« Radio to MT: Migrating the data »

This evening I used Bill Kearney's Exporter tool. Exporter is a Radio tool that exports Radio data into into a text format that MT can import. The process is: you install the tool into Radio, run it, FTP the resulting file to the server on which MT is installed, then run MT's import function.

This all worked very smoothly for me, with one hitch. When I write blog entries in Radio, I always use the sequence "<br>\n<br>\n" to separate paragraphs, because I don't like the way that Radio translates blank lines — it uses an unclosed <p> tag. (I don't use the crappy IE WSYIWYG edit control because I use Mozilla instead of IE, so I've gotten into the habit of writing raw HTML when I blog, which I hope will become unnecessary.) Anyway, it seems that the default behavior of MT's import routine is the add a break tag every time there's a newline in the imput text, so I ended up with too much vertical whitespace between paragraphs.

I wrote to Bill, who suggested massaging the file before feeding it to MT, but I read the MT import format spec and saw that they have a per-item setting that controls this behavior ("CONVERT BREAKS"). I added a line to Bill's code to insert "CONVERT BREAKS: 0" for each item, and I got the result I wanted.

I gave Rick's tool a very quick try as well, but it crashed. It would be premature for me to blame his code, since operator error is certainly possible. I may give it another try at some point, but the fact is, the work is already done.

Next steps:

• See if I can import comments (I know MT can do this, the question is whether I can get the data from pycs.net).
• Templates: port the one I'm using now, use one that someone else wrote, or write my own? I'm leaning toward porting.
• Investigate the whole redirection issue, both in RSS and on this site. For the latter, I'm considering writing a Radio macro that, on any given page, would generate JavaScript to redirect (or at least point) the reader to the closest equivalent page on the new site. For example, if someone goes to the page for today's date, the macro would compute the URL for the equivalent archive page on my MT site.

---

« Morning notes »

I had a tooth extracted this morning, which was a new experience for me. It was a mature wisdom tooth that had cracked a couple of years ago. It was repaired at the time, but I was warned that it would eventually have to come out. It was pretty quick. Now I'm waiting for the anesthesetic to finish wearing off.

Good news and not-so-good news on the AmphetaDesk front. The good new is that Morbus sent me a new version (2.27) of the XML::Parser:Expat module, and it cut the memory usage on my system from 108MB down to 31MB, which I can live with. Also good news: based on what he told me about his plans for the eventual 1.0 release, it sounds like it will have the features that I'm looking for. The not-so-good news is that he's a busy guy these days, so it may be a while before those features surface. I'd be tempted to pitch in and help, except that (1) I'm not proficient enough with Perl to be immediately effective and (2) I'm just too busy with my own work to make that kind of commitment right now. But I'm willing to be patient. While I wait, I have the option of continuing to use Radio as an aggregator, or I might choose to use (at least for a while) one of the crop of cool new 3PDNAs.

On the Radio-to-MT front, I got comments from Rick ("techno weenie") and Bill Kearney pointing me at tools that they have written to translate Radio data into MT's import format. I'm looking forward to trying them both, and I will report back here on the results. Thanks to both of them.

---