December 22, 2008
Formatting changes
I'm trying to experiment with the site formatting, so strange things are likely to occur for the next little while.
UPDATE: OK, finished. I just discovered that I had a typo in my CSS file.
Anyway, the change I just made was to make the active part of the page 30 pixels wider, which means that 800*600 pictures now display fully and don't bleed off the right side when you open individual posts. (They'll still be truncated on the main page. I'd have to add another 255 pixels to fix it there.)
Posted by: Steven Den Beste in Site Stuff at
09:41 PM
| No Comments
| Add Comment
Post contains 94 words, total size 1 kb.
December 03, 2008
Stack Overflow

This was curious. It turned out to be caused by ClustrMaps, so I've removed that from the main page.
Is SiteMap still evil?
UPDATE: It seems to have happened to me because our DNS went out. Strangely, I could ping the DNS with 100% return rate, but lookups all failed. I was able to load chizumatic because I had the DNS lookup cached, I think, but looking up ClustrMaps failed, and that resulted in this strange IE error. (Firefox didn't pop up; it just was stuck waiting for a response that never came.)
It took Comcast about an hour to get the DNS working again.
(Please note that this is not a request for advice or suggestions.)
So I went out for a walk. The creek is higher again, and I went down and looked at the dam. The beavers have completely blocked the main channel and raised the spillway there so that it is now dry. As a result the water is now running around it, over the sidewalk, and the water behind the dam is maybe 6 inches higher, which means it's flooding a larger area again.
UPDATE: Late last night I received the following from someone working at ClustrMaps:
(please copy/paste to your site if you want)
Hi-- it wasn't us... really! The DNS failure would have caused the same problem with any other resource needed externally, and you just happened to be pulling in ours. Our reliability rate is fantastic these days, and I'm wondering if you could kindly edit or 'strike through' the sentence 'caused by ClustrMaps', since, even with your update paragraph, a search engine trawl just now highlighted 'stack overflow caused by ClustrMaps', which is very off-putting to any prospective users!
In case it helps, I've also updated your account to ClustrMaps+ (better zoomed-in continent maps, no ads); you don't need to change the HTML or thumbnail map, which would stay the same).
Many thanks for your consideration!!
It's a bit scary that they noticed so rapidly, and that they thought I was important enough to respond to! Yikes!
This is just an itty bitty anime blog, and it doesn't even get to borrow all the Google cred I built up with USS Clueless, since it's on a different server, at an entirely different URL. (I don't even know how to check that, by the way.)
Regardless, I had already been thinking about removing ClustrMaps from the page, and I don't think I'm going to put it back. But I did want to give them their chance to respond. It would seem that the stack overflow is how IE responds to certain cases of DNS timeout, and it isn't specifically related to ClustrMaps.
Posted by: Steven Den Beste in Site Stuff at
12:00 PM
| Comments (6)
| Add Comment
Post contains 452 words, total size 3 kb.
Posted by: Pixy Misa at December 03, 2008 03:09 PM (PiXy!)
2
It's been long enough now so that I think the county really has thrown in the towel. Probably a hundred saplings they planted in the park have had their roots underwater for two months now, which must sure have killed most of them at this point.
Posted by: Steven Den Beste at December 03, 2008 03:44 PM (+rSRq)
3
(Please note that this is not a request for advice or suggestions.)
Just what or when is ever a request for advice or suggestions in your blog, Mr. Den Beste? =P
Do you ever catch the beavers doing their job?
Posted by: CronoLink at December 03, 2008 10:29 PM (v0ZDu)
Posted by: Steven Den Beste at December 03, 2008 10:35 PM (+rSRq)
5
I've had intermittent problems with Verizon's DNS. And they're fun, too, because I'll try to browse a page, and get a network error, and hit refresh, and the page loads. I've got a friend who runs his own caching DNS server, tho, so I made that my primary server. I'm thinking about changing it so that VZ is primary, and the friend's is secondary, in hopes that this will eliminate the problem due to the stack falling back properly.
The ACM has not evaluated this statement, and it is not intended to diagnose, cure, or treat any network condition.
Posted by: RickC at December 04, 2008 08:12 AM (8nFIS)
6
Is it wrong and horrible that, reading the title of this post, first thing I thought is "Hey, that happens in Divergence Eve all the time!"
*runs*
Posted by: metaphysician at December 04, 2008 10:12 AM (h4nEy)
Hide Comments
| Add Comment
November 19, 2008
This is a small bug-test. Please excuse
Google search bait the first: nosferatu__111
The idea is to see how long it takes Google to notice, and to see if it finds and indexes both search baits. My prediction is that it won't see the second one, which is below the fold.
more...
Posted by: Steven Den Beste in Site Stuff at
04:53 PM
| Comments (12)
| Add Comment
Post contains 57 words, total size 1 kb.
1
It's 731pm, Pond Standard Time, and no hits on either.
Posted by: Wonderduck at November 19, 2008 05:31 PM (hplPV)
2
The point of the first one is so that I can tell when the googlebot processes this entry. I'm not really too concerned with how soon that takes place; that isn't the point of this experiment.
When it does, what I'm betting is that it still won't see the second one. I think that the googlebot isn't processing our "more inside" sections, based on my inability to use Google to find some older posts of my own. I wanted to definitively prove or disprove that.
Posted by: Steven Den Beste at November 19, 2008 05:34 PM (+rSRq)
3
Just scrolling back through my posts to find some extended entries, and searching for them in Google.
This one worked and as you can see, it not only found the main post, but all the different archives and folders that it got posted into.
This one worked too, but found only two of the four locations for some reason.
This one worked fine.
Posted by: Pixy Misa at November 19, 2008 06:36 PM (PiXy!)
4
Pixy, the problem seems to be in "more inside" sections. I tried using this post just now as a test subject. That link goes to the dividing line; everything below that wasn't on the front page.
Searching for "Takuto is beginning to collect" hits. That text is above the fold.
Searching for "upper right corner supposed" fails.
What's curious is that the first one returns the link to the full body of that post, which includes the second string. Yet Google doesn't know about the second string.
Anyway, whenever Google gets around to finding this post we'll be able to prove it for certain. My prediction is that only the first of the two search baits will be in Google's database, but it will link to the full version of the post which includes the second one anyway.
Posted by: Steven Den Beste at November 19, 2008 06:44 PM (+rSRq)
5
I'm now officially confused. Why does it work for Ambient Irony but not for Chizumatic?
Posted by: Steven Den Beste at November 19, 2008 06:49 PM (+rSRq)
6
Anyway, I just downloaded eps 5 and 6 of Macademi Wasshoi so I'm going to go watch at least the first of them.
Posted by: Steven Den Beste at November 19, 2008 06:50 PM (+rSRq)
7
DON'T WATCH EPISODE 5!!!!!!!!!
Posted by: Pixy Misa at November 19, 2008 06:59 PM (PiXy!)
8
Well, it's only Don who's unhappy with Maca 5 and 6. Aroduc says it's been great.
Posted by: Pete Zaitcev at November 19, 2008 07:01 PM (/ppBw)
9
Aroduc is an evil pervert!
...
Well, I admit, the last few minutes of episode 5 weren't bad.
...
Pixy is an evil pervert!
But one with
taste.
Posted by: Pixy Misa at November 19, 2008 07:25 PM (PiXy!)
10
Now I'm even more confused. Google has found this post, and it also found both search baits.
So what the heck is wrong with my older posts that it doesn't see the "more" sections of them?
Posted by: Steven Den Beste at November 19, 2008 08:50 PM (+rSRq)
11
Mysterious are the ways of Google.
Posted by: Pixy Misa at November 19, 2008 08:53 PM (PiXy!)
12
I guess there are some things man was not meant to know.
Posted by: Steven Den Beste at November 19, 2008 09:08 PM (+rSRq)
Hide Comments
| Add Comment
November 13, 2008
Google cruft
The googlebot is hammering me again today. It's looking for stuff like this:
/Chizumatic/cref.shtml/Chizumatic/translation/translation/tmw/else/tmw/reviews/Shingu.shtml
/Chizumatic/cref.shtml/else/tmw/translation/translation/else/reviews/NinjaNonsense.shtml
/Chizumatic/cref.shtml/translation/nitpicks/reviews/tmw/else/tmw/ALDVD06.shtml
/Chizumatic/cref.shtml/Chizumatic/translation/nitpicks/else/translation/nitpicks/reviews/UFOPrincessValkyrie.shtml
None of those is a real path here; but the way my server was set up before they'd all return the "cref.shtml" entry (the second item in the path) and it's loaded with sub-paths. (It's the index of all my reviews.)
Last time this happened I had to block the googlebot in my firewall for a while. Then one of my readers showed me how to set it up so that my server would return a 503 for all those paths.
What I'm hoping is this: that all those idiotic paths are ones the googlebot has in its history and it thinks it is supposed to visit again, and that when it does this time and gets a 503, that it'll tag the path as a bad one and never try to visit it again. hope hope hope
hope hope hope because it's been visiting bogus paths in that non-existent directory tree for the last four hours and it doesn't show any sign of stopping any time soon. I sure hope it doesn't treat the 503 as a temporary thing, and revisit all of them again and again over the upcoming weeks in hopes of finding out if the 503 has been cleared... (Maybe I should have set it up to return a 504, eh?)
It's getting kind of lonely around this part of the blogosphere these days. Shamus has carpal tunnel and may have to curtail his blogging. Wonderduck has medical problems in his family. Ubu has gotten infected with politics. And I'm not feeling so good, myself.
Posted by: Steven Den Beste in Site Stuff at
06:13 PM
| No Comments
| Add Comment
Post contains 278 words, total size 2 kb.
September 26, 2008
Gay? Gayer? Gayest?
Man, the things people put onto the internet. (via)
"wonderduck is officially gayer than chizumatic!"
Posted by: Steven Den Beste in Site Stuff at
12:58 PM
| Comments (4)
| Add Comment
Post contains 18 words, total size 1 kb.
Posted by: madmike at September 26, 2008 01:10 PM (mV2q+)
2
When it comes to blogs, the way this is rated tends to "reward" longevity. The Duck has been blogging under that name a lot longer than I have under "Chizumatic".
Posted by: Steven Den Beste at September 26, 2008 01:13 PM (+rSRq)
3
And Houblog is about
8 times gayer than Bridgebunnies. However,
this one is not even close--which is fitting, considering all the brokenness..
Posted by: ubu at September 26, 2008 01:23 PM (qfBA/)
4
The strange thing is that most of the hits for "gay" and "Wonderduck" have nothing to do with my blog. Oh, the first two are (one of which is your own darn fault, Steven), but after that? Not me.
Not that there's anything wrong with that...
Posted by: Wonderduck at September 26, 2008 03:42 PM (AW3EJ)
Hide Comments
| Add Comment
September 22, 2008
HTACCESS tech help, please
I reenabled the googlebot, and now it's trying to traverse that non-existent file tree again.
I want to add a rule to my .htaccess file. This is what I think it should say:
RewriteCond %{REQUEST_FILENAME} /Chizumatic/cref.shtml/.*$ [NC]
RewriteRule \.shtml$ /common/404.html [R,L]
Except that "REQUEST_FILENAME" isn't the right keyword to use in the first line. What I want is the request filepath, instead. Is "REQUEST_FILEPATH" the right keyword, or is it something else?
The Apache documentation for HTACCESS is a travesty, by the way. Just thought I'd mention that.
Posted by: Steven Den Beste in Site Stuff at
10:35 PM
| Comments (11)
| Add Comment
Post contains 90 words, total size 1 kb.
1
I've read comments to the effect that the documentation for Apache is mainly the code. Not trying to hijack the thread here, but this, of course, is one of the things people gripe about WRT open source. (And plenty of closed source software has the same problem.)
Posted by: RickC at September 23, 2008 06:13 AM (p1CHu)
2
I once got behind the 8-ball with my boss because he thought the MS Access '97 Users Manual was suffcient documentation to be able to write a program to parse text input, discarding bad results in VBA -- and I failed to do so. The episode contributed significantly to a stall in my career.
Software: the only "engineering" discipline wherein structural failure isn't just tolerated, it's expected.
As for the .htaccess I have even less idea, sorry. 
Posted by: ubu at September 23, 2008 06:43 AM (dhRpo)
3
Noticed something interesting today: I have been reading Chizumatic on and off for a week or so via Opera Mini on my WinMo smartphone. When I do, I get the "no bandwidth theft" logo instead of the banner image at the top of the page.
Posted by: RickC at September 23, 2008 06:55 AM (p1CHu)
4
Firstly I don't quite understand how .htaccess got involved into something that affects the whole document tree and should be in httpd.conf. Secondly, why use mod_rewrite at all? I would just match against a pattern using something like DirectoryMatch and then "Deny all". And finally, the root cause seems to be something in the configuration that generates the recursion in the first place. Why not fix that instead? Just a thought. I don't have a lot of experience configuring webservers.
Posted by: Pete Zaitcev at September 23, 2008 07:08 AM (/ppBw)
5
Firstly I don't quite understand how .htaccess got involved into something that affects the whole document tree and should be in httpd.conf. Secondly, why use mod_rewrite at all?
Because it's already on, in order to prevent LiveJournal and Fark users from deep-linking my graphics files.
I would just match against a pattern using something like DirectoryMatch and then "Deny all".
Good for you, but I don't know how to do that.
And finally, the root cause seems to be something in the configuration that generates the recursion in the first place. Why not fix that instead? Just a thought. I don't have a lot of experience configuring webservers.
Neither do I, and that's the problem.
Noticed something interesting today: I have been reading Chizumatic on and off for a week or so via Opera Mini on my WinMo smartphone. When I do, I get the "no bandwidth theft" logo instead of the banner image at the top of the page.
The top rotation picture is hosted on my own server, not on Pixy's. If you don't send one of a small number of explicitly permitted referers, you get the "don't steal bandwidth" image instead of the real top rotation picture.
Probably your browser, or your proxy, are lousing up the referer.
Posted by: Steven Den Beste at September 23, 2008 09:56 AM (+rSRq)
6
"The top rotation picture is hosted on my own server, not on Pixy's. If you don't send one of a small number of explicitly permitted referers, you get the "don't steal bandwidth" image instead of the real top rotation picture."
Yep, I knew that. I figured it was likely either a proxy error or something wrong with opera mini. If you have any interest in examining your logs, I hit the site today, last monday, and a couple of other days last week, around 5AM your time. (The only reason I mention it is if it turns out the referer is ok, I might file a bugrep with the Opera folks.)
Posted by: RickC at September 23, 2008 10:10 AM (fZUrq)
7
Except that "REQUEST_FILENAME" isn't the right keyword to use in the first line.
I think REQUEST_URI is what you want.
Or, for a one-liner, this should do it:
RewriteRule ^/?Chizumatic/cref.shtml/ - [F,L,NC]
The "^" means to match against the start of the string (so it wouldn't match /foo/Chiz... for example), and the question-mark is because depending on whether you're using this in an .htaccess file or in the .conf file, the leading slash may be implied -- I forget, so the question-mark makes it work either way. The "F" means "return a 'forbidden' error", and the "L" means "if this matches, then make it the last rule we process for this request."
Posted by: Anthony DiSante at September 23, 2008 10:49 AM (xJ4r5)
8
"REQUEST_URI"
Did you mean "URI" or "URL"?
Posted by: Steven Den Beste at September 23, 2008 12:57 PM (+rSRq)
Posted by: AnthonyDiSante at September 23, 2008 01:36 PM (xJ4r5)
10
Your one-liner seems to work perfectly. Thank you very much for your help!
Posted by: Steven Den Beste at September 23, 2008 02:56 PM (+rSRq)
11
You're welcome! I've been frustrated many times by that horrible Apache documentation, so I'm glad to share what I've learned to spare someone from having to spend any more time in those docs.
Posted by: AnthonyDiSante at September 23, 2008 03:46 PM (xJ4r5)
Hide Comments
| Add Comment
September 18, 2008
Googlebot
Look at this crap:
(In order to deuglify the front page, I'm putting this below the fold.)
more...
Posted by: Steven Den Beste in Site Stuff at
04:51 PM
| Comments (3)
| Add Comment
Post contains 182 words, total size 3 kb.
1
Sounds like that page should be specifically called out in your /robots.txt file. I know some search engines only pay attention to the first matching entry in robots.txt, so if it's already there, that could be why it still gets searched.
Yahoo once tried to index, in parallel, every recipe in my cookbook search engine, which wrecked my machine for most of a day, and it was because they only honored the first block in robots.txt.
-j
Posted by: J Greely at September 18, 2008 05:36 PM (9Nz6c)
2
Except that I want Google to see "/Chizumatic/cref.shtml" and everything beneath it, at least everything that's really beneath it. That's the index of my reviews. The problem here is that they've somehow gotten onto the idea that "cref.shtml" is a directory name, not a readable file.
Apache on my machine is ignoring everything after the second term of that URL, and returning the cref file every time.
Posted by: Steven Den Beste at September 18, 2008 05:54 PM (+rSRq)
3
I believe you'll get the right results if you include the trailing slash, e.g. "/Chizumatic/cref.shtml/", but you might want to test it with Google's
webmaster tools.
(side note: the "Allow: /" in your robots.txt seems odd; Google reads Allow before Disallow, so it should be a no-op, but the behavior doesn't seem to be defined)
-j
Posted by: J Greely at September 18, 2008 09:24 PM (2XtN5)
Hide Comments
| Add Comment
September 14, 2008
Doing chores
Sorry for not posting much. My project right now is to try to do another update to the top rotation. So far I've got:
95 Kirameki Project
54 This Ugly Yet Beautiful World
71 Melancholy of Haruhi Suzumiya
53 Early Reins
12 Green Green
34 Kiddy Grade
16 MoonPhase
I'm probably going to get a few from The Third and I may take a look at Ikki Tousen.
So I may end up with about 400 new ones. The current rotation is 753, so that would be a bit more than half. So what goes, to make room? I've already started a bit of trimming, but I've only gotten rid of about 90 images so far. Mostly that consisted of pictures from Seven of Seven which I find I didn't like seeing. But I'm going to have to get ruthless if I'm going to eliminate half the current list. Probably I'll get rid of most of the Gravion images that remain, and Yumeria can go. But there are a lot of things I don't want to lose (Shinobu, Rushuna) so it'll be painful. I'll probably seriously cut back on High School Girls and Magical Play and UFO Princess Valkyrie and I'll probably also have to trim a lot of Magikano. And some Ramen Fighter Miki.
Ah, hell; I have to delete a lot of everything. Maybe it would be easier to identify the pictures I want to keep, instead of the ones I want to eliminate. I'll think about that.
UPDATE: Deciding what to keep, instead of what to get rid of, worked a hell of a lot better. I've got it down to 305.
You know, it really shouldn't make any difference, but psychologically it isn't the same. Probably it's because there are pictures I do like, pictures I don't like, and pictures I don't really care about. If I do it as weeding-out, I only get rid of the ones I don't like. But if I do it this way, effectively I also get rid of the ones I'm neutral regarding.
Posted by: Steven Den Beste in Site Stuff at
11:20 PM
| Comments (1)
| Add Comment
Post contains 345 words, total size 2 kb.
1
Once images are pulled, do they ever reappear? I mean, do you ever think, "I should put some Happy Lesson pictures back up"? (Not a request; I'm just curious.)
Posted by: BeckoningChasm at September 15, 2008 02:45 PM (fnoZ9)
Hide Comments
| Add Comment
August 27, 2008
I thought they'd forgotten about me!
Good Heavens!
I'm being Slashdotted! (Of course, it's from a comment, and in a somewhat obscure sub-board, so my poor old server isn't being horribly hammered. But it's still strange to see.)
Posted by: Steven Den Beste in Site Stuff at
06:51 PM
| Comments (2)
| Add Comment
Post contains 38 words, total size 1 kb.
1
*runs from the shockwave of the San Fran flywheel*
*cough*
Posted by: metaphysician at August 27, 2008 06:59 PM (12xNi)
2
Well, that one's not set up to be linked to.
Posted by: Steven Den Beste at August 27, 2008 07:02 PM (+rSRq)
Hide Comments
| Add Comment
August 25, 2008
Not quite what I had in mind
My most recent post managed to seriously anger one of my regular readers. That isn't really what I want from this blog, so I've deleted it.
UPDATE: Pixy: bug? I didn't delete that post, I changed its status to "hide". It's gone now on the front page and links to it fail now, all of which is correct behavior.
But comments on that thread still appear in the right column on the front page.
UPDATE: Just to make clear: I deleted it because it made someone angry, and I don't want to make people angry. I apologize for crossing that line.
Posted by: Steven Den Beste in Site Stuff at
05:04 PM
| Comments (2)
| Add Comment
Post contains 108 words, total size 1 kb.
1
Yeah, that's definitely a bug. I'll get that fixed today.
Posted by: Pixy Misa at August 25, 2008 06:07 PM (PiXy!)
2
Odd. It shouldn't happen at all. I'll need to poke around a bit more.
Posted by: Pixy Misa at August 26, 2008 05:11 AM (PiXy!)
Hide Comments
| Add Comment
46kb generated in CPU 0.044, elapsed 0.1509 seconds.
46 queries taking 0.1288 seconds, 118 records returned.
Powered by Minx 1.1.6c-pink.