The Middle

Got my Tiger installed, it rocks.

Also caught an afternoon screening of The Hitchhiker's Guide to the Galaxy today. Good, but not great.

I should probably do something this evening that doesn't involve being a complete super-dork.

Heard of it, but hadn't tried until now. The verdict? Works like a charm. FedEx RSS Tracking

Every install of typo comes with a handy feature, an RSS 2.0 feed for comments, by default located at <site>/xml/commentrss/feed.xml (or /comments.xml in the case of my site...)

However, there appears to be something wrong with this feed, as the one true reader completely ignores it. I've subscribed to my own feed, Robby's feed, and scoop's, but none have updated in the last 5 hours. However, I know that at least two of the three have actual new content, just from hitting them with a browser.

Browsing my access logs, it appears the useragent isn't even hitting the feed, but from what I can tell, it's perfectly valid

Could the mightly bloglines be doing something wrong? I doubt it, but it is possible. Time to file a report

Update, 1:30AM EST: So now the bloglines plumber is out fixing things, which makes this awfully difficult to debug...more tomorrow, I guess.

Note: This post has been brought to you by the new-fangled typo admin interface, just as a test, mainly.

Ok, switch made, rails routing magic in place to make syndication formats continue to work, and let all my old permalinks function properly. So what's left to do?

• Extended entries. This was an MT feature that I used to use, but have since stopped doing. Still, there's a lot of incredibly valuable posts from 2 years ago with extended text that typo just ignores. So I'm going to add those back to the main body in one fell swoop, I suppose.
• Something new with the layout? Maybe...
• I need to run Markdown or Textile on all the old comments, and do that by default for new ones. The import was just fine, but it's all plain text with no paragraph breaks thrown in, as MT used to do.

And I'm sure there are 97 other things I haven't found yet...

Made the switch, let's see what happens.

With the monumental 2.0 release of typo late last night, and the corresponding bug fixes this morning, it's time to make the switch full-bore.

The MT comment/trackback spam had reached a deafening level, and I really didn't want to leave the site in 'everything disabled' mode, which was my stop-gap solution previously.

I'm going to roll with typo for awhile, see what the load/spam numbers look like, and have fun hacking at the blog again.

This is going to result in stale RSS .91 and 1.0 feeds, sorry to both of you who are subscribed. I'll probably redirect them to the RSS 2.0 or atom feeds real soon now

May 19th excuse cards: I won't be at work today because...

I could take him: Star-Kick robo-foosball.. it’s all over

The insanity continues with the US Air Travel system, as today was officially the day the TSA banned lighters from aircraft. Because after all, "Terrorists constantly seek innovative means to cause harm, and this requires appropriate risk mitigation".

Additionally, this ban applies to your checked luggage as well. You know, the place where it's perfectly legitimate for you to carry an Uzi.

Welcome to the slippery slope, folks.

If you're a greasemonkey type of guy, check out Interchange. Use it to get to a google map version of your stored yahoo maps addresses. Very handy.

And now you have a satellite picture of the author's apartment, too.

I've been doing more playing with typo, this time trying to wrestle in a way to handle a default category for blog entries. In my old MT setup, I had a category that I used to resrtict entries on the front page (originally named 'Blog'), and I wanted to continue to use this functionality, since I imported all of my old MT entries.

Digging into the code, my first inclination was to modify the articles controller(app/controllers/articles_controller.rb), changing the index action here:

def index
  @pages=Paginator.newself,Article.count,10,@params['page']
  @articles=Article.find_all('published !=0','created_at DESC',@pages.current.to_sql)
end

to this:

def index
  @pages=Paginator.newself,Article.count,10,@params['page']
  @articles=Article.find_all('published !=0','created_at DESC',@pages.current.to_sql)
  @articles=@articles.find_all{|art|art.categories.any?{|cat|cat.name=="blog"}}
end

Obviously, this is far from ideal. Now I'm hacking up the internal code of my blog software, and in a very ugly way at that. On top of all that, this completely breaks the nice pagination that I was previously getting for free.

Fortunately, there's a much cleaner solution. Open up that default routes file, config/routes.rb, and change the following default from this:

map.connect'',:controller=>'articles'

to this

map.connect'',:controller=>'articles',:action=>"category",:id=>"Blog"

Now I have nice default category filtering for my entries, and I don't have to do a terrible hack to make it work. Just a teeny-weeny little routing tweak.

I can live with that.

New site setup, checking the trackback-love: Rails is cool

Like most people running Movable Type these days, I'm in a constant stuggle to deal with spamming problems, even for a site like mine that's never updated, much less read by anyone.

So what's an aspiring young man to do? Install typo, obviously! Check out the new site at typo.andywismar.com. It's got a lot of really nice whiz-bang front-end features (Post a comment, then check out the search box. Just try it, trust me), and I'm hoping that the super-spam bots will pass by without mt*.cgi links everywhere. We'll see about that one.

More than anything else, this was an excuse to continue my Railseducation, and it was nice to hop into some code and be able to hit the ground running, rather than developing another simple CRUD app. I'm finally starting to think in 'ruby mode', where weird blocks and collects and '@@' start popping into my head. I've got another writeup in the works with the various changes I made to get everything right around here, but I'll save that for later. Moving forward, that's the part that scares me the most. With new releases and features being added everyday, I could spend a lifetime keeping up and porting my customizations to each release. But hopefully things will settle down a little. I even managed to GIMP up a little header graphic with fancy gradients, which effectively quadruples my lifetime graphic design achievement.

I also used the fabulous mt3.rb script detailed here to import my whole MT history. I think I've got some import quirks on some of my really old entries, but that's a problem for another day. Right now, I'm going to try to squash out one or two more bugs in the setup, and possibly move the whole old site over. Or move the old site away, and replace it with this one, at least.

Even if this is a horrendous failure, it did manage to get me a Rails/FastCGI setup on my production box, so watch out.

The Google Blog folks have announced a satellite view on google maps, which seems to essentially just be "Keyhole on the Internet".

Very cool, check out my office. And say hi to Mom and Dad.

All the normal clicking, dragging, & zooming google maps rules apply, it's just real pictures instead of meaningless lines representing roads. Very cool.