Because I needed to alter the site, I had to learn perl (took two 8-hour-per-day weeks), and I'm glad I did. But there are still places in slash that simply leave me breathless with an aching head.
I started installing a slash 2.0.x site coming from 3 years of NT/Netscape/Coldfusion programming experience. I'd never heard of CPAN, knew perl was a programming language but not much else, and that apache was a pretty powerful web server. I'd only been dinking with Linux for about nine months. Took me a solid week of 8-hour days to get it up and running. I completely reloaded the server twice in that time (sometimes I get so screwed up that "Nuke it from orbit, it's the only way to be sure" is the only way I can get back to a good starting place :).
The book is great because it gives you a "you are here ->" spot, and a map to where you go. I learned quite a bit, but altering it took learning the code.
I think it was Krow that mentioned he'd start working on really documenting the code itself soon. I think this would be superb. I for one would love to have documentation that took me step-by-step through what users.pl did. Is it a big job? Hell yes! That's why nobody's done it till now. :)
Some of my projects are nearly as ambitious and big as slash. I think one might actually be that big. If someone put a gun to my head and said "explain how it works! In writing! So a moron can understand it!" I'm not sure I could. So I do know just how big the project must be. But it would sure go a long way toward helping folks like me get the most from the system.
AMCGTLD.COM... where science fiction, cats, and anger combine! [amcgltd.com]
[Part I [microcontentnews.com] | Part II [microcontentnews.com] | Part III [microcontentnews.com]]
Comments on Part I
I don't think the world needs any more definitions for the term "weblog". There is a whole industry revolving around defining and redefining it, Dave Winer's weblog [scripting.com] being just one example. People like Dave Winer have hit the media's radar screen, and some technology oriented journalists have mentioned him (and other weblogs like his) in print. This gives their end of the weblog universe a certain degree of legitimacy.
The problem with the "media-recognized weblogs" is that they are an echo chamber. They generally talk to each other and about each other. The number of topics they discuss is quite limited, and as a group they have become even more concentrated thanks to relatively recent discoveries like Googlebombing [microcontentnews.com], blogdex [mit.edu], and Daypop News [daypop.com].
I think these issues matter because the media is defining weblogs in terms of these media-recognized sites. The authors of these sites put up a good front by acting as if they are outsiders, but most of them have direct or indirect access to mainstream media outlets if they have something important to say. With the exception of Slashdot [slashdot.org] and Plastic [plastic.com], sites that are published using Slash are not in this group of media-recognized weblogs. This doesn't matter unless we want to use the collective popular recognition of Slash-based sites as a marketing tool for Slash itself. But, if that is one of the goals of our project, we need to do a better job of recognizing and promoting each others sites.
I commented on the book "Running Weblogs with Slash" [slashdot.org] in the review I wrote for Slashdot. If you are interested in what I thought of the book and it's positioning, please read the review.
Comments on Part II
I agree with krow and chomatic that Slash is a web publishing or a web application platform, depending on your perception of the strategic value of the plug-in architecture. In the first year to 18 months that I used the ancient version of Slash that I started with, I focused on parts of the system that did not suit my needs, or those of my clients. When the project was a lot smaller, I fantasized that I could maintain a fork from Slash and develop it into the platform that Slash 2.x has become. But, the plain fact of the matter is that OSDN [osdn.com] / VA Software [vasoftware.com] have many good choices with respect to the development of Slash as a strategic asset. We are fortunate to be able to leverage off of all the work that their employees have done.
The plug-in architecture may be the single greatest incentive to stay on the Slash platform for a lot of developers. I am frantically trying to get a Linux infrastructure together so that I can move an application I built outside the Slash framework over to Slash, so that I can leverage the backend user services that Slash provides. This would be a great deal more difficult if the plug-in architecture did not exist. In fact, if plug-ins didn't exist, we wouldn't be attempting this migration.
It's interesting to hear how the authors of the Slash Book came to be involved in the Slash community. Some of us know the story of CmdrTaco and Hemos well by now, so, it's intriguing to hear stories about some of the other community leaders.
With respect to the sites the Slash Book authors like, come on guys-- you could have mentioned CTDATA [ctdata.com]. After all, I plug your work every chance I get :-).
If you ask me where I'd like to see Slash development go in the future, I'd like to see some focus on tools to help build a community. It would help a lot if there were more powerful tools available than "mail this story to a friend". For instance, I would like to see a system where real email addresses of other participants are revealed to registered users who have opted in and maintain a certain level of karma. That would be a priviledge similar to some of those that have been implemented in PerlMonks [perlmonks.com]. I think things like that could go a long way toward making Slash the community building site of choice for user groups and alumni associations.
Comments on Part III
P2P publishing: I spent a long time publishing personal reflections on the attack on the World Trade Center over at CTDATA. These evolved into The War on Terror [ctdata.com] section of CTDATA.com. That effort was cathartic and it built a small audience for CTDATA that went beyond people interested in Perl and Slash, but it wasn't a ticket to the big time in the blogging world.
I'm not sure that any weblog that does not already have a large following can build an audience around a topic that captures the attention of other webloggers and grassroots journalists anymore. Events like September 11 happened before the media-endorsed weblogs realized how important they are. Now, many of these weblogs are engaged in seeing how many of their headlines make it on blogdex. This has decreased their editorial focus to the point that their audiences can't depend upon the truly popular ones to stay with an interesting story as it develops over a period of days. Heck, even the Googlebombing stuff played itself out in a week or less.
I think headline syndication using RSS has reached a level where it needs a technological breakthrough to reach the next level. We've been doing headline exchange, even on small sites for at least a couple of years now. Headline aggregation has been more than adequately covered by blogdex, Daypop, Weblogs.com [weblogs.com], Syndic8.com [syndic8.com], and NewsIsFree [newsisfree.com]. Do any of these syndication sites make anyone stand up and say "Wow!" the way Google did when it first started? Certainly not.
One thing I'd like to see tried is headline aggregation hooked up to collaborative filtering. This would give you something like "The people who liked this feed also thought these sites were interesting." But, this is yet another derivative of headline aggregation.
The Microcontent interview was well done. It's great to see some members of our community recognized for their expertise. This is a positive side-effect of getting a book published about our platform. These happy events will keep occurring if the book sells well, if we keep our development pace high, and if we keep producing high quality web content with the Slash engine. We all have a part to play in making that happen.
Chatham Township Data Corporation [ctdata.com]