Saturday, October 12, 2002

A common design pattern: brochureware + documentitis: My alma mater illustrates a recurring theme in school Web sites: news comes in PDF files, links to which are buried inside relatively static brochureware-style Web sites. So for a parent to get current info, for example the timely news that Oct 12 is parent conference day, they would have to drill down 2 levels from the home page to click on the current newsletter, wait for the heavyweight Acrobat plug-in to launch, and then go to page 2 in a document clearly designed for print, not online viewing. OK, it works (after a fashion) but I think the usability is poor, and it affords absolutely no community-building effect. I'm not complaining about using PDF format - for distributing electronic documents that are, generally, to be printed, PDF is the obvious choice. The VHS school web site above has a very nice link to a collection of printable forms in PDF format. But for online information access PDF just sucks. And as someone who was once Adobe's engineering manager for Acrobat, and helped invent some of its features, if anything I should be biased in favor of this technology. But PDF really is a format for "electronic paper", and the computer display is a very different medium. It's funny I just now imagined that probably Jakob Nielson would have a rant on this and yep he does. Of course Adobe has long had a different view but it's just not reality. Acrobat's great electronic paper and 99.99% of Adobe's Acrobat-related revenue is all about print. Adobe passed on the opportunity to get PDF built-in to Netscape very early on as a peer format to HTML, in which case PDF might actually have become relevant as an online format, leading to fewer explicit-layout hacks in HTML. But that's another story.
Out in the real world: OK there's a thousand failed school web sites I could pick on, but instead I'd like to highlight one that works, but yet exemplifies some of my concerns: Evanston Township High School has a detailed, up-to-date site. I don't care for its organization or presentation but there's no doubt about it, it's current. But, at what cost? It's pretty clear that this Web site benefits from some serious dedicated webmaster resources, as indicated by the heavyweight publishing process described in their Web Site Guidelines And, how used is it? Sure there's news but it's buried away on sub-pages. And when you get down to the departmental or individual activity level its frequently stale brochureware. I'm suspicious that paper's still the primary source of current info and news for most students and parents. And there's little to no visible indication of an involved community - ok there's a set of discusion forums but it appears that only 1 out of 15 has an active discussion going. And, this is is a high school, with budget for technology staffers and presumably a quorum of tech-literate students. Our school by contrast has to deal with zero dedicated tech staffers and a K-5 student body. Don't get me wrong: I admire the accomplishment of the obviously hard-working folks who contribute to the ETHS web site. But I'd love to find a way to get there faster, cheaper, and better. I'm more and more convinced that a Weblog-centric implementation architecture with an RSS news feed information distribution mentality is the way to go.
Dynamic vs. static HTML generation: Dan at IMX has a short article highlighting scalability concerns with dynamically-generated HTML.
Another John Hiler article anecdotally illustrates the disruptive technology posed by Web log software. While it's a great story, I feel the real disruptive potential of Web log software is not merely undercutting high-end CMS systems but rather making content management for the first time a practical technology for small-scale SOHO/school/personal Web publishing, and thus radically transforming the market for low-end Web publishing tools (diminishing the need for HomeSite, FrontPage, etc.).
Personalization - is "MyBryant" needed? the majority of Web log software is oriented towards automating creating of essentially static web pages. While there are "community" features (such as comment systems) the typical Web log site doesn't provide for personalization. That is, each visitor gets the same experience. On the other hand corporate portal packages such as Plumtree inherently support personalization, as do some Web log software. John Hiler over at Microcontent News has a good taxonomic survey of Web log software.

My preliminary take is that, while each parent, student, and teacher definitely will want to access Bryant Web information in different ways this does not need to be implemented by a personalization system withint the Bryant Web site itself, given that "deep" personalization isn't required. The primary need appears to be "I want to see the info that's relevant to me" choices which the news feed features built-in to the Web log model already deal with. For example, I want to stay on top of Mrs. Paulson's Kindergarten class, the Chess Club, as well as school info - but I will be able to simply subscribe to the appropriate Web logs, aggregating them in my news reader of choice, or getting email updates. And, even more simply, bookmarking as favorites the relevant three "nav entry" spots for my needs.

Friday, October 11, 2002

Can it be long until Microsoft offers blogging? OK the image of blogging is a heady mix of body-pierced indy poets, feisty lefty/anarchist commentators, and evangelical open-source zealots so this may seem unlikely but really, it seems to me that blogging features make perfect sense in the Microsoft world view. blogware undercuts expensive full-featured CMS systems with something that's easier to use for the masses at the cost of more constrained capabilities: basically the M.O. of Microsoft in every market that they enter. After all, Microsoft is much more about structured information creation (e.g. Powerpoint, Excel, Visio) than free-form blank-canvas creativity. So in some sense Web logging is a better fit for Microsoft than FrontPage.

Tactically, blogging capabilities would seem a natural add-on to Outlook, which is already calendar- and collaboration-centric. It's also a great fit with the Microsoft Web services strategy in general, and Passport in particular. Not to mention a fit with MSN/Hotmail on the consume end. Naturally they will borrow heavily from the Blogger API etc. but ultimately decide they need to create Microsoft-specific interfaces. In fact I'm kind of surprised all this hasn't happened already - is BillG slipping up?

I find it interesting that - if one presumes that Web log authoring will say be built-in to Office 2004 and Web log server technology will ship bundled with Windows Server that the question of "should Bryant use Web log tools in creating its web presence?" could almost be a non-sequiter like saying "should Bryant use word processing tools in creating its printed documents?"
Stakeholders of school web sites

the primary information authors: teachers, administrators, librarians, PTSA and otherwise involved parents, activity leaders, children

the primary information consumers: parents, children, teachers

I find it interesting that much information communication in the school fits the 1:many author:reader pattern. E.g. teacher to parents of his/her students; chess club coordinator to chess club participants and parents; principal to school community; librarian to school community. Or in the case of school news or PTSA news it's perhaps several authors to many readers. Of course by using the term "reader" I don't intend to deprecate the commenting and other feedback and dialogue opportunities that can and should be encouraged. But it's simply a fact that this 1:many pattern is a match to the general organizational structure of a school. And, happily, is a great match to the assumptions built-in to blogging tools.
More on tool choices - of course to the degree that the Bryant site needs full-scale portal features Slash should also be considered. Perhaps even a step further down the difficult-to-administer open source slipperly slope from Movable Type, but clearly scalable and not narrowly blog-specific.
Tool choices. Pragmatically it probably comes down to Movable Type or Manila since both offer server solutions with multiple-author CMS features. Manila probably has a slight edge on supporting more general web site construction and Movable Type has a slight edge on adoption rate. But it may come down to whether a Windows server solution or a PERL/Unix solution is preferable to those who will have to setup and maintain the installation.
Well maybe it's not just convenient tools: another benefit is that Web logs inherently promote more usable Web content. Latest-information-first presentation, short pithy information "chunks", with "Read More..." links to details, cross-links to background information and other relevant sources, visible information creation dates to help provide context: these best practices for Web publishing are all well-supported by Web log tools and reinforced by the patterns of existing Web logs.

the "blank canvas" provided by traditional web publishing processes does not inherently guide content creators. but busy teachers and administrators don't necessarily have the time and inclination to become masters of Web publishing style - nor does a school web site have a compelling need to break new ground in Web publishing memes. so using tools which naturally guide communication into known, useful channels at the expense of some sacrifice in expressive freedom seems to be, on balance, a good tradeoff.
A very stale elementary school Web log site raises many questions. Probably have to call them and get a take on what happened. School board meeting minutes hint of staff turnover and technology challenges - and there's no sign of that this web site has been superseded by another - but it's all very murky. It looks like the school bloggers got significant help which makes the zombie state even more disturbing. Of course as per the below it's clear that blogware in and of itself isn't a silver bullet but still, this makes one wonder...
Of course a blogware-based Bryant website could still fail. There are cultural issues here - if teachers and administrators simply don't really need to disseminate timely information online, or if even a much-decreased level of required effort still is too much for overworked, underpaid educators, if more parental kibbitzing is more of a negative than a positive, and/or if there really aren't enough connected Bryant parents then it just doesn't matter. But - if that turns out to be the forordained outcome, then (assuming we're going to have any Bryant website at all) setting up up some blogware and slapping on a Bryant theme is likely to be orders of magnitude faster and cheaper than any other solution. So in other words, Bryant-via-blogware is definitely the right implementation architecture if it results in a vibrant, living web presence but it's probably still the right approach even if it doesn't!
Over on Syllabus Phillip Long has a good summary article about Web logs in education, including numerous links. His comment that "Blogging software makes the expression of writing, including the incorporation of hyperlinks for publishing on Web pages as easy as word processing" perfectly captures the primary benefit of blogging that I see applicable to Bryant at the school and classroom level.

I don't think that Bryant should blog because the Semantic Web is the Next Big Thing, because blogging is a transformational filtering mechanism in the dissemination of information, because the blogosphere is now the wellspring of creative expression on the planet, or any other high-minded philosophical motivations. Rather, I think Bryant should consider using blogging tools to foster creation of a live, community-building site that works, simply as a newly available alternative to using more cumbersome Web publishing tools and processes which risk our school site ending up as underused brochureware, if not stale and dead.

Will Richardson of has a good presentation on Weblogs in the Classroom

The presentation covers many possible uses of Web logs, but my primary thought for Bryant - a Web log implementation of school and class home pages - is exemplified by
Delano High School and
British School of Amsterdam

Thursday, October 10, 2002

Truant school web sites abound: as an August 2002 NY Times article opined, many school web sites "are nothing more than a bulletin board of lunch menus and teacher rosters that are rarely updated or are not used at all" where "visiting the site is not part of a daily routine for students and parents, particularly if only a few of their teachers have classroom pages. Even if students and parents do visit, they often find outdated information, like homework assignments from last week, or worse, the previous year". A key reason for this sad state of affairs: "educators and companies providing Web site services underestimated how difficult it was to gather, format, present and refresh" information.

While some in the educational 'Net community have taken issue with the dour tone of this article, it struck a real chord with me in regard to Bryant Elementary's situation, given our prior experience with "web site mortality".
The background: I am an involved parent at Bryant Elementary, a public school in Seattle, Washington. The principal has chartered a team to create a new web presence for the school, as the current web site went badly stale (as did its predecessor). The default plan is likely to involve implementation of the new web presence as a traditional HTML site using standard desktop page-editing software and publishing processes. I fear this is a recipe for yet another dead Web site down the road.

Hence, the goal of this blog is to be a Socratic exploration of the pros and cons of a possible alternate approach. Namely, implementing the new Bryant web presence via weblog software.

Declaring my bias up front: I believe that for the needs of an elementary school web presence will be best met with systems that are extremely usable by the time-constrained and technology-challenged teachers and administrators. The blog model of automatic generation of presentation from easily-authored content is very appealing. The general organizational model of weblogging, with its focus on chronological news items, seems to fit both school-level and classroom-level communication needs. And, unlike say a corporate web site, there does not seem to be a compelling need for a public elementary school to create a highly-differentiated web presence. Thus I am inclined to favor the "school site as weblogs" solution.

However, as a blogging newbie I'm not yet personally convinced, much less able to convince others, that the information model and presentation needs of a school website are a good enough match with today's blogware tools. Nor on the real cost-benefit tradeoffs of blogware vs. traditional page/site design tools. Hence, the exploration.