I can’t believe SQL*Plus has survived for 5 minutes, let alone decades. And for some other equally unanswerable reason, free replacements are hard to come by. Toad is the most popular, but it’s this so-called “freeware” that expires every month.

Anyway, I think I’ve found an open-source tool with a past, and hopefully a future. Eclipse has a plugin called SQLExplorer that uses JDBC to connect to Oracle databases.

UPDATE: This seems to be still alive, I updated it today to version 3.5.0RC2. I had to download the plugin from sourceforge, extract to the Eclipse 3.3 directory, then re-set up the Oracle drivers, as detailed below…
(continued…)

The other day my coworker Gamini asked me if there were a C++ exception handling mechanism to “continue” from where the exception was thrown. I thought I remembered some obsure keyword… wrong, C++ exception handling follows a “termination model”, the stack is unwound as you go, the damage is done before you hit the exception handler, and there’s no going back. In this case, there are some other languages (Smalltalk, Lisp) that do a better job than C++, by providing a “resumption model” of exception handling. When an exception is thrown, these languages walk back up the stack without destroying it, or preserve it in some other way, so that a “resume” is possible if the exception is “tolerable” (after you log the condition or take some other moderate action). Wikipedia has a good discussion. And of course boost peoples have an even livelier one, as expected. :>

Spent some time today exploring UnrealEd. It’s an excellent editor for UT2004 mods/maps/models/scripts/etc. that really let’s you jump in and mess around. It’s included as part of the standard UT2004 install – look for UT2004\System\UnrealEd.exe. Couple it with these awesome video tutorials and you’ve got no excuse to start cranking out that great new mod. My goal: set up a friendly game of tag that the girls and I can play.

Been tweakin’ my life away with my silly google-in-a-frame portal. You put your search request into a top-left frame. It then drops google results in small little nuggets down a thin left pane, under the request. When you hover over each link, it gives you the link google details in a tooltip (ala DHTML). When you click, of course, the right pane fills with the result. You can flip around in the results at lightning speed. Next I think I might try working on prefetching the first few results. I’m trying to get to the maximum google consumption rate that my brain can manage. Let me know what you think so far…

From the “you mean it’s working??” dept…

Woke up this morning to the latest admin adventure: none of my mysql-driven sites were working. Tail’ed the log and found out I was out of disk space! Well I’ll be damned, the next time I tell MythTV to record every episode of SNL it can find, I’ll be more careful – that was only a couple days ago!

Everything but b2evolution came right back up after cleaning a little space and bouncing mysql. phpMyAdmin showed that mysql was stuck on the posts table. A simple “repair table posts” right in phpMyAdmin, and we’re all better now.