Recently in Software Development Category

Seth has been writing about many aspects of development for the Mozilla browser for a while now, and he’s finally gotten published! Seth contributed a number of hacks to the new Firefox Hacks book from O’Reilly.

Sincere apologies to Seth for letting this post sit in my drafts folder for so long before getting it out there - congrats man!

At work, we recently “upgraded” from using CVS for all our source control to using Subversion. Subversion is what all the cool kids are using, but we chose it for several reasons, not the least of which being intelligent tag and branch management. With CVS, managing our release branches was becoming more and more hairy, while creating a new branch with Subversion is almost trivial. Other bennies:

• Subversion runs over http, making browsing the repository incredibly easy, if a bit spare.
• Versioned meta data called Properties
• Subversion uses a property called externals to identify that a module should include checkouts from other modules. This means that Our main portal app can automaticlly checkout each application from the repositry in the correct location without us having to checkout the portal, then cheack out each app into the portal directory.
• The optimized update/commit code uses diffs in both directions, making updates fast and compares almost instantaneous.
• svn commands are sensble, easy to learn, and based on CVS commands where appropriate
• Version Control With Subversion

I’m trying to get the php accelerator MMCache running on a Mac running Mac OS X 10.3.7, and it’s failing. I’ve followed the install instructions on Jacken’s weblog, and tried the install process described in this Appletalk Australia article.

I did have to install fink, as described in Jacken’s post. After completing the install, when I restart apache

sudo apachectl graceful

it says httpd started, and I do not get any errors in apache’s error_log, but httpd afterward is not running.

myhost:/tmp/turck-mmcache-2.4.6 steve$ ps -ax | grep -v grep | grep httpd
752 p2 S+ 0:00.00 tail -f /var/log/httpd/error_log

Restarting apache again, error_log includes the following line:

[Tue Feb 1 09:25:45 2005] [warn] pid file /private/var/run/httpd.pid overwritten -- Unclean shutdown of previous Apache run?

mmcache.so is in /usr/local/php/extensions/no-debug-non-zts-20020429/, as described in the recommended modifications to php.ini:

; Turck MMCache
zend_extension="/usr/local/php/extensions/no-debug-non-zts-20020429/mmcache.so"
mmcache.shm_size="16"
mmcache.cache_dir="/tmp/mmcache"
mmcache.enable="1"
mmcache.optimizer="1"
mmcache.check_mtime="1"
mmcache.debug="0"
mmcache.filter=""
mmcache.shm_max="0"
mmcache.shm_ttl="0"
mmcache.shm_prune_period="0"
mmcache.shm_only="0"
mmcache.compress="1"

I would be very interested to know if anyone has gotten this working or, better yet, if they had this same issue and found a workaround.

You can’t use DHTML to change the contents of <table> or <tr> elements in IE6.

MSDN says this: “However, because of the specific structure required by tables, the innerText and innerHTML properties of the table and tr objects are read-only.”

Of course, the Safari and Mozilla/Firefox engineers seem to have been able to make the “structure required by tables” work for them, as my javascript works just fine there.

NetNewsWire and MarsEdit public betas: “Here’s what’s new in 2.0 and here’s the download page.

(Via inessential.com.)

Ok, I’m going to geek out for a minute here… I’ve been listening to Adam Curry’s feed of audio posts today, and noted that he often mentions links to other’s sites on the “show” (he is a blogger, after all).

It occured to me that it would be interesting if Adam’s DSC posts included lists of the links he mentioned, as footnotes or something. Then I thought of a better idea: encode the link in the comments field of the mp3 file — trickier to implement but oh-so-very-cool. Looking ahead, I could see a piece of audio-posting software that would let you drag a link onto it while recording, and it would add the link to the output file’s meta data, perhaps also adding the time the link was inserted:

[hh:mm:ss] Harold Gilchrist, Poolside <http ://radio.weblogs.com/0100368/2004/08/18.html<

Then (again thinking ahead) our media reader/audioblog client would read the list and present it to us so we could follow along while listening. Yeah, that would be cool.

I want I want I want I need.

See it in action.

Dave writes about what I’ve always thought was a good idea but have a serious reason to care about now: we want to be able to subscribe to audio content feeds via something like RSS and get it downloaded directly to our iPods. I’d love to be able to have |NetNewsWire| automatically add any .mp3 enclosures to a special playlist in iTunes.

I’ve found some links to my content floating around the web of the form www.redmonk.net/discussion/thread.html$msgnum=2353 So I decided to route them to the same script that maps some of my other old Conversant URLs to the new site, and now the discussion-style links work after a fashion - they get 302‘d to the correct url - in this case redmonk.net/archives/2003/10/30/winter-lawn-redux.