live website creation and promotion

Returning to Wiki post, I might choose Wiky, it suits better in my case (example).

( via )

Tried to install and configure IBM DB2 on our virtual dedicated server. It just doesn’t work well - half of operations were aborted due to memory limit. So we’re moving to 1and1 dedicated root server (Fedora Core 4, Plesk 8, 1Gb RAM, unlimited traffic). They have 2Gb RAM option as well, but I don’t think we will really need it soon, even though DB2 is not very lightweight software

P.S. DB2 is just fantastic, I’m able to do things that MySQL can only dream of, and it still works fast, despite of the fact that I didn’t set up any indexes yet - except of primary keys in tables.

I’m thinking about the idea to add personal page to each book author, movie actor and director, and so on (as well as adding a separate comments to all them). The question is - how to implement it? I vote for wiki. Then I need some wiki engine, that I can easily change for my needs - perfect variant would be if it has functions for working with database (database layer) that I can redefine - to use my users info instead of internal usebase, to store data in my database. Or even better - if I can take just the wiki-part (working with separate page, storing history) and use it alone, without all other functions. Does anybody know such wiki engine? Or maybe you have something else to suggest?

[I’ve already built such mini-wiki-engine for eMule Plus forum, but I prefer to use something more powerful if possible]

Update. I’ll probably look into TiddlyWiki first - when I get to this feature at all.

Auto-increment (as in MySQL) is called Identity in DB2 (value generation parameter). And it wasn’t too easy to find it out…

Somebody is trying to copy one of my projects…

I’m having constant problems with MySQL, it is very slow now, especially when there are dozens of search bots parsing my sites.
I also found it very annoying that simple subqueries take a lot of time to execute - it works about 10 times faster after I separate them and use two queries instead of one, passing the data from first to second via PHP script.
It appears to be a well-known problem for MySQL 4.1 - I’m not sure if that’s still happening with 5.0 version - but I don’t have too much time to make thorough tests, so I’m going to move to PostgreSQL, as it seems to be a much better solution. Right now my database code is not too complex (yet), so it won’t take long.
Any comments?

P.S. I understand that this move also means ideology change - things should be done with stored procedures, foreign keys, triggers and so on - so part of business logic is moving to DB. But actually this is exactly how it should be, IMHO.

Update. I’m seriously thinking about IBM’s DB2…

Fixed a problem that caused duplicate data entries being added if you used browser’s back and forward buttons after adding a new book/movie/program to your list.
It was easy, actually
Fix: don’t send all those ‘page expired’ headers on the page after POSTDATA has been submitted.

Added AdSense tracking to all websites at my Google Analytics account, will report here any interesting results as soon as they appear.

Fixed RSS feeds for all Banks of Ideas. Now it’s possible to track all newly added ideas with any RSS reader. Later I plan to add RSS feeds for comments too, which will allow to be totally up2date.

I’m going to write two small helper projects soon. It seems that I don’t have another choice.

One is statistics tracker - I don’t need the fancy numbers, “browser wars”, “OS wars”, screen resolutions data - all this can be checked at any big site statistics, or even at NetCraft - those numbers don’t make any sense per-site, they are global. What I really need is number of visitors (daily, monthly), returned visitors, number of people who performed specific actions daily (saw tag clouds, used send-to-friend feature, etc), number of RSS readers daily (but not readers version counters), ability to see those numbers as charts, compare how they corellate. And visitor paths - to see where people stuck on site. Those are features needed to improve sites’ usability (Plus - I already track referrers in code + check search engine queries).

Another one is translations management. I’m adding new features and new languages (and new sites too), and it’s becoming very annoying to maintain all translations for all sites. I promised such project to eMule Plus team a long time ago and didn’t fulfill my promise yet. Maybe that’s the good time to do it.

P.S. Those two projects have a good chance to become open-sourced when I finish them.