Solipsism Gradient

The Solipsism Gradient is reviewed at PageBoost:

“I just saw /bb/viewtopic.php. First-rate.

(…)

It must have taken a decade to perfect the page. The URL has 39 characters. This length scores best in usability studies. Seeing Solipsism Gradient, I’m simply so inspired. What a splendid page! The color scheme is impressive.

If only my aunt would have a cool page like that. Well, I expected the creator to achieve only the best. Simply stunning. The page contains 669 links, a stable amount. There are 106,092 characters in the code, which is a swell length for the Firefox browser. The HTML is highly accessible. What a wonderful, wonderful web page.”

— Max Williamson, Daily URL

You too can have your weblog reviewed! Thanks to John Walkenbach for this.

Humor

From Boing Boing:

If you’ve ever had an existential moment and wondered if you were possibly a brain in a jar being fed an elaborate simulation, this is the gizmo for you…

The Solipsism Gradient is on the high side today. I could use an extra CPU or two, certainly.

Solipsism Humor

You may have noticed some dozen trackback spam entries here, on various topics, over the last week. Had a merry time deleting them all, as doing consecutive deletes uncovered a bug in my post-deleting patches… I’m still undecided on whether I should spend time on fixing the bug or on perfecting the workaround I found. In any event, if you’re the unmentionable who’s doing this, your links won’t work, as they were all being tagged with rel=”nofollow” anyway. Which I implemented as soon as it came out. So you might as well give up.

On a similar note, attacks from Santy and its descendant worms are decreasing daily, since Google now seems to be actively blocking its searches. So that’s one less thing to worry about, I hope.

Meanwhile, the Creation Robot has posted a list of The Top 100 Mac OS X Applications and Nudge is on it. Thanks! Now… why aren’t my other apps on it…?

Work on RBSplitView is proceeding well; besides implementing several suggestions from users, I hope to finally have animated collapse/expand in place. So far the method of collapsing a subview has had to be changed completely; formerly I was moving a collapsed subview off screen, keeping it at its original size; now I’m keeping it in place, collapsed to a zero dimension, and jiggling other stuff around to keep it from squishing its own subviews out of shape. Once that is done, animating various intermediate stages should be easy. I think.

Hm, there’s more meta stuff. My Technorati tags now finally seem to be working; I overlooked that they have to be implemented in the RSS feed also, and have to look “just so”. A very interesting experiment. They’re also producing RSS feeds for tag updates, which should make tags very useful once adoption spreads.

Speaking of RSS feeds, my subscription list (on the left) has been updated once again. My stats show that only a third of you are reading this weblog over RSS; I wonder why, since other bloggers report much higher ratios, some as high as 90%. I’ve been asked why I don’t have a Atom feed; now that the final spec seems to be well on its way, I’ll do it as soon as I find some spare time.

A full month on DreamHost is now over and I’ve been looking at their comprehensive statistics and access logs.

Several interesting facts stand out. There have been unexpected traffic peaks on Jan. 17, 25 and 26. A significant number of requests came from wb5.stanford.edu (171.64.75.198 (apparently some sort of webcrawling robot called WebVac?), but they’ve been off the air again lately. Only 29% of you are reading this over a RSS reader, and most of you are using NetNewsWire. The most popular browser (33%) is Safari, but about 20% are still using Internet Explorer. Mac users are, unsurprisingly, more than twice the number of Windows users.

The overwhelming majority of referrals came from Google and similar index sites, while the rest was equally divided between VersionTracker and the link I placed on Martin Hering‘s site because of the XRay name mixup. There have been several attempts at referrer spam which of course didn’t work, as I don’t publish automatic referrer lists.

A couple of thousand visits came from search engine users searching for the expected terms: “x ray” and “xray” – strangely, more for the first variation, which I never use myself. A surprising number of visitors came from pages cached by Google, and another significant fraction was searching for “viewtopic.php”. The latter search is well-known to be used by the Santy/Spyki worm. In fact, the various variations of these worms still seem to be active, and were responsible for 44.4% of all accesses to my website! Still, worm attacks have dwindled to 15% over the last few days, so I hope this will die down soon. In any event, nearly all such attacks are being rejected outright by some mod-rewrite magic and the few that may come through will have no effect, as I did the necessary updates as soon as the first attack happened last December.

Some comment spammers have appeared too, and I promptly deleted their comments. I also implemented the new nofollow tag which is now automatically and retroactively applied to all links posted by non-registered visitors, so they won’t have any effect.

I’m also seizing the opportunity to patch little defects here and there, but much still remains to be done…

Last Saturday I was fiddling around with software delivery options – in particular, BitTorrent – and learning a little bash and Python in the process.

Although the tests were mostly successful, I still need some more time make the system fail-safe. In the meantime, yesterday evening I noticed that I hadn’t gotten any e-mail – even spam! – over the whole day; an extremely unusual event, since my daily average varies between 50 and 200 e-mails, and would be much more had I not early learned to subscribe to mailing lists in digest mode.

To make a long story short, my tests had conflicted with some other tests I’d conducted earlier in the month, and space on my trusty provider had expanded beyond the now-too-small limit I had set previously and promptly forgotten about. The duty neuron has been adequately chastised.

Therefore, if you sent me e-mail between Jan. 16 and 18, and it bounced back with “maildir delivery failed: error writing message: Disk quota exceeded”, kindly send it again? My apologies.

BitTorrent Python <- These? Ah, I just implemented support for Technorati Tags. Click on one of these words to see more weblogs mentioning these topics.

The source code page is up, with version 1.0 of RBSplitView as the first item. I’ve also opened up a support forum for source code. Enjoy!

After some very tense days, I’m up and running on the new hosting provider: DreamHost. Prices are good, support seems to be fast and quite good, and there’s lots of options. More if you buy before year’s end; if you do so, be sure to mention my domain name as the referrer, or click through from the link above; I’ll get a modest commission.

The Santy worm, now apparently renamed Spyki, is still pounding the servers in ever-new combinations. Still, the security hole remains closed and I have added some new stuff to avoid having the server respond to the worm, which should ease the load quite a bit. If you have seen the site responding very slowly or even not at all during the last two days, it was a hardware problem and should be fixed by now.

Some other things are still shaking out but everything should return to normal on January 1st. In particular, Bill Bumgarner has tested RBSplitView and made some interesting suggestions; so have several other developers. It seems I should put up a new “source code” page with that and other goodies, so that will be my first priority.

Rafael Fischmann wrote:

It’s not that strange, Rainer… you have changed your phpBB settings to redirect to this weblog page whenever a user access your viewtopic.php page, something that doesn’t happen when you access viewtopic.php in a default phpBB installation.

That’s not what I did; I just made my weblog topic the default topic. Since 99% of accesses to my forums are to my weblog page, that should make little difference – at least in terms of traffic. Then again, Google doesn’t measure traffic.

Anyway, since the topic argument is not a part of the URL proper, but of the arguments (the part after the ‘?’), Google shouldn’t consider that for relevance either…