April 19, 2003

The Blog Street Journal and content-management systems

In the past few weeks I've been surprised with the attitudes and beliefs held by many bloggers. I've discovered that few, if any, have much hands-on experience with technologies such as Apache, PHP, MySQL and even Linux.

Case in point: a recent discussion between myself and another individual at BlogShares regarding content-management systems (CMS). I'm participating in a new project called the BlogStreetJournal (BSJ), which itself is not yet a clearly defined project, but that should change as time goes by (hopefully). This past week the development team decided amongst themselves which CMS to use. I assumed several things before the vote was taken:

- First, I thought that since those on the development team were familiar with not only blogging but how to use the forum (phpBB, open-sourced) and understand how the BlogShares system works, that they were also familiar with the state of information technology as it stands today.

- Second, assuming that I thought none of them would promote products that are not part of the open-source 'free-software' paradigm.

- Third, that if all else failed a reasonable and rational discussion would take place to decide the merits of each CMS.

If nothing else, I assumed that everyone (or at least these individuals) understood how the business plan extolled and practiced fervently by Microsoft, Oracle, Sun & Apple was not only a dead-end solution, but outmoded. SourceForge, the Open-Source Developer Network, Fresh Meat, Savannah and dozens of other similar organizations have displaced and destroyed the inefficient business models that these companies wished to use (similarly, how alternative mediums have leveraged independent bands against the power of the RIAA and the Big 5).

Nope, I was wrong, however let me briefly discuss various aspects of the open-source universe today.

For traffic intensive websites that need security and versatility, there are three mature and highly robust PHP/MySQL solutions that are also entirely open-sourced: Mambo, PHPNuke and Postnuke. I've personally used all three and can vouch for their ease-of-use, speed and stability (we use Mambo at Collectrix and used Postnuke with the original Movementarian). The individuals running the show at the BSJ however had set a time frame that required an immediate decision, so the choice for a CMS downgraded from the solutions that large companies use and support, to those running small blogs (even though you can implement the Nukes and Mambo in the same amount of time).

There are currently no less than three maturing open-source blogging applications on the market and readily usable: Cafelog, Pivotlog & Nucleus. Collectrix (the company I work for installs a modified version of Cafelog for our clients by default); additionally I've used Pivot and Nucleus and can vouch for their usefulness.

Those were not chosen however and the decision was made to use Movable Type (which in my opinion will be displaced in the next couple of years, because of poor licensing decisions).

Why do I propose and suggest that commercial software is doomed? It's actually not a new suggestion or proclamation. Apache continues to defy Windows IIS, PHP is killing ASP, MySQL and PostgreSQL not only have stopped the deployment of the Microsoft SQL server but are also hurting Oracle and DB2. Mozilla and Konqueror will kill IE within the next few years and OpenOffice will do the same to MS Office.

Nevermind your average consumer at home, it's the IT departments at businesses that will be won over and have been. No more expensive or prohibitive licenses (MT and pMachine are 'free' for non-commercial uses, that is not winning the biggest customer base); say goodbye to lengthy waits for bug fixes and security holes (bug and security fixes deployed the same day for Linux/BSD, plus they're continually debugged by thousands of individuals around the globe continually, versus a monolithic entity in Redmond); another plus, multiple-vendors - you're no longer stuck with Microsoft or MT or pMachine or Radio; and lastly, any feature that is deemed useful can merely be 'copied' over to and integrated into an open-source project.

On that last point, this novel aspect of open-source software is that if you get fed up with a particular widget, you can program something and integrate it into the program. If a developer dies or gets tired fo the project, the source is available for others to use and modify. A blogging case in point is with Cafelog (b2). Michel Valdrighi, the creator disappeared several months ago, but that has not stopped the development of the software. Similarly, Linux and Mozilla can be modified to however you would like it (so as long as you can code). One alternative to this of course is the Microsoft method, where you cannot modify the code to fit your specific needs (the Shared Source Initiative is nothing like the Open Source Initiative).

At any rate, the BSJ will use Movable Type for the time being. There really is no advantage to using it though, it simply cannot compete with PHPNuke or PostNuke in terms of development, community or price.

Posted by Tim at April 19, 2003 07:15 PM | TrackBack
Comments

I put together a 'definitive' resource for the Steal These Button collection and added your rss comment ones. I hope you dont mind :)

Posted by: Taylor at April 20, 2003 01:18 AM

Hehe, lookin' sharp there. I might have to "borrow" the Nucleus and Blogroll ones.

Posted by: Tim at April 20, 2003 02:09 AM

Interesting comments, and a very well-thought-out perspective on open source v. proprietary developments!

As one of several individuals involved in the BSJ decision to go with MT instead of an open source code base, let me suggest that although in the programming community at large, there may well be a stronger community to provide code support and development, our team was, is, and will be limited to those who volunteer their time and skills.

Unfortunately, it appears as though right now, the number of people who have coding skills to support the blogging (and other CMS) platforms you mention are primarily, well, you, and one or two other volunteers.

At the time the decision was made, we compared that to the larger number of volunteers who had experience with MT, and with other packages (some of which were open, some not).

Since the licensing fee itself was not an issue (with someone willing to pay it for us), the primary issue was one of "who will make it run?"

And in my quite possibly flawed judgement, I extended this to, "if the entire development team left tomorrow, and a whole new set of volunteers stepped in, what is the most likely platform I could find someone to run among that volunteer group?"

Given the vast number of bloggers with MT experience, you can see why my thinking was slanted in that direction. As I told someone else just now on the BSJ private message boards (where you have access, though your blog readers may not), it would be different if I could handcuff you to the server for the lifetime of the BSJ project! But, alas!, you might decide to leave tomorrow, and I'd be left with a code base that I couldn't find volunteer support for among the BlogShares and BSJ communities.

Now, having said that, if you're willing to be handcuffed, I'd be glad to switch over to a 100% open source code base that you're comfortable with! But starting with issue #2. We've got deadlines to meet, and I don't think I'd be comfortable switching over at this point, even if you're already warming up the handcuffs and coffeepot....

WDave

P.S. It's not likely I'll get back here to read any comments you have on this; if they're useful to the project, may I ask that you copy us in the BSJ message board on any related things you post here in your blog? (Or at least post us a URL to here?) Thanks!

Posted by: WDave at April 26, 2003 12:45 AM
Post a comment









Remember personal info?