Media museum

I want a media museum. It is a exhibit museum. It is not Flickr.

It holds media of all types — text, photo, video, audio. I lay out media around a theme — here is a custom-designed page that presents video, audio, text about Chicago. Here’s another that uses some of the same elements, but it’s exhibit about wayfinding.

The exhibits can tour. You can use have some of my pieces to use in yours, but provenance (not scarcity!) is enforced.

Maybe you can comment on things. Maybe you can’t. I suppose you could suggest new pieces for my collection, but it’s my choice whether they appear right away or not, or if I even read your suggestions. Or maybe on this one exhibition we can all work together, because I’ve invited you.

Here’s what the museum looked like in its last revision, if you wanted to know.

Here’s everything in the highest resolution I’ve got.

The future digital humanities museum.

Sign my guestbook?

Posting local civic records online

Mostly a way to prove that I’ve been thinking about this.

Civic records are documents produced by city government. Substitute any public organization for city, if you wish. Documents might be PDFs, or Word documents, or plain text, or excel files, or large-format architectural drawings, or video recordings of meetings, or nearly anything else.

It’s worth thinking about the ways cities can post these things online.

Ann Arbor uses SharePoint. I dislike SharePoint because it’s an expensive, proprietary product that is tied with the Microsoft way of doing business. It has functions (files, wikis, blogging, comments, whatever) that resemble the standards the rest of the world uses. They’re poor approximations.

But the truth is, it works. It integrates with whatever Microsoft file server and authentication and whatnot already exists. Its search can be configured to include the contents of files as well as webpages. Some research has gone into its interface. It’s approachable by City staff familiar with other Microsoft product.

(technical sidenote: SharePoint search does not query its own database; instead, it produces a search index by CRAWLING ITSELF at administrator-defined intervals)

Your Microsoft administrator can probably take some expensive classes and be able to do nearly whatever you want with SharePoint.

That solution maximizes the ease-of-use by City Staff.

The best open option for running a City website that comes to mind is Plone, which is based on the open-source Zope server. It has quite powerful versioning features, and a user interface that is not difficult to master.

What it does require is significant implementation time. The civil servants who will be making it work need to know the language it is written in. They also need to know how to manage a server. Time will need to be spent integrating the city’s login system with whatever the City already uses.

Unlike Drupal, upgrading Plone generally does not break all the work you’ve done before.

That’s immaterial, though, since both are tools in the hands of experienced wielders; the problem being cities don’t have money to hire those folks and if they do they’re busy doing other things.

More thoughts later on architecture. There’s a layer above all this also warrants some thought.

1) Who gets to add and subtract from the website?
2) How do you announce additions and subtractions?
3) Where are documents kept? How?
4) What about public commentary?

1) Who cattle-prods staff into posting stuff to the website? Who decides what kind of information must be posted, and how often?

Maybe you create a central information czar whose job it is to frustrate overworked staff by demanding documents. It becomes part of their annual reviews, have you provided relevant information to the Czar? That won’t go over well.

Or you hope that staff will be conscientious enough to put things up in a timely fashion. This assumes a] that they aren’t overworked (they are), b] that they won’t withhold or delay things that are important to them, c] they know how to use a computer, or d] they feel that they have a public obligation.

2) Announcing new information doesn’t seem so hard. But there are some got some sharp edges once you want to go past the minimum.

Okay, so you make an RSS feed of changes to each collection of information, whatever that may be. And you CMS provides an RSS feed of changes to each individual piece of content. Great.

What about all those folks who don’t know what RSS is? So you set up an automated email system.

Now you’ve got this email thing to code into that Content Management System you’re using. And the vendor isn’t about to add it.

The solution to that is to get a $10,000 Knight News Challenge grant or an equivalent and pay a] someone offshore or b] local programmers and have them do that for you. Then you contribute your addition back to Plone or Drupal or whatever so people can use it (add $5000).

Is that enough? Should you organize people by neighborhood, so they can get info about their neighborhood without having to wade through all the crap about people living 2 miles away? How much extra tagging burden does this add for staff?

Outsource that work. Pay someone in who-knows-where $0.25 a document to add predefined tags that categorize document. Does this violate the fair labor practices part of your city charter?

3) Putting documents somewhere simple is important. But what’s simple? Create a directory by topic (aka budget, planning, …) or organize things by the city org chart? For ex, the cable commission in Ann Arbor is stored in the Community TV section of the website. Does this make sense? Should there be one section on the site just that holds subfolders for each commission, regardless of its affiliation with a city unit?

What do you do with documents when they’re outdated? Do you throw them away? That makes no sense. Move them to some sort of archived position. Maybe you can use your storage system’s revision tracking feature

Stop posting PDFs as scanned files. It makes you look like a luddite if you don’t know how to remove Tracked Changes or combine multiple files into one.

That has the added benefit of making the files searchable, which is something people do with Google, not your website’s useless search engine. Why do you think you can search your own site? You cannot. You do not have an army of Computer Science Ph.Ds. Do not devote resources towards it. Use those resources to make your website searchable by other peoples’ systems.

Let’s say the City budget is this huge PDF. If you want someone to find info in it, what do you do? Do you just say, download the PDF, and go to page 144? Or can you post that PDF in such a way that you can link right into page 144? Should you post the parks section of the budget on the parks section of the website? If you take that latter route, you now have many pieces of the budget that all need to be kept up-to-date.

4) Does it make sense to allow citizens to comment on City documents on the City website? Should people be able to add annotations to that budget? Well, they’d at least need to be clearly marked as separate from the original document. So maybe the official version lives on one part of the site, and another section of the site is devoted to public markup and commentary. You’ll need to take some time to integrate anti-spam features. Or maybe all that’s the role of newspapers, to post those documents in a form that people can use in public discourse.

Coming up when I feel like it:

The difficulty of going past the minimum

legistar and posting documents on line

Also, how creating a system like this is quite possible in receptive communities, and what to think about that.

How long do you need to be embedded in the organization before you have the legitimacy to propose undertaking these features?

The tricky issue of video, and how others have partially solved the problem

Wasted space

Michigan Marketing and Design responded fast to Jacob Nielsen’s post on all-caps, right-justified navigation (the RC hasn’t, yet…)

I’m surprised he didn’t mention the 50 pixels of wasted space on either side of the navigation (highlighted green in the image below). Whitespace is necessary, I understand, but in this case it could be well used to increase the font size (and therefore readability) significantly.

2008-04-29_1217

New Michigan site of the week:

A post on the University of Michigan’s Community Information Corps (CIC) website announces the launch of Open.Michigan.

The CIC is a place for School of Information folks that “provides students with readings, lectures, practical engagement service opportunities, research projects and social and professional networking connections to launch them into careers as public interest information professionals”

Here is the Open.Michigan homepage. From this screenshot, what is the purpose of the site?

2008-04-28_1850

Heaven knows.

Design lesson: explain the purpose of the site on the homepage. “Connecting the global learning community” is not a sufficient explanation.

Ostensibly, it:

represents the diverse collection of Open initiatives on campus – from open access publishing and open archives to open source software and open standards. The site provides greater visibility to the various projects and attempts to expand the dialogue between campus participants and external collaborators.

That is not really a global learning community. It’s an internal site for the people who are already in the system. Cf. the text of the about page, which (unlike the homepage) actually says what the project is about:

Open.Michigan provides a clear view of the many places and ways U-M contributes to our world’s knowledge

the U-M Health Sciences Global Access project and the dScribe project; in open source software, the Sakai and SiteMaker projects; in open archives and publishing, Deep Blue, digitalculturebooks, MBooks, and OAIster projects; and in open standards, the IMS Global Learning Consortium work and conferences.

Do you know what any of that stuff is? Browse further, and woah, you’ll actually learn! I’m not quite sure that the site is so much a “part of an emerging paradigm for participatory education on a global scale” as it is a static collection of project abstracts, but it’s a start. A first prod at some sort of aggregation at Michigan, and a laudable one.

Some of the descriptions need a buzzword pruning, though:

This new program will combine continuous, formative and summative assessment of higher order educational outcomes with flexible learning paths for achievement in nine defined competency domains.

One thing interesting from the introductory post on the CIC is a link to a Wiki on some bizzaro Med School server that has lots of training documents. Of note: we learn that rich presentations are converted into (get this) JPEGs for dissemination through Open Courseware. Beautiful, and very Michigan.

Thinking about RC information architecture

I wanted to think out loud about the information architecture of the RC site. Probably not interesting to you, but a valuable exercise for me, and I don’t apologize.

It’s bad right now, for many reasons.

One is that since everything is in  the left column and not well delineated,  you get a nasty mess of links once you’re a couple levels down in the site. And that nasty mess exists on every page — it’s not pulled from a template.

But what really irks me are the large categories. I wasn’t a part of the initial design process, but it looks like some misguided notion of personas was pushed. But this isn’t a strict B2B site or intranet where all users have well-defined roles. Prospective students want to know what classes the RC offers and what groups exist, as do alumni, staff, and friends; but that information is, illogically, all under “current students”. And “about the rc” is found under “prospective students”. Arg.

I’m not in a position to completely rewrite those categories yet, but a stab in the right direction would probably look very traditional:

  • about
  • academics
  • rc life (players, benz, )
  • rc community (gallery, PALMA, SLIP, urban org)
  • alumni
  • calendar / news

But with cleverer nouns.

Even that raises problems; where does EQMC go, or internship listings?

Where to put the top-level navigation is even trickier. In the end, it’ll go up horizontally at the top. That seems inevitable. But then  we’ll have another Cornell / UChicago / Community High. Boring! (but usable?)
Don’t get me started on the file structure — it’s flat, and everything’s in one folder. That’s actually a good thing, as we get ready to move to a platform that supports a large amount of open source software — there’ll be less fiddling to do. At the same time, it means that (with some grunt work) we can reorganize the navigation — without worrying about moving files into different folders, which would break links and cause a whole different set of problems.

The project management software I want

Craig sends it an email like this:

“Matt will update the photos and captions on the homepage by Tuesday. Here are the files he needs: [attachments]”

And I get an email right away (and whenever else I want) that says:

“Update the photos  and captions on the homepage by Tuesday. Here are the files: [attachments]”

There would be a project page where I can log in and check all that stuff, since I lost all those emails and then turned off notifications for a week.

When I’m done, I just send it an email saying,

“I updated the homepage. Took 1 hour. Craig now needs to find new photos for next week”

And the system knows what I’m talking about.

Cornell Redesign Blog is back

The Cornell Redesign Blog, a valuable resource for anyone in academic digital development, is back as View Source. The writing retains its unified, entertaining voice and its plethora of embedded hints and strategy. Nice to once again hear from such a talented group of confirmed humans.

Mailing lists: UM vs the Real World

UM:

screenshot_4.png

Real World:

screenshot_5.png

That’s pretty much my job… bringing the Real World into the U of M.

Today’s reason(s?) to not like Drupal

You can define custom content types with custom fields using the CCK plugin –

BUT! you need a second plugin (“contemplate“) to define how they display, and that’s not obvious.

Even with that plugin, there are aspects of article display you still can’t change. I call this the “System wants to insert random crap effect“. It’s prevalent in most poorly thought-out systems.

Also, default URLs are yucky– if I’ve got a content-type called “projects”, entries should have urls of the form “/project/name”, like every other CMS, not “/node/112″. Sure, the latter is easier for the developers to implement… but really, who made that decision?

I still don’t understand why Drupal is so popular when many other systems “just get it” right out of the box.

[update] Here’s some code that Drupal produced today:

<div class="field field-type-text field-field-people">
<h3 class="field-label">People:</h3>
<div class="field-items">
<div class="field-item"><?php print $field_people[0]['view'] ?></div>
</div></div>

What?!

UM Math dept. starts to listen on web homework

Many UM undergraduate math classes have homework based entirely on the web — this semester, the dept. is actually looking to see if the system is working. This analytical spirit is rare, and I applaud it.

Here is the UROP project description (more on UROP later, hopefully I don’t have to file a FOIA request)

This [project] will specifically seek to answer the following research questions:

1. Do students spend more time working on the on-line homework than they do working written homework, and is this affected by the inclusion of the homework as a part of students’ course grade?

2. How does the use of on-line homework change the manner in which class-time is used?

3. Do students in classes using on-line homework perform better on the uniform exams in the course, and do they perform better on questions that specifically resemble homework problems in the course?

4. Do students in classes using on-line homework demonstrate better skills at finding antiderivatives and evaluating definite integrals?

5. Do students work a greater percentage of the homework when it is assigned as on-line homework instead of pencil-and-paper homework, and do they work on problems requiring greater thought when those problems are presented in the on-line format?