a piece of paper nearby.

Wonderful Tools for Web Developers

I love the initial stages of creating a great new website. The part after the planning, when we know the objectives and some of what the website users want, and before we start building the system. Usually I am sitting down over at a coffee shop with a nonprofit and my geeky developer friends, and we are furiously brainstorming on paper, old receipts, napkins - drawing lots of sketches of pages as we explore what is effective.

There are many terrific tools that make it really easy to bring the sketches to life online. I used to turn to Visio, Illustrator and Photoshop to get our thoughts into something beyond a sketch, but found these tools just too complex to whip out depictions of what we had discussed, and difficult to share and collaborate with the files created.

After my latest napkin sketch session, I discovered Twiddla. Although it lacks the energy and ambiance of the coffee shop and its napkins, it is a wonderful tool for groups to collaboratively view and sketch on top of web pages and other uploaded images. The people I invite do not need an account, and can quickly join an audio conference through the computer, without dealing with call-in codes and other login nonsense. We happily set about defacing websites and brainstorming design directions, and can do screen captures to preserve the digital graffiti for later use.

After twiddla-ing, I want to take all the great ideas and extend them in different ways. Visiting the same websites we defaced earlier using the FireFox browser, I use the ColorZilla and Firebug extensions to view and manipulate how the web page actually looks, as though I were editing the page. ColorZilla is a free extension to FireFox that gives you a simple point-and-click tool to discovering what the color is of anything on a web page, in RGB, hex and other codes. I can use Firebug to view all sorts of things about how a particular web page is coded, including its stylesheet, and edit this code right in the browser. While sometimes I wish I could change some of the terrible sites that are out there, Firebug only changes the site for you at the moment. I can copy the code for a color I like on one web page, then visit another and change the stylesheet to use that new color, and see right away how it looks online. If I messed up the page, no worries, just reload and it all comes back to normal. Great stuff!

I love how Gliffy takes Visio and strips it of almost everything complex, leaving a simple flowcharting tool that just works. I can draw lines linking boxes and triangles with ease, and label quickly, and when done, share this with the whole team. I can also use the available templates to start with a web page frame, and adjust the columns, drop in boxes representing features we discussed and have a variety of sketches ready in no time.

Mapping Blues: Where is the Data?

Online maps are great. After working on this mapping report, it became clear that half the battle with mapping is finding the right data. Recently I was asked where nonprofits can find data about their community. "There are so many sources," I sputtered, before realizing I did not know many of them by name. "Which ones have you used?" she asked. In a Palin moment, I said, "Well many of them, I will get back to you with some examples." Ugh...

While you can buy data, there is so much public data out there to check out first. Resources for data can be broadly applicable, or very specific for narrow uses. Here are some resources I have found useful for finding data, often for free, to use in my mapping projects:

U.S Census: Yes, most everything that the government collects in those giant census surveys is available to download. Tons of demographic information, boundary files for outlining states, counties and cities, and many other resources for free.

Data360: Lots of great data sets, slideshows and graphs contributed by users. Its helpful to click through graphs and slideshows that are close to what you are looking to map to see how others approached the problem. As an occasional surfer, I find this contributed graph of shark attacks somewhat comforting.

UNdata: Tons of international data covering , demographic, health, trade, education information and more - a particularly good resource for international development organizations. The site provides multiple options for filtering data, showing/hiding columns, and for exporting to work in multiple mapping/GIS software.

Freebase: Another resource for data contributed by users. I find this site easy to navigate, and has some really interesting data. All contributed data can be associated with topics, grouped together by types, which in turn are grouped together by domain. This structure allows you to find, say, data on the topic Dick Cheney as well as on the type U.S. Vice President - providing various perspectives on related information.

Huge Data Lists: Not for the faint of heart, these are goldmines for finding public data resources, but you will have to dig. Check out the publicdata tag in delicious, or this giant list from Peter Skomoroch.

I would love to hear about your data and mapping projects, and where you found useful information for them.

Web Meets World

The folks over at O'Reilly, who brought us the term "Web 2.0" and lots of great software books for geeks, are having a benefit auction in conjunction with their annual Web 2.0 Summit this November in San Francisco. Themed "Web Meets World", they are soliciting donations and donation ideas, along with recommendations for charities that should benefit, all through their Facebook "Web 2.0 Summit" community.

I like the focus this year, which is looking at how web 2.0 can be used for solving pressing world problems. While I like my Apple iPhone, and love how it can emulate a cigarette lighter, its these kind of silly applications can really bring you down when you turn attention back to global warming... is this all people can think of? The Netsquared folks over at TechSoup have done a great job supporting web 2.0 initiatives for social good, I hope that the influential folks at O'Reilly can contribute some more momentum here.

One of my favorite organizations, Witness is reported to be one charity that will benefit from the auction. They provide video cameras to human rights groups, and technical assistance to bring their messages out. I used to work for Witness over 10 years ago when it was three people. It has grown far since then, and is a terrific program and resource for rights groups worldwide.

Really Simple Databases

It just seems that too many databases are too complex to use. How many years have we had to develop simple databases, that just work? Most groups I work with can quickly list off the information they truly care about in their systems. Yet only a fraction of these groups manage to get it out of their systems in a way that does not induce fits of anger at the idiocy of the routines required to get it done.

I had the distinct pleasure recently of working again on a Kintera driven website. Kintera is a very comprehensive solution, and has lots of levers and dials to flip and turn to help drive it. My head begins to hurt as soon as I log in. It took me twelve clicks to navigate through the system to edit the first web page, and five clicks for subsequent pages. There are lots of terms, menus and buzzwords that remind me of those signposts with the 26 arrows pointing all different directions to different places. I don't mean to pick on Kintera - lots of systems are similarly maddening.

So where are the easy systems? There are no absolutes here, so much depends on who is using it, and for what purpose. However, some systems are just plain easy. Here are a few appealing examples:

Bento (filemaker.com/bento): From Filemaker, this is a personal database tool that really appeals to the side of me that does not want to work for a living. I have barely touched the demo for five minutes, and I had a simple donor management system ready to go using one of their many preset templates. Their latest version works well for folks moving from Excel, and ties into address books and mobile handhelds nicely. Of course, simple has its price. Its Macintosh only, and only one person can use it at a time, and you can't go crazy building complex relationships between your data.

DabbleDB (dabbledb.com): I created a demo account, and had a simple database of all my favorite chocolate shake places in the bay area created in 8 minutes and 38 seconds. That was just a little bit longer than their demo video, which very nicely explains the basic features. This is what someone should have done a long time ago, for folks just looking for basic contact management, accessible anywhere you and the web are. Apparently you can get a bit crazy developing charts, pivot tables and maps, but I expect this will take a fair bit longer than 8 minutes to work out, and I need some dessert.

Blogger (blogger.com) and TypePad (typepad.com): Two great programs for starting a blog or small website. This wonderful Idealware blog uses blogger, and makes posting a snap. While it did take me about 1 hour and 30 minutes to set up my first website using TypePad, most of this time was spent debating over the free templates I could use to make it pretty. Both systems remove any knowledge of web code and most technical jargon so you can get to work getting a website up and running, and change content quickly and easily.

What are your favorite, super-simple web & database systems?

Managing Documents on Your Macintosh

It's taken me two years to convert, but now I love my Macintosh. Many of the nonprofits I work with do too. However there is no getting around the fact that there is only a limited set of options for business software that runs on a Macintosh. While I would love to have all my nonprofit friends upgrade to the fancy new Mac laptops that will run Windows too, the cost and performance is still prohibitive.

I wanted to share one struggle of mine to find appropriate business software that runs on Macintosh systems. Recently I was challenged to look for a document management solution for a Mac based office. They have approximately 250,000 documents currently being imaged to PDF format. The main criteria are:

(1) $2000 budget
(2) 6 users
(3) Must handle about 8GB of PDF documents without crashing
(4) Can store some basic document descriptive information - such as Title, Author, Category.
(5) Should provide some kind of preview feature when browsing documents

Focusing on hosted solutions - I looked at freely available online systems such as Magnolia CMS (magnolia.info) and Alfresco (alfresco.com) - they could definitely do it, but I needed a brilliant and horribly underpaid software engineer to build it for me within budget. I looked as some online file storage solutions, such as Box.net, Xythos (xythos.com) and DocumenTree (documentree.com), but these focused on storing files in folders and provided very limited ability to store basic document descriptive information.

Eventually I turned to locally installed systems. FileMaker was out - the license costs + server software alone would break the budget. I found a raft of locally installed systems recommended by various companies who do document scanning, but all of these were PC only and most were quite expensive.

Do I just need to admit that my requirements are unrealistic, or are there more options I am missing?

So far I have found a few options that come close, but I am still looking for other leads. The best match so far is DEVONthink (devon-technologies.com). These are the same folks that built EasyFind - a very cool Macintosh OSX file search tool. DEVONthink can handle the larger amount of files, offers a preview function, and some basic document description data and a promise to do more on this point. I should be able to pick up licenses for about $900. Another possibility was Papers (mekentosj.com/papers/) - this is more for researchers who are gathering articles and other publications from online resources, but does has features that come close to what I am looking for.
Syndicate content