slate blog

Those folks who are interested in higher ed might find this interesting. UC Davis did a survey of higher ed institutions and their CMS choices. You can check out the results of the survey along with a blog post that summarizes the content.

My take away stats:

• 129 folks responded which I think is a good catch. The list of institutions. Didn’t realize our very own OIT responded.
• 42 out of 84 CMS deployments identified were of the campus-wide variety. I’m sort of surprised this number is so high. Depts. move faster than us centralized resources.
• The vast majority of those surveyed have only rolled out 0-40 sites with their solution. So we got that beat :)
• Along with that (because of that?) most CMS installs don’t have very many people involved with them.

The last two points are kind of disheartening for me. At least in the sense that if you spend all time I’d hope it’d reach more. But maybe this is my big school-bias showing.

There’s lots of extra info if you check out their blog post along with their own points.

“User interface (UI) can make or break a Web app project. Simple and easy? Users will flock to your project. Overly complicated and confusing? Your project will wilt. Using slate — a Ruby on Rails-based CMS being developed at West Virginia University — as our example, we’ll be looking at lessons learned in developing a good Web app user interface. To help demonstrate our points we’ll share before and after example screeenshots from slate. We’ll share practical steps any developer can take to improve their current applications user inferface. We’ll also discuss issues and ideas a developer should keep in mind when tackling a new project to get the UI (nearly) correct from the very first release.”

Credit to Amy Hoy’s slash7 site and 37signals’ book Getting Real which both played a large part in forming my views on UI and providing material for the presentation.

Updated!
I updated the presentation on Oct. 20th 2007 so all the notes should be there. I also added a few more slides and added a real example for the wireframe stage.

Just a reminder that I’ll be at HighEdWebdev 2007 up in Rochester from Sunday afternoon to Wednesday morning. I’ll be giving a talk on Tues. at 11am-ish on user interface design in web apps. I wish I could give a tease on the talk but, to be honest, I haven’t written it yet. Yes, I’m that guy… the one who doesn’t have his presentation pre-printed for folks to take notes on. Feel free to stop me to talk about Rails, slate, or whatever. See you there!

We’ve had two projects recently where slate has really shined. The reason why is the speed with which we were able to react to and finish projects. The first was the College of Law’s Dean Search site. The second was the new WVU Alert site. Lets look at Dean Search first and why it was a success.

Dean Search: All About Multi-tasking
The site was requested in the AM. We had it delivered by close of business for review. The really killer thing was the multi-tasking aspect of the process. We were able to set-up a site in slate, start entering content, and had a rep from CoL have access to it even before we had a design done and approved. We were able to slap the design on after the fact without any hiccups or issues. I’m so used to a linear process in web development or at the very least in producing a site. You create a design in Photoshop, get it reviewed, and do the template all before you can start entering content. Here we were doing it all at the same time. It’s pretty sweet how our themes work just like that auto-magically. Now let’s look at WVU Alert.

WVU Alert: Programmable Templates
Tuesday morning I walked into the office to learn that we would be rolling out a new text messaging alert system Friday morning (today!). I hadn’t seen the product, had no idea what was involved, but knew we had to produce. After reviewing the product it was an easy decision to add it to the existing Emergency Information site being run out of slate. The next question was, “Who would have access to sign-up for the service?” The answer, “Only WVU personnel but it had to be extremely easy to sign-up.” Our solution relied upon one of the coolest and unexpected features of slate’s templating system. It’s the ability to drop in raw Ruby in a template which allowed us to detect referrers and IP addresses as well as authenticate users against an LDAP system. It only took an hour or so to implement but it’s a really nice customization for a site all without touching the real code base of slate. If you’re off-campus you can figure out for yourself how we’re making it easy. If you’re on-campus, well, it should be a breeze to get to the sign-up form :) Oh, and sign-up if you haven’t already ;)

I’ve been spending a little bit of time today cleaning up slate and making sure we have Google Analytics for all the sites. Just wanted to share some numbers that I’ve found out.

We now have 71 live sites. We had 15 live sites as of Feb. 11. So we’ve added 56 sites in ~33 weeks or ~1.69 sites a week. Check out our updated list of sites. We still have quite a number of sites waiting in the wings. To me the 1.69 number is pretty crazy and we’ve done a fair bit of work outside of just the sites in slate.

Over two weeks (Sept. 19-Oct. 3) we served up 139,313 pages. That’s a little more than a page every 10 seconds on average during that time period. Also, it’s 2.78 times more pages than we were serving up around Feb. 11 (69,000+/wk vs. 25,000+/wk). So on average over a month we’re now serving over a quarter of a million pages with slate. Not to shabby.

Overall slate currently has the following:

• 3,549 pages
• 4,753 resources (e.g. images, pdfs, etc. but doesn’t include themes)
• 526 blog articles
• 34,434 submitted comments though a fair number have been marked as ‘rejected’

All I have to say is wow!

For those out there who want to use TEH but just want the JS because you’re not using Rails you can now download it . I’ll be honest, I didn’t spend a heck of a lot of time putting it together so don’t expect amazing directions. It does seem to work and I think the example (included) is pretty straightforward. Enjoy!

PS – Looking at the TEH javascript made me a little wistful. How scary is that?!

I’m going to be presenting at HighEdWeb 2007 in the middle of October in Rochester, NY. I’m giving a talk entitled “User Interface is King: Developing a Web App UI.” Not going to be anything earth shattering and I should stop procrastinating on putting the slides together (they’re due tomorrow). That being said, I’ll be in Rochester so if you’re going or you’re just in Rochester drop me a line in the comments.

Sorry, an article that’s not related to slate. Had to put together some search engine optimization tips for a meeting I had earlier this month. Figured I could share them. I’ve culled these tips from several sources and I don’t guarantee you’ll be ranked #1 by Google… but these have to help some.

1. Text, Text, Text: Obviously the key to getting indexed is having content in a format that can be read by a search engine. Provide an introduction for anything that is graphic-heavy or relies on Flash. Avoid JavaScript-created links or, at least, provide an alternative way to get at the link. Avoid frames.
2. Keywords in body text: When you?re writing your copy make sure you include the words you expect people to search on. If possible and makes sense include them multiple times though don?t overdo it. Obviously keywords should also be put in the keyword and description metatags as well.
3. Fresh Content: The more often you update your content the more the search engines will search your site and the more likely you?ll pop up the listing. From personal experience this works really, really well.
4. Focused Content: The more content you produce on a particular topic the more of an ?expert? your site may become assuming you?re not just rehashing the same thing over and over.
5. Markup Accordingly: Use H1, H2, etc tags. Alt tags and title attributes also help a lot. If you?re not using XHTML look into it.
6. Be Light: Don?t bloat a page with a lot of text. Keep pages down in size. The larger a page the more likely search engines will skip over it as spam.
7. URL Paths Matter: Directory names are looked at as keywords by some search engines (e.g. Google). So make sure they make sense and use dashes or underscores to separate words.
8. Domains Matter: Similarly to URL paths your hostname/domain matter. By default .edu?s are trusted so that?s a bonus for search engine rankings. At the same time make sure your hostname is relevant to the topics you feel people will search for.
9. Inbound Links: The more trustworthy sites that link to you the higher you?ll appear in the rankings. This was Google?s original claim-to-fame. Tough in higher ed but don?t be afraid to advertise what you?re doing to relevant outside organizations. Also include addresses in press releases or similar material.
10. Statistics: Pay attention to your stats with a specific focus on the keywords being used to find your site. Adjust your material accordingly.
11. Be Patient: It will take a while for any material changes to happen. The more you update your content the faster it will happen though.

Sources:

• Top 12 Search Engine Optimisation tips for public sector websites for 2007
• Whitehat SEO tips for bloggers
• Search Engine Optimization for Blogs

Actually iTunes U isn’t the only project recently that’s taken us away from slate but it’s one of a number of them. It’s been an interesting summer.

Join the Conversation
This summer marks a transition at the University (that was pretty clever, huh?). One of the cool things with this transition (for me at least) are the forums that have been held around campus and around the state. We’ve live blogged the local campus events including pictures near-real time. We’ve also been using QuickTime Broadcaster to do live streams of all of the forums. As an aside, seriously easy to set-up that I can detail if anyone cares. slate has obviously been involved as the basis for the site itself and the corresponding blog. I can’t say we’ve done the streams via slate but does a link count?

iTunes U
We’ve actually had a contract with Apple in regards to iTunes U for a while now but didn’t take advantage of it (read: didn’t have enough time). This summer we’ve cranked out a fair bit of content for the new system (not all there yet) and really tried to reach out to groups on campus to push forward and provide a comprehensive view of the University. I recommend you check it out. Chris was a big part of this as well as a number of folks across UAM that I’m not listing only because I don’t know if they would want their names listed.

Mike’s Blog
The blog of the new president of the University, Mike Garrison. Does your University president have a blog? I’m honestly curious.

Another cool project that may be our first true integrated project and definitely my first national project should be coming online on Friday with another the week after.

Let us know what you think of Join the Conversation, iTunes U or Mike’s Blog.

Just a heads up for those folks who care about making their web apps compliant with Safari. You should check out the latest build of WebKit. WebKit is the open source base for Safari. Surfin’ Safari has a good article on the very significant changes to the Web Inspector. While I haven’t had to debug anything yet I’ve been very impressed with how it highlights items that are being inspected and the network window is very, very cool. Also neat is that it’s available for Mac and Windows.