This Entry Has Been Deleted
Categories
Month: May 2003
Categories
Vindicated
Categories
Epistula V MPT
- GPL
- Not yet, maybe one day. Currently BSD.
- FreeDB
- MySQL, check. Soon (I hope) anything that PEAR::DBI supports.
- Output to XML
- Good idea.
- non-crufty URLs.
-
- http://base/2003/05/02/oneMeaningfulWordFromTheTitle, Not yet.
- No system-specfic cruft. Check (/journal doesn’t count)
- No Queries, Check
- No file format, Check
- Seperate page per entry, Check
- Daily URLs
- check.
- Monthly URLs
- check.
- Yearly URLs
- check. Cool 🙂
- Each entry has a title, a category string, contents, time posted (auto-generated), and one or more objects
- Check.
- Summeries
- Not yet.
- Faceted Categories.
- Well, you can currently narrow down any entry via all the metadata I hold on it, so maybe check.
- Hierarchical Categories.
- Check 🙂
- Automatic subdivision of categories
- Neat idea, but would require keywords so we could tell which should go in which new cat, so maybe not for Aquarionics
- Multiple categories per item
- Check. (By the requirements, when I include such things as GPS data this becomes more complicated, but for now, check)
- RSS for Weblog
- Check.
- RSS per cat.
- Check
- Ping weblogs.com
- Check
- Automatically convert Slashdotted entries to static pages, and switch back to dynamic generation once the traffic subsides.
- A page that is being slashdotted becomes static because of the Funky Caching. Some day soon, I want cached pages to be served as static, rather than the minimum of PHP that this currently uses, so Checkish
- Creative Commons.
- I don’t like them, but others might. Noted.
- Web Interface
- Check
- LinuxSTEP Interface
- Not Check. Never heard of it until today 🙂
- metaWeblog API
- Blogger API, yes. Not metaWeblog, yet. This has less to do with my thoughts on it as a standard, than it is that, basically, I don’t like closed standards, something that’s not helped by my irrational irritation with the person who invented it. Not that BloggerAPI is more open, but the development of it is at least community lead. So one day, when I’m bored, this will happen. Not before.
- Accept entries sent by e-mail
- Don’t like the authentication issues on this one, but noted.
- Make it easy to send entries from a mobile phone
- Hmm. SMS Gateway.
- Me like.
- Import entries from Blogger, Radio, Manila, Movable Type, etc.
- I’ve imported them from Blogger (I changed my template to a psudo XML one and ran regexs to turn it into SQL). But yes, noted.
- Sam added PingBack, TrackBack and Comments to this list.
- Check.
So I agree with everything except the metaWeblogAPI thing. So, where’s the conflict? You can’t have drama without conflict! So my conflict is with the metaWeblogAPI (MWA). My problem is, basically, that it was written by Dave Winer, a man who’s technical abilities I respect, but who is to any social software as a megaton bomb is to a hippy commune. I’ve never seen him work with the community. I’ve seen him work with Biiiig fuckoff companies to get his standards, and Biiiig fuckoff organisations to get his message across, but not the community. I’m going to write the whole RSS fiasco off as a contest of egos, but basically some people developed RSS 1.0 as an open, community created standard (They shouldn’t have done, because it wasn’t theirs to fuck around with, but I digress) which Dave looked at, smiled fixedly at for a while, and said “Some good ideas”, took the good ideas, put them and RSS0.9 out as RSS2, and continued to develop it as a closed standard.
But as I say, I’m writing that off. What he did to the Blogger API thing, however, was to look at it, impliment it, and then run with it. We had a name for that back when Microsoft was doing it to HTML and RTF. We called it “Embrace and Extend” and we meant it as an insult. It’s one thing to develop a competing standard against someone in the same field. It’s quite another to take the pre-existing standard, add stuff until the original author’s app won’t support it, and publish it as your own work. Esspecially when said developers are working on V2, which supports everything your brand new thing does, and was developed with the support and recommendations of the people who were implimenting it, communicated via an email list.
That was a rant. It’s over now. There’s a point in there somewhere.
Categories
2blog||!2blog
I’ve recorded Burningbird’s Request twice, and am not happy with either. It will happen.
I’ve recorded four audioblog entries today.
I’m not happy with any of them.
I’ve posted 155 entries this year, 854 since April 7th 2000, when I started using Blogger.
Since mid last year, when I started “Blogging” rather than writing a diary (The difference is small, but is mostly about awareness) I’ve written one single short story, and one new site.
William Gibson is right, Blogging interferes with writing. Burningbird was right, once you start blogging it’s [D]ifficult to focus on my books and my articles
. I now know the book I’m writing, I can see the plot in front of me. But all the time I’m writing this, I’m not writing that, and Aquarion The Website started out as a place to put my fiction online.
Don’t panic, I’m not going anywhere immediatly. I’m here at least until the house thing happens.
But after that?
We’ll see.
Categories
Not Cool
So, Aquarionics has been yo-yo-ing all afternoon, because some tin-pot server over in America decided it wanted to spider my site at a couple of hundred pages a minute.
Now, each of these articles on a page is made out of a few DB accesses. When you build a page, you have four DB queries per article, for the attachments, two for the crossreferences (logs and the xref table) and one for the comments. Plus one query that gets all the relivant items for display on the page. This is why the category archives, which can have up to five hundred items on a page, don’t show comments, attachments or crossreferences. Not that a query is much overhead, thanks to the wonder that is persistant connections.
Epistula’s Funky Caching (Originally billed as ‘Vinette style caching’) means that once a page has been built it’s saved and served as static content until the content changes, or the Admin deletes the cache. Last night I modified the page class that generates the HTML code, and so cleared the cache (This was for the referer stuff).
This afternoon, said tinpot server (Which actually belongs to AltaVista and therefore really should know better) grabbed 4000 pages in about half an hour. This had something of a knock on effect on performance because Sneaky was querying the database for [E]2 (And I had somehow disabled persistant connections at this point, so 5 connections per entry), as well as once per file downloaded (complete with all user information) to the apachelog database set up earlier this week
At this point, pol took down the webserver to find out what was wrong and install throttling stuff, but still, at a decent rate of hits (which Aqcom – which is the default virtualhost for the server – gets) the mod_sql_log was DoSing the box, as Apache wouldn’t serve what it couldn’t log. How the current level is killing it where yesterday’s wasn’t, I’m not quite sure. So pol has installed the ionCube Accelerator for PHP, and taken off the sql_logs for the time being.
Isn’t life fun?
Categories
Wibble
More audiobloggery.
- Lemmings!
Oh, and I finally got around to blogrolling Cathy, despite having been reading the blog for months. You can’t tell (The blogroll not being on this design, must…not..redesign…) but it’s there on my start page for all to see.