In my haste to fix the Mozilla history bug, I failed to regression test against Konqueror, leading to a bug that hindered loading of the NW codebase on that browser. Things now seem to be fixed for all Konq 3 verions.
Not quite sure how I missed it, but damn
Matt needed some code to do DHTML tooltips for the CIRDB, and since netWindows includes code for this, I volunteered to extract all the netWindows dependencies from that code and make a stand-alone version. The results are now available for download.
As usual, the code is copyright myself and released under an MIT/X11 license. See the included test file for information on how to set up the script and for sample usage.
Matt has put the NW tree widget to good use, providing a tree-based interface to his netWindows CVS tree checkout. You can checkout the PHP code for it too.
One of the things Matt suggested with regards to the tree widget was an "opened" flag that determines whether or not a node that has children is expanded by default. It seems pretty obvious in retrospect, but it's a good thing Matt said something or I might never have done it. That said, support for the "opened" attribute on inline tree widget declarations is now in CVS.
I've received a couple of responses about the blogger-replacement post. Some have suggested Moveable Type, but I have my missgivings about it.
Firstly, it's written in Perl. My opinion on Perl is that it has lots of wonderful uses, but that Perl's utter unreadability is the reason that Python has been successful. Some may call this a cop out, but frankly I don't feel like auditing lots of Perl, and that's what I'd have to do if I went with MT. Security is a big deal for me, and Perl's illegibility (to this programmer) makes MT an unsafe choice.
My other concern about MT is that it's not Open Source. The license isn't stupendiously bad, but if I'm going to host an application, I'd prefer that it be Open Source.
I've started working with a couple of scripts that use ZODB as an OODBMS backend, and I think I have code to meet most of the rest of my requirements lying about (save the spellcheck thing). We'll see how far I get in the integration/rework before I give up and start reading MT source code = )