Time Heals All Wounds.. And Then Kills the Patient
<Previous Next>
Evening
Evening
Sun Jul 20 13:46:54 2003
POUND progress
Topics:

POUND, the successor to my current netdiary software, is now nearly ready for primetime. Of course, it's not nearly done, but it's pretty much at the point that netdiary is. Why don't I switchover right now? Let's first review where it is now.

pndc is about half-implemented. It has absolutely no error checking, but it can post to the database, redate entries, and has retrieving stuff back out half-implemented (well, implemented, but not to spec). It can also give lists.

The schema is more ambitious than the current code to manipulate it. It has provisions for logins, comments (which I might still rework), and topics. It doesn't yet support access logs, which pndc will eventually support viewing. On the other hand, I might decide, out of space and performance concerns, not to bother with such logging.

The cgis are coming along well. It is indeed possible to cleanly handle all the page stuff through SQL that was ugly to do by hand. OTOH, I realize now that I was doing it wrong, and doing it all over by hand wouldn't be nearly as bad. The CGIs also lack the error checking they should, but they respect private flags, have better page handling, and have per-entry viewing options (which will eventually include comments, when comments are supported).

So, what's wrong with the code that I don't put it up now? 1) It's insecure. I'll need to look at the schema, and set it up to have various accounts, in order to feel safe running this thing on a public server 2) It doesn't yet have the cool time-of-day icons that netdiary has. This is easy enough to add, but it was kind of hackish the old way, and I'm thinking about better ways to do it.

Stuff I haven't started on yet: User logins, RSS feeds, personalization of color schemes, web form to post new entries, backups, files

Oh, Friendster is getting a lot more interesting now that I have some friends on it. W00t.