WP has definitively retained the MT post IDs and spot checking indicates that they are addressable via URL in the form of <domain>/?p=<1 to 4-digit post ID>.
So I just need to set up a rewrite rule in .htaccess – there are some varying approaches to managing the unpadded ID form – MT output six-character ID strings so post ID 39 was incorporated into the URL as 000039 and 5769 was 005769.
I keep waffling about whether I should concentrate on skinning this site first or if I should fix the URL redirects first. I guess since I need to move this site from the deployment address to the final address redirects come first.
I haven’t been writing here because of the long drawn out nature of the migration. I have written some stuff on FB that I should bring over. This is a list for me to refer to regarding topics for here:
Links for reference to accomplish the URL rewrites needed to preserve old search-engine links and to combat general bit rot:
WP official info on MT imports (old but with a lot on redirects)
Mudita on the migration and redirects (2005 updated in 2010)
Basically, there are three components: using .htaccess, ensuring that the WP import has access to the old MT post numbers (something I don’t think I spent time on at import), and determining the best way to combine this data. One possibility is actually to do a cold .htaccess mapping on a post-by-post basis but that would involve forcing 404s to run through a text search of about 7000 lines which sounds like a bad idea.
(from Facebook)
My mom, quietly interacting with her iPad, suddenly makes a sound like a cat moving from concern to anger and then suddenly says with quiet, voice-creaking intensity:
“I hate you.”
I have never heard my mother say these words or emote like this. I am taken aback. I ask, analytically, have you ever uttered this in the context of an interaction with a computer that requires the use of a mouse and/or keyboard?
“Probably not,” she says.
Great, I gave my mom a Christmas gift (years ago) that leads her to experience and express emotional states she has literally fought her entire life. You’re welcome!
OK, that should do it. Now I gotta make a list of cleanups and bugs to squash.
Twitter entries cleaned up and reposted. Whatever the caching issue is that intermittently delays individual entry page updates seems to be affecting the site but whatever the entries are correct in the DB, which means, I think, that they are also correct in the WXR.
(NOTE: This was the final entry written and posted in Movable Type. It retains the MT post ID and is addressable via a WP URL that incorporates the post ID, http://michael.whybark.com/?p=6759 at this stage on the deployment domain.)
Aggravatingly, the locked-out text entry fields in the MT UI has recurred. Posting using the quick post method resolved it last time. Let’s see if that’s reproducible.
The 750-odd Twitter derived entries have been cleaned up in an export file. Next step is weighing how to integrate that data either here or in the next CMS.
Finally spent an hour doing diffs on the various entry-export data piles, and very happily, the WXR-format export routine seen here involving setting up a new index template is now proven to work. The missing entries in the export are simply the entries set to draft status.
So the next procedure is to work on correcting the stored copy for the Twitter archive entries and reimport them into the MT DB. Once I get them into human-readable format I’ll be good to migrate in earnest.
In other computey news I have been taking a javascript course from UW online. The class meets in a classroom downtown on Mondays at 6 and the lectures are presented in a distance learning format as well. It’s the first computer-programming class I have taken since PHP in 2001 and the first distance learning class I have ever taken.
I was able to take good handwritten notes in the PHP class. Handwritten notes are the gold standard for effective speech-based learning presentation, but the classroom standards and needs have changed sufficiently that I have found it to be completely impossible to take notes with a pen and paper, copy and paste code samples into an open note document on the laptop, follow along in the online presentation, ask questions via chat window, and develop and debug example code in the browser as the classes are being taught.
So I’ve moved on to all-typing note taking and consolidated the cut-and-paste doc with actual interactive note taking. However, and this is really kind of an issuer I think, the predominance of detailed presentation slides made available as PDFs largely obviates the need to recopy the data as it is presented. This is inherently unfortunate, because it means that rather than reiterating the data presented in my mind prior to consigning it to a record-keeping medium, I am largely preoccupied with making sure my interaction with the available text preserves it accurately in a mechanical sense.It certainly speeds up the presentation of the material but it must necessarily degrade the quality of apprehension needed in order to effectively capture it.
I suppose this reflects the way we use and have integrated computers into our day to day lives as information retrieval and sharing devices – I would imagine that hardcore date-based memorization in disciplines such as art history is essentially de-emphasized outside the area on an individual’s academic concentration, where the dates will be memorized anyway by necessity and exposure. For everything else, there’s wikipedia and more-vetted sources.
New cousins had to put down a big dog this morning. Been working on little something for them. It’s coming along nicely.