"Count what is countable. Measure what is measureable. What is not measureable, make measureable." -- Galileo

Monday, December 31, 2007

Plone Metrics Person of the Year

Ah, yes! The moment you've all been waiting for -- the announcement of this blog's Plone Metrics Person of the Year. With a distinct honorable mentions for Alan Runyan and Vidar Andersen, co-founders of Plone, this year's award, the first ever, goes to Alexander Limi for his tireless efforts both in front and behind the scenes.

The Plone Reference Library and Amazon Statistics

I was poking around in Amazon.com recently and realized that it had some numbers that would be worth kicking around. Plone holds up well in the "number of books about it" category (6) for open-source CMS, although Drupal comes close with 4 and SharePoint blows everyone away with 48.

In terms of Amazon sales rank, Aspeli's book is doing well at 40,410. By comparison, SharePoint for Dummies comes in at 17,382 and Pro Drupal Development at an astonishing 6,168. The full set of data from Amazon, looking at number of reviews, customer ratings, and sales rank are:

Title Author Sales Rank
Professional Plone Development Aspeli 40,410
Building Websites with Plone Cooper 93,139
Definitive Guide to Plone McKay 139,560
Content Management with Plone: Handbook for Authors and Editors Lotze,Theune 213,164
Plone Live Pelletier, Shariff 455,539
Plone Content Management Essentials Meloni 584,584

Title Reviews Rating
Professional Plone Development 5 5.0
Building Websites with Plone 7 3.5
Definitive Guide to Plone 15 4.0
Content Management with Plone: Handbook for Authors and Editors 0 ---
Plone Live 5 5.0
Plone Content Management Essentials 8 3.5

By way of comparison, here are the numbers for Drupal:

Title Author Sales Rank
Pro Drupal Development VanDyk, Westgate 6,168
Drupal: Creating Blogs… Websites Mercer 34,996
Building Online Communities Douglass, et al. 138,071
Practical Drupal Mansfield 534,402

Title Reviews Rating
Pro Drupal Development 36 5.0
Drupal: Creating Blogs… Websites 16 3.5
Building Online Communities 26 3.5
Practical Drupal 0

Interestingly, sales rankings bounce around hourly. "Pro Drupal Development" had a 3,648 ranking last night and this morning it had fallen to 6,168. A couple Plone titles changed places overnight.

Amazon restricts raw sales data to authors and publishers only, so what are we to do with these mysterious "sales rank" numbers? I suggest a figure of merit (FOM) that is one over the sales rank (larger sales rank number means lower sales). One finds that the sum of all Plone sales rank FOMs is 1/6 that of the Drupal total sales rank FOM. This may mean that roughly six times as many Drupal books are being sold as Plone books on Amazon. I should mention that 81% of Drupal's figure comes from a single title, "Pro Drupal Development."

But some caveats are due. "Definitive Guide to Plone" is available free on the Web and "Plone Live" is best obtained by online subscription. These would certainly depress the Amazon numbers. Also, one should take into account how much online documentation is available--better free documentation would mean less need for that reference book on the shelf.

One last statistical pernambulation: take the number of reviews and multiply it times the average rating in number of stars. Plone totals 162.5 and Drupal comes in with 327. Over half of Drupal's total comes from VanDyk and Westgate's book

The oddest statistics were the category rankings that Amazon had. Meloni's book is the last place Plone title and yet it ranks #2 in Content Management. Meanwhile, "Practical Drupal" clocks in a #11 even though it has a slightly higher overall sales rating.

What, if anything, can the Plone community do about this? First, support your authors--buy their books. Second, review their books, rank them, and perhaps most importantly, tag them so that they pop up high in the category searches, especially "Content Management."

Amazon is a wealth of objective data. I'll come back to this data set and look at other CMS at a later date as well as reviewing how Plone texts are moving in the statistics.

Have a Happy and Prosperous New Year!

Thursday, December 27, 2007

Plone Metrics Person of the Year

Well, I've survived Christmas and the New Year is fast bearing down on us. As it seems with all other endeavors this time of the year, there needs to be a "Person of the Year" identified. Of course with this Plone Metrics blog, that esteemed personage will have with it all the honor, glory pertaining thereto. Needless-to-say, there's no money in it ;-))

My nominees are:
  • Alexander Limi, founder and continuing champion
  • Jarn, for selflessly removing the word Plone from their original "Plone Solutions" name
  • Joel Burton, for continuing the bootcamp vision and for maintaining uml.joelburton.com
  • Jon Stahl, for his work with the non-profit community
  • Martin Aspeli, for getting the first Plone 3.0 book out the door
The Award Committee (me) will now evaluate the nominees and make a pronouncement by 1 January 2008. Please comment on your opinion of the person or group who has most effected Plone (positively or negatively) in 2007.

Sunday, December 23, 2007

A Very Plone Christmas to All!

Here's wishing everyone a Merry Merry and a Happy Happy!!

And to all a good night!

Sunday, December 16, 2007

Internet Mensuration in General

I'm not the only one struggling with measuring aspects of Web usage. I ran across this quote in a piece referenced by Yahoo!News in which advertising executives were wondering why ad spending isn't higher considering how much time people spend online.

"This industry looks like it can't get out of its own way," said Steve Wadsworth, president of The Walt Disney Co.'s Internet group. "We need measurement of the audience and their use of the system that's clear, simple and actionable for a marketer. You need comparability with other media."

As Internet executives hash over clickstreams, page views and user panels, 2008 is sure to see even more evolution of the way online audiences are measured.
More to come....

Open Source Alliance in the News

Today there's a bit in the news from OSA. Here's a quote from the article in Yahoo!News, which is just a repeat of the material in PC World:

Participants cited a variety of interoperability scenarios and concerns, including single sign-on for identity management.

Another involved user interface interoperability. "A lot of people try to plug open-source projects into an enterprise portal, and they want a unified look and feel," Sartorio said.

Customers also cited cross-platform portability and data integration challenges.

In addition, the study found that:

-- Lower up-front cost was the most important driver for adoption of open-source products, but this was tempered by concerns that spending on support and services would be greater. That feeling in turn was mitigated by a belief among respondents that open-source products will become more mature over time and easier to support.

-- Few customers cited the ability to customize source code as a selling point for open-source applications, preferring instead that it handle their needs out of the box.

Its particularly odd to me that customers want out-of-the-box solutions. While I'll admit I'm always happy when I find a product that does exactly what I need Plone to do, more often than not, we're off to UML and ArchGenXML for a custom component. One size does not fit all.

Anyone interested in this should track down the OSA material at http://tinyurl.com/2884eh. There OSA states:

The top criteria were (1) total cost of ownership, (2) whether the solution meets requirements and solve business problems, (3) the vendor’s ability to support them and (4) the ability to interoperate with their existing environment.
Customers were generally satisfied with the first two items, which should have been a no-brainer -- open-source by definition is lower cost (unless someone is cooking the books on total life-cycle costs) and requirements-driven criteria should be a certainty. Vendor support and interoperability come in 3rd and 4th, but appear to be the discriminators.

The Plone community and support we'll save for another post and today I'll concentrate on interoperability, since that is the MSM's big thrust. OSA identifies six issues with interoperability:

  1. Single signon -- done.
  2. Data integration -- need to research deeper into the report on this topic.
  3. Portability -- done.
  4. UI customization and portal integration -- no problem.
  5. Content management integration -- I'd love to see an easier way to get two-way synchronization between external RDBMS and Archetypes.
  6. Component compatibility -- one of the areas where Plone 3rd party developers excel but could still improve.
Plenty more good stuff in just the 11-page summary report, but that will have to wait for another day.

Friday, December 14, 2007

Native Heart

I've been corresponding with Frank Odasz of Lone Eagles Consulting since I met him at last summer's Digital Pathways Conference and he's had some excellent ideas about using CMS to engage Native Americans. These same concepts translate well to my work with women's empowerment and economic development in rural areas of the Middle East.

Since then I've contacted Openia and they've provided a free Plone site for us. (Thanks, Openia. You guys rock!) We're just getting set up, but the plan is to leverage existing systems like YouTube, LinkedIn, Monster, and Flickr with Plone as the hub. Right now its a prototype running on the good wishes of Openia, but we're on the prowl for a sponsor who will cover the bill for a monthly .org domain and full-up Plone 3.0 site where we can add the products we need (for example, Plone4Artists) and have space for the multimedia content we hope to attract. Stay tuna'd...

Meanwhile, one of the critical needs for pushing Plone forward are low-cost and free hosting services. Also, a how-to would be extraordinarily helpful on setting up a personal server on a home computer using, say, a standard Comcast connection, some simple scripts to handle IP addressing issues, and services like no-ip.com.

Thursday, December 13, 2007

OpenLogic Census

Without a doubt OS developers should be elated with the recent news item that the OpenLogic is attempting a global OS software census. Check it out and be sure to take part when the census opens early next year.

Meanwhile, I'm happy to report that Plone WorldKit, even in an alpha release, is working fine on a production machine. This weekend I'll try to get a couple other Plone GIS systems installed and running, but for now, the gauntlet has been thrown.

Finally, in my College of Santa Fe Computer-Math Science class--Analysis of Algorithms--I've begun exploring a Kalman Filter implementation that I may be able to use with Bullard's measures of effectiveness. Its based on a pre-publication copy of an article by Tom Arnold, but it looks feasible. Stay tuna'd...

Sunday, December 9, 2007

Plone Downloads, Parte Duo

Looking at the download stats on SourceForge shows a disturbing downward trend, but the truth of the matter is that the download link in Plone.org now points to LaunchPad.net, where statistics apparently are not made public. My thought is that LaunchPad downloads are siphoning away stats from SourceForge, so we have an incomplete picture of the Plone hits. I'll have to turn to the Plone Foundations guys to get the numbers from the Plone.org server logs. Just goes to show that there are apples and oranges things happening all the time in the software world.

Saturday, December 8, 2007

SourceForge Downloads

Its a windy Saturday morning in Albuquerque and I'm about to head out into it for a bonsai workshop with Michael Hagedorn. But before I do that, I'm waiting on the coffee pot to finish brewing and that gives me a couple moments to hop over to SourceForge and see how the download stats are doing.

The graph at right shows calendar year 2007 (ignore December since its an incomplete month). Obviously, since the high watermark last Feb-Mar, things have tapered off. In fact since the rollout of 3.0, downloads have been quite weak.

Does this mean that Plone is falling out of favor? I doubt it, but coffee's done, so my explanation will have to wait until this evening.

Thursday, December 6, 2007

Measurement and Uncertainty

Today I came across Mike Cassady's piece entitled "Does Lack of Measurement Mean Uncertainty?" on a Sandia internal website (sorry, no public URL). He began by posing that logical positivism rests on an assumption that "to measure is to know; not to measure is to remain uncertain." He suggests that they rely on the fact that "We are comfortable using numbers as an assessment tool.... but we are suspicious of any assessment approach that doesn't rely on objective data computed numerically."

But then, unexpectedly for a place full of engineers and scientists, Mike goes on to say "If you can't measure it numerically, it's probably ten times more important than anything you CAN measure numerically." He goes on saying, "customer sets often live in the gray area of 'customer experience' rather than the firm scientific feather bed of 'measure is meaning' [and] that [programmers] are going to have to broaden their receptivity to qualitative and judgment-oriented measurements."

Hmm? Where does that leave us?

Let me make another observation about a phenomenon called an "information cascade." Here's the nitty gritty from Wikipedia:

Definition: An information cascade is a situation in which every subsequent actor, based on the observations of others, makes the same choice independent of his/her private signal.

Erroneous Mass Behavior: In an information cascades everyone is individually acting rationally. Still, even if all participants as a collective have overwhelming information in favor of the correct action, each and every participant may take the wrong action. The probability that everyone is taking the wrong action is less than 50%, but it is easy to construct examples in which everyone is wrong with 30-40% probability.

Fragility: A little bit of public information (or an unusual signal) can overturn a long-standing informational cascades. That is, even though a million people may have chosen one action, seemingly little information can induce the next million people to choose the opposite action. Fragility is an integral component of the Informational cascades theory!

So, what does this mean for society? Cascades predict that you can get massive social imitation, occasionally leading everyone (the "herd") to the incorrect choice. (Because everyone knows that there is very little information in a cascade, cascades are "fragile"; a little bit of new public information can make a big difference).

For someone trying to decide on the best CMS to implement, they might just look around and see what others (the "herd") have done and make a sub-optimal choice. From this I conclude that we most definitely need to temper judgement-oriented decisions with cold hard objective knowledge. At the very least, we need enough rigor to ensure that an information cascade doesn't occur.

Wednesday, December 5, 2007

Rhythm and Business

Norman Kelley's book is entitled "Rhythm and Business" after the Rick James quote, “R&B stands for rhythm and business.” So I'm taking a moment to look at the business side of CMS, especially Plone. One of my measures of effectiveness is the economic health of 3rd party companies.

A quick visit to plone.net shows 225 companies listed in 47 countries. Drupal is covered by Drupal Yellow Pages and it has a nice Google map interface. They have 50 map pins. Searching for "alfresco companies" on the other hand turns up (in the top 30 search results) no unified list of active 3rd party companies. Joomla has a Joomla Yellow Pages with 203 companies identified.

It looks like the number of 3rd party companies doing business around an open-source project is a good proxy for the economic health of that product's business environment. The Plone Foundation should be applauded for their proactive business development on behalf of the Plone community.

In conjunction with this discussion, I should mention another one of Matt Asay's pieces on open source--this one dealing with the Red Hat business model. Its worth a read for anyone working in open source.

And finally, if I still have your attention, here's a chunk pulled right out of a recent posting on Jeff Pott's blog:

Here are the open source projects/companies that made eContent Mag’s eContent 100 this year:

  • Alfresco
  • Automattic (Wordpress)
  • Drupal
  • eZ Systems
  • JA-SIG
  • Liferay
  • Plone
  • Tiki Wiki
  • Typo3
Plone is on the list for the second consecutive year. eContent only produces an alphabetical listing, so the only number we can pull from their data is "number of years listed." More on this in a subsequent write up as I data mine their listings.

Monday, December 3, 2007

Yet Another Plone Portal

Another call; another request for a portal. About four hours later, Debbie has crafted another Plone site, albeit with a default skin, but with workflow, security, and folder hierarchy are all in place.

Now we see that with just a few new portal requests, our regression line has become a 3rd order polynomial. Its R-squared is over 99%, meaning that the curve fits the data extremely well. I can now revise my forecast of Plone sites for calendar year 2008 -- we should be at 50(!) by a year from now.

Saturday, December 1, 2007

Plone Products

I recently found a year-old posting on Jon Stahl's blog about the need for objective product reviews. Turns out, I have an excuse to start doing that.

Since I have a customer at work who has a need for a map-centric view of facility data, I'm looking at the various products for GIS integration with Plone. Here's a quick rundown of what's available off-the-shelf at plone.org for geospatial data presentation.

Product Version Most Recent Release Date Days Since Last Update Status
Maps 1.1 9/5/2007 87 Active
PloneWorldKit 0.2 5/5/2007 210 Torpid
Plone Place Map 0.10 2/21/2007 283 Torpid
Plone Google Maps 0.2 9/28/2006 429 Moribund
AT Google Maps 0.5 7/19/2006 500 Moribund
GeoLocation 0.3 6/9/2006 540 Moribund
MapLocation 0.1b41r3 3/13/2006 628 Moribund
PrimaGIS 0.5.1 2/18/2006 651 Moribund

I've defined 'torpid' as a product without a major stable release and no activity within the last six months. 'Moribund' here means no major stable release and no activity for over a year. As I work my way through actual installation and use, I'll have genuine, objective remarks to make.

I may add Pleiades Geocoder and OpenLayers after I take a look at them. PrimaGIS may get elevated--if I can get it installed on my Windows testbed (sheesh, what a mess of dependencies). I saw the demo/tutorial at World Plone 2006 in Seattle, but have never gotten it to run for myself.

I'm tending not to favor the products that use the GoogleEarth api at first blush. The project is an internal application and so would need the Enterprise version of GoogleEarth if we went that way. Right now I'm thinking PloneWorldKit or PrimaGIS would do the trick.

PloneWorldKit works fine on our 2.5.3 production server, but doesn't install on a 3.0.3 testbed. Something about not finding the AddPortalContent permissions, but the fix for CMFCore.permissions syntax doesn't solve the problem. Also, the annotation capability for PloneWorldKit is very rough around the edges.

More on this as I dig into deeper levels of these products.