Zempt Account Profiles

Our posters have been using Zempt, a third-party MT posting tool, with great success. However, we have a number of computers used by a number of librarians, and setting up Zempt profiles for every librarian on every computer would be an extremely time-consuming job, and since it's a manual procedure, prone to (transcription) errors.

So here's how we're doing it (we're lucky that every machine has the same Windows OS):
1.) We install Zempt on one machine and set up all the desired user profiles.
2.) We export from that machine the registry key [HKEY_CURRENT_USER\Software\Zempt\Sites] and save it to a floppy (unfortunately we do not have common network folders).
3.) We then install Zempt on the other machine(s), and after the program is installed import that key into the Windows registry(ies).

Be aware that this procedure (I think, I'll double-check) transfers all user profiles from the source computer, and erases all previously entered user profiles from the target computer.

Zempt stores the user profiles in the registry, and not in a data file in the Zempt (or other) folders. I've kept these instructions purposely vague-- if you're not familiar with the Windows registry, you probably shouldn't attempt this. Also, as always, make a backup of the registry before modifying it!

Old pages now redirect to new

The URL for our original no longer used blog (http://www2.sls.lib.il.us/MVCC/blog.htm) now redirects to the active blog portal (http://www.morainevalley.edu/lrc/blogs.htm). We did this rather than delete the old material because search engines still point to the old. We moved all entries for all categories to the appropriate new blog and deleted all entries from the old. The past is behind us.

Main blogs page now carries feeds

A fine idea, the previous message "RSS to blogs.htm page?" The LRC main blogs page now (or by the end of the day) carries headline feeds from the 4 public blogs with general appeal, so it acts as a stripped-down aggregator. Eventually, we expect our patron community to begin using aggregators, but the leap to the full experience (blogs, blogging, posting, feeds, aggregators, rss format wars, etc.) might be a bit too much for right now. For a rundown on the mechanics, see the entry Feed over, Frankenstein. Also, a shameless link to Arbitrary Content, my web page, which contains links to RSS/JavaScript info on a few sites.

RSS to blogs.htm page?

Would it be possible for us to post recent entry titles on the blogs.htm page? This could work as our own agregator for students, faculty, and staff. I know we want our staff to move to aggregators, but I imagine that it'll take our general public a little while (months, years?) to get to that point?

Multiple Users...Solved

Ok, I played around and got it everyone added to Zempt at the Info Desk. I was just a little slow.

Multiple Users on Zempt

Is it possible to get multiple users on the same instance of Zempt for the reference desk?

Comments Coming UP

AHHH!! I installed Zempt on Barb's and Leslie's machines. I forgot to turn comments off on Leslie's machine, so comments came up. I asked her to remove them. However, I turned comments off on Barb's machine and they came up. Do we have to turn them off for each post?

Feed over, Frankenstein!

I like this, a lot. We've taken the headline-only feed from the Frankenstein blog and use it as (dynamic) content in the Library's main Frankenstein event page (the section titled "Exhibit News Links"). Each headline links back to the appropriate blog item. We don't have the type of server access we need to do this ourselves (at least right now), so we're using Alan Levine's demo, RSS Feeds via JavaScript. Cool to think the feed leaves the SLS server, visits Maricopa Community Colleges in Phoenix, and then returns to Moraine. A happy kind of outsourcing.

Colors and Graphics in Place

Tracie delivered the sharp graphics and color schemes, and we're ready for prime time. We added a blogroll (bottom left column) which, we hope, will add extra interest and drive a bit more traffic. Tomorrow we present to all the librarians during inservice...

Third Party Posting Program

We took a look at Zempt and went for it immediately. Why? Posters won't have to open a browser and point to the MT site (Zempt immediately opens the add/edit item window); it includes a spell check; and the interface is simple and fairly intuitive. It would be nice if the editor was WYSIWYG, though, and we'll have to install and maintain all the instances, but still...

Also, it's free.

Contact Information

I added contact information near the bottom of the right column for two reasons: (one) that column holds links & general info, so the addition seems appropriate, and (two) that column will not grow so quickly as to bury the contact/responsibility info at the bottom of a long list. However (and there always seems to be a however), I ran into some type-sizing problems.

First, in order to make the contact information (there's a lot) appear reasonable in an 800x600 window, I had to reduce the font size, which makes it fairly small at higher (1024x768) resolutions. Second, browsers will not introduce line breaks into long literals, so at 800x600 and lower resolutions the email address libraryweb@morainevalley.edu prevents the proper width sizing (20%) of the right column. This makes the left column too narrow and the list of recent entries visually confusing. So I chose not to spell out email addresses but to use the HTML "mailto" link tag instead.

Images must be reposted

We exported the data from the original (single, multi-category) blog and imported that data into our new (separate) blogs. However, the only data "exported" and then "imported" for images was the image URL-- the image files are still sitting in the old blogs' directories. So before we ask Jenny to delete the old blog, we have to remember to re-upload, to the appropriate blog(s), all the item images.

No More Categories. All is changed!

SLS has given us separate blogs for each topic, so we no longer have to do category engineering. Very nice! No per-category feeds, no individual category archive/index templates, no debugging per-category searching, no error-trapping for uncategorized posts! We'll retain the posts relating to these topics as an archive of sorts, but life is now a lot simpler. Once the new layout gets OK'ed we'll use it for all six blogs and ask Jenny to delete our original multi-category blog (we don't want search engines finding it).

We also added a new page to the Library's site introducing the blogs, explaining their purpose(s) and offering links to each (it hasn't been published yet so it's still a dev instead of a www. It's at: http://dev.morainevalley.edu/lrc/blogs.htm.

Category Searches Fail

The category searches don't work. In some instances they do, in others no matter what category initiates the search the results come from the Blog Development category. This may be related to an error that's returned during a site rebuild; it may not be. We'll look into it this week.

Per-Category XML Feeds

Anders Jacobsen's article Optimizing Movable Type - Category XML Feeds showed us the way (his article on MT XML Syndication is also clearly written and very informative). We followed his step-by-step instructions and they worked like a charm: we had a new template (which we named "Category RSS Template" as Anders suggests) that generated an XML/RSS feed for each of our categories.

But we want two feeds for each category: one that carries the full text of a post, and one that carries the headline (item title) only. To do this, we made a copy of our new template and named it “Category RSS Headlines Template.” When setting up the configuration of this headline template (the "Blog Config > Archiving > Add New..." step), make certain that you give the file a unique name-- we used "categories/<$MTArchiveCategory dirify="1"$>/indexh.xml" (We named the first feed "index.html").

Then, to get the level of content we wanted (full/headline) we changed the <description></description> section of the “Category RSS Feeds” template to:

<description><$MTEntryBody encode_xml="1"$></description>

and removed the <description></description> section of the “Category RSS Feeds Headlines” altogether.

So far so good, but both feeds (full text and headline) carry the same <title> value, the name of the blog. Easily solved: in the full text template change:

<title><$MTBlogName encode_xml="1"$></title>
to
<title><$MTBlogName encode_xml="1"$> - <$MTArchiveCategory encode_xml="1"$></title>

and in the headline template change

<title><$MTBlogName encode_xml="1"$></title>
to
<title><$MTBlogName encode_xml="1"$> - <$MTArchiveCategory encode_xml="1"$> Headlines</title>

The <title> value will then include the category name, and the fulltext and headline feeds will have different <title>s.