My blog http://james.seng.cc/ used to ran MT now running Drupal

Theme is done using simple.theme (http://drupal.org/node/view/8365), a theme I wrote that mimic MT-tag-style theme.

Migrated using mt2drupal (download from CVS contributions/tricks/mt2drupal or http://james.seng.cc/files/mt2drupal.tar.gz)

Comments

Arcanum-1’s picture

I used Drupal 4 Blogger to convert my old MT blog to Drupal and it worked great. I also used the MT 2 Drupal script and it imported all of my entries perfectly (and very quickly).

My site is not completely done yet (I have to change the logo and a few other things), but you can see it here (http://blog.aranami.net). I like Drupal way more than MT and I've had a lot of fun learning some PHP and lots of CSS through installing it.

I'd love any comments or suggestions on my site if anyone has the time. I modigied the marvin_2k theme. I'm not a professional web designer so I know it's not really that great. It's still a work in progress.

Prometheus6’s picture

Still in development because I have some pretty specific ideas. I've imported some 2500 MT posts using mt-to-drupal.cgi...really sweet. That and the work on the line break conversion that James did saved a lot of grief.

Here is the site in progress. Chameleon with minimally modded marvin-2K CSS, and since I'm a minimalist it will stay that way.

What I'd like to draw attention to is a module that I'm offering James as sort of a quid pro quo: it produces weekly archives (nicely paginated, respects your nodes per page setting) and a block of links thereto for the sidebar. If you check it out, notice there are two "Blogger" accounts blog/2 and blog/3 (I'm not rewriting URLs yet). If you check them out, you'll notice the sidebar block only picks up the posts for each blog when you in them, but on the main page (node) it picks up all of them. It should, in fact, work with pages, books and stories but I don't have any to test yet.

The module was developed with Drupal 4.4.1 and will be GPLed when it's all nice and clean.

garym@teledyn.com’s picture

I did the migration of my blogs over from MT to Drupal CVS, and overall the jseng script was wonderful, a breeze to use (aside from having to hand edit, but that's a security feature!)

But I still have some problems, some minor, one of them really critical ...

  1. the migration preserves passwords, but does not (cannot?) re-encode them with MD4(); this isn't any surprise (because you can't unencode the MT passwords to know what to encode) but I think should be mentioned in the docs; if it is, my apologies, I didn't read that part very carefully ;)
  2. the docs should also point out that you need to create your admin account, then enable the blog module, then create an empty taxonomy for the blog module -- error handling in the script is pretty good, so you learn all this pretty fast.
  3. would be nice to have some mod-rewrite rules for your new migration so here they are, and you'll need to adjust these to your purposes ...
      RewriteRule ^mt/index\.(xml|rdf)$ /node/feed [R]
      RewriteRule ^mt/archives/0+([1-9]+[0-9]*)\.html /node/$1 [R]
      RewriteRule ^mt/archives/cat_WHATEVER.*\.html$ /taxonomy/term/X [R]
    
      <em>(where X is your corresponding drupal taxo number)</em>
    
  4. And this one is the really killer ...
    All my headlines now still link to /archives/NNNNNN.html instead of /node/NNNN!

    Help! ... I didn't notice this at first; I'd tried the rewrite rules on old MT archive URLs, but didn't think to try the titles before I announced that the new site was ready to go (ok, it's just my personal blog, no biggie).

    But what do I do now? Where does Drupal get the URL for the node link on the summary pages? I'd thought it was generated from the node ID, but the node record itself has no 'archives' string anywhere in any of the columns! This is very mysterious.

    Equally mysterious, I tried inserting the following RewriteRule above the normal Drupal redirect, and it doesn't work ...

      RewriteCond %{REQUEST_FILENAME} !-f
      RewriteCond %{REQUEST_FILENAME} !-d
      RewriteRule ^archives/(.*).html$ index.php?q=node/$1 [L]
    
    

    This is an exact duplicate of the old-style translation rules for converting import to aggregator but still gives me a page not found error, even though if I hand-edit the url replacing archives with node and cropping the .html, the page loads just fine. Any advice with this is very welcome, although it is a hack-fix to the more basic problem of having those node-titles link to /archive/NID.html.

  5. As a further migration aid, I have created a basic template for PHPTemplate that eases the migration from MT to Drupal; you can start up with your old MT styles-site.css renamed as style.css and instantly get your page up and running. It's an ugly page, but you get a page to start you off and you can then tweak it to correct the layout. It's not polished, no screenshots, but I've put it into my garym sandbox (themes/unm for un-Movable) just in case anyone's interested.

garym@teledyn.com’s picture

I'm not sure exactly what I did to make it work, but using the following rewrite rule, which I'm certain I tried before, it now rewrites the archives links correctly:

  RewriteCond %{REQUEST_FILENAME} !-f
  RewriteCond %{REQUEST_FILENAME} !-d
  RewriteRule ^archives/0+(.*).html$ index.php?q=node/$1 [L]

This is still a hack, because I really want to just have it generate correct URI.

But a further clue: I notice this only happens on those blogs that were converted (using the jseng script) where the base dir of the MT blog was a sub-path! Interesting ... if the blog pages were /mt/archives/NNNNNN.html they get translated correctly, but if the blog was homed at URI like /mt/lions/archives/NNNNNN.html the migration script mis-assigns them to a Drupal URI of /archives/NNNNNN.html.

but where would Drupal be storing this information? there's no 'archives' string in any node or variable records.

The quest continues ...