Patrick wrote:
"Don�t be surprised if for brief periods the entire site is in Chinese."

That would be pretty cool.

Utterly illegible to probably ninety percent of your readers, but pretty cool.

I sure picked the wrong week to give up sniffing glue....

Dave, I'm afraid.

I'm afraid, Dave.

Dave, my mind is going. I can feel it.

[insert sight gag here of a bunch of undisplayable chinese characters]

不過中文有那麼多興趣．

[Actual sight gag of Chinese characters, if your web browser does Unicode.]

Best of luck.

At least it's not in Middle English.

ما هذه كتابة صينية

Ожидание! Это не китайско!

Ya know, I had a feeling I'd come back here to exactly this sort of thread.

MT 3.33 installed. Seems to work so far. Fingers crossed.

What, no Greek? Armenian? Mayan?

Sun wah nutakô�hsee w�ôpanâô��tuw�âm?

Teresa, I doubt we can get you drunk enough.

� incr�divel, ma�s o sitio agora est� escrito na lingua galega. O esp�ritu de Rosal�a, cica�s, se apoderou das ordenadoras.

Not working so well. We can't publish new (or revise old) front-page posts; the error message is:

Content-Type: text/plain; charset=utf-8 Got an error: Can't call method "prepare" on an undefined value at lib/MT/ObjectDriver/DBI.pm line 103. at lib/MT/App/CMS.pm line 906

Ni hàú ma?

In my copy of that file, that has to do with author ids & permissions.

WAG: did you get an author associated with John M. Ford's posts in the database? (I see that he appears in the author line in some posts, but I don't know if you guys use static publishiing and just haven't republished those since the database problem earlier.)

Otherwise, my standard first-line troubleshoot is to re-upload the complained-of file, triple-checking that it's in ASCII.

Yes, pericat very kindly fixed the problem with Mike Ford's posts losing their byline.

The DBI.pm problem appears to be fixed, by the expedient of googling the error message to see who else has had the same problem.

Although, since your test post has been and gone, I take it you fixed it. So nevermind.

I'm glad it worked without extreme amounts of pulling-of-hair.

Whan that Septembre with hir nightes cool
Recall the children by hir bus to school,
When cattes hath not this weke vacuumed been
And Buffy reruns hath been lately seen,
And memories of woes past are tofaden,
Than longen folke hir software to upgraden.

(#10) Greek or Armenian (or Aramaic) are no problem, but Unicode doesn't have Mayan.
OTOH, you could have Balinese, Linear B, Cuneiform, or Shavian.

Don't know if it was an intended side effect, but although webiste links for posters still come through OK, email addrs do not.

1010010 001 00100011001001 010101101 01001000101 010101111 01010 10100101010 1010101001 010

but I digress

01010001001001 0101 01 0001001 01 0101 01 01 0 010101010 01 101 0100101 0 101010 101001 0101001 011 01000

I think you'll all agree with that logic.

ᚻᛁᛖ᛫ᛊᚳᚫ᛫ᚦᛖ᛫ᚻᚫᚱᛞᚱᚪ᛫ᚻᛖᚩᚱᛏᛖ᛫ᚦᛖ᛫ᚳᛖᚾᚱᛖ᛫
ᛗᚩᛞ᛫ᛊᚳᚫᛚ᛫ᚦᛖ᛫ᛗᚫᚱᛖ᛫ᚦᛖ᛫ᚢᚱᛖ᛫ᛗᚫᚷᛖᚾ᛫ᛚᚣᛏᛚᚨᛞ᛭

Well, the pages now load fast enough that I'm not distracted by something shinier and wander off to, say, clean out the meat drawer in the refrigerator.

Well, the grape-eating moose in the particles does seem to have lost some weight.

"Don't know if it was an intended side effect, but although webiste links for posters still come through OK, email addrs do not."

Nope, not intentional. And now fixed.

(?:G(?:[^\\G]*(?:\\.[^\\G]*)*)G|o(?:[^\\o]*(?:\\.[^\\o]*)*)o|o(?:[^\\o]*(?:\\.[^\\o]*)*)o|d(?:[^\\d]*(?:\\.[^\\d]*)*)d|\ (?:[^\\\ ]*(?:\\.[^\\\ ]*)*)\ |t(?:[^\\t]*(?:\\.[^\\t]*)*)t|o(?:[^\\o]*(?:\\.[^\\o]*)*)o|\ (?:[^\\\ ]*(?:\\.[^\\\ ]*)*)\ |h(?:[^\\h]*(?:\\.[^\\h]*)*)h|e(?:[^\\e]*(?:\\.[^\\e]*)*)e|a(?:[^\\a]*(?:\\.[^\\a]*)*)a|r(?:[^\\r]*(?:\\.[^\\r]*)*)r|\ (?:[^\\\ ]*(?:\\.[^\\\ ]*)*)\ |t(?:[^\\t]*(?:\\.[^\\t]*)*)t|h(?:[^\\h]*(?:\\.[^\\h]*)*)h|i(?:[^\\i]*(?:\\.[^\\i]*)*)i|n(?:[^\\n]*(?:\\.[^\\n]*)*)n|g(?:[^\\g]*(?:\\.[^\\g]*)*)g|s(?:[^\\s]*(?:\\.[^\\s]*)*)s|\ (?:[^\\\ ]*(?:\\.[^\\\ ]*)*)\ |a(?:[^\\a]*(?:\\.[^\\a]*)*)a|r(?:[^\\r]*(?:\\.[^\\r]*)*)r|e(?:[^\\e]*(?:\\.[^\\e]*)*)e|\ (?:[^\\\ ]*(?:\\.[^\\\ ]*)*)\ |p(?:[^\\p]*(?:\\.[^\\p]*)*)p|r(?:[^\\r]*(?:\\.[^\\r]*)*)r|e(?:[^\\e]*(?:\\.[^\\e]*)*)e|t(?:[^\\t]*(?:\\.[^\\t]*)*)t|t(?:[^\\t]*(?:\\.[^\\t]*)*)t|y(?:[^\\y]*(?:\\.[^\\y]*)*)y|\ (?:[^\\\ ]*(?:\\.[^\\\ ]*)*)\ |m(?:[^\\m]*(?:\\.[^\\m]*)*)m|u(?:[^\\u]*(?:\\.[^\\u]*)*)u|c(?:[^\\c]*(?:\\.[^\\c]*)*)c|h(?:[^\\h]*(?:\\.[^\\h]*)*)h|\ (?:[^\\\ ]*(?:\\.[^\\\ ]*)*)\ |s(?:[^\\s]*(?:\\.[^\\s]*)*)s|o(?:[^\\o]*(?:\\.[^\\o]*)*)o|r(?:[^\\r]*(?:\\.[^\\r]*)*)r|t(?:[^\\t]*(?:\\.[^\\t]*)*)t|e(?:[^\\e]*(?:\\.[^\\e]*)*)e|d(?:[^\\d]*(?:\\.[^\\d]*)*)d|\ (?:[^\\\ ]*(?:\\.[^\\\ ]*)*)\ |o(?:[^\\o]*(?:\\.[^\\o]*)*)o|u(?:[^\\u]*(?:\\.[^\\u]*)*)u|t(?:[^\\t]*(?:\\.[^\\t]*)*)t|\!(?:[^\\\!]*(?:\\.[^\\\!]*)*)\!)

And memories of woes past are tofaden

I think that last verb is, um, not misconjugated, but the prefix doesn't seem to fit it. I'd probably go for 'yfaden' myself. (Now that I've stopped giggling!)

I'm still getting comments threads opened in the middle, rather than on the comment I clicked on in the comments listing on the main page. (Just so you know.)

I wonder what that post from xeger was supposed to be?

Nancy, what you describe is an inevitable consequence of the method Dori Smith used to give us user-choosable style sheets. If it's intolerable, the only alternative is to stick with the default itty-bitty type.

As I've been saying with increasing frequency, the technical maintenance of Making Light is now officially beyond my skills.

:Raises a glass in the general direction of Jim MacDonald:

נו? יש "בלוג". יש אור.
מזל טוב

Well, damn.

It worked pretty well in the preview.

נו? יש "בלוג". יש אור.
מזל טוב

Interesting. One preview translated to unicode, which looked OK. But editing the text and previewing again, while it looked OK in the preview, resulted in, well, "מזל טוב", in the actual post.

And in previewing it again, it looks right in preview, but I bet it will turn into that strange mixture of Czech, Icelandic, and mathematical notation in the actual post.

Movable Type is officially weird.

Ni hao!

Is it just me, or is the "View all by" broken?

After fiddling about with various diagnostics, I think I figured it out:

While there's a
<meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1">
header in the page itself, both in the regular page and the preview page, the preview page, which is generated by "qlbr.cgi", apparantly throws in an "UTF-8". That is, the HTTP header includes this:
(Content-Type: text/html; charset=utf-8)
Which apparantly overrides the content-type in the HTML page header itself.

So actual unicode (UTF-8) characters will look right on preview, but be turned into chicken scratches (iso-8859-1) on the regular page.

Nielsen Haydens, you might want to modify the site to be uniformly UTF-8 throughout, for the sake of consistency, so that people playing silly buggers in different character sets will see the same thing throughout.

Or not. Your choice.

Yes, "view all by" is broken, as discussed in several other threads. It's been broken since our hosting provider moved us from one server to another last week.

Test comment, for arcane purposes.

Look! Arcane porpoises!

#31 ::: Patrick Nielsen Hayden mused:
I wonder what that post from xeger was supposed to be?

That's actually what it's supposed to be... I gave in to a regex.

I wonder what that post from xeger was supposed to be?

Vg nccrnef gb or n erthyne rkcerffvba juvpu zngpurf 'dhbgrq' fgevatf jvgu onpxfynfu rfpncrf, jurer gur dhbgrf znl or nal punenpgre bppheevat va gur cuenfr "Tbbq gb urne guvatf ner cerggl zhpu fbegrq bhg!".

.i lo nu di'u mifra kei .e lo nu dei fange te bangu kei se mukti lo nu ri mapti ja'o lo casnu poi dei pagbu ke'a

xeger's post looks like a very long regular expression written in Perl's extended pattern notation. I think it matches the phrase "good to hear things are pretty much sorted out"

(For the record, my previous comment was "But Chinese is so interesting.")

Xopher, wo hao, xiexie ni.

Are the arcane porpoises major or minor?

I just want to say that when I posted #4 above:

[insert sight gag here of a bunch of undisplayable chinese characters]

I was relying on intelligence from 16 different agencies and a guy named "Curve Ball" who assured me that the remark would be greeted with flowers and chocolates and the entire thread would become a western-style democracy within 5 to 10 hours.

#48 ::: Michael Weholt added:
I was relying on intelligence from 16 different agencies and a guy named "Curve Ball" who assured me that the remark would be greeted with flowers and chocolates and the entire thread would become a western-style democracy within 5 to 10 hours.

So sorry.
Your post has been refused
by the great firewall of china.
Please, try again,
some other day.

Xopher, wo hen hao, ni ma?

BTW, just wanted to mention that if you wanted to buy yue bing this is the week to do it. The Moon Festival is this coming Friday. At my local 99 Ranch Market, they'll be running out of the good ones pretty soon.

𐑲 𐑣𐑨𐑝, 𐑝 𐑒𐑹𐑕, 𐑤𐑷𐑙 𐑢𐑭𐑯𐑑𐑧𐑛 𐑨𐑯 𐑧𐑒𐑕𐑒𐑿𐑕 𐑑 𐑐𐑴𐑕𐑑 𐑦𐑯 𐑖𐑱𐑝𐑾, 𐑯 𐑯𐑬 𐑲 𐑕𐑰𐑥 𐑑 𐑣𐑨𐑝 𐑓𐑲𐑯𐑩𐑤𐑰 𐑜𐑪𐑑 𐑢𐑳𐑯. 𐑔𐑱𐑯𐑒𐑕, 𐑐𐑨𐑑𐑮𐑦𐑒!

Okay. So, now we know MT doesn't like Shavian.

Oh, dear, and I'm going to Seattle tomorrow... on the other hand, do I want to go to Uwajimaya when the Mariner's are playing their last game of the season? Maybe I'll just go to the Arboretum bulb sale as previously planned and leave the mooncakes until some later year.

Purely geeky here - but I'm doing my best to keep my itchy fingers away from excessive ciphertext...

R29vZCB0byBoZWFyIHRoaW5ncyBhcmUgcHJldHR5IG11Y2ggc29ydGVkIG91dCE

hmm. the livejournal feed isn't showing up on my friends page. is this just me?

περὶ ποιητικη̂ς αὐτη̂ς τε καὶ τω̂ν εἰδω̂ν αὐτη̂ς, ἥν τινα δύναμιν ἕκαστον ἔχει, καὶ πω̂ς δει̂ συνίστασθαι τοὺς μύθους εἰ μέλλει καλω̂ς ἕξειν ἡ ποίησις, ἔτι δὲ ἐκ πόσων καὶ ποίων ἐστὶ μορίων, ὁμοίως δὲ καὶ περὶ τω̂ν ἄλλων ὅσα τη̂ς αὐτη̂ς ἐστι μεθόδου, λέγωμεν ἀρξάμενοι κατὰ φύσιν πρω̂τον ἀπὸ τω̂ν πρώτων.

Well, arcane porpoises are better than cross porpoises.

Sorry, Patrick, I don't usually read the technical threads in depth, so I must have missed it.

Test:

Don�t be surprised if for brief periods the entire site is in Chinese.

Patrick, I didn't know if you had tried to fix it, or considered it a problem at all. I don't have a problem with it; I was just hoping to give a data point that might help.

I would put up with a lot of weird formatting to read/ hang out at this site.

#59 ::: Nancy C wrote:
I would put up with a lot of weird formatting to read/ hang out at this site.

Heh. Yes - this is one of the few places where I'm actually willing to do technical work[0] (if called upon [1]) to help sort out whatever the problems are.

[0] This often being thankless (which is okay - getting it working is what makes it worthwhile) and irrationally full of psychodrama (which isn't).
[1] To clarify that psychodrama is nicely disemvoweled here, and that I'm not speaking about here :)

Interestingly, it kicked out a few older posts to the Electrolite feed that I never bothered to drop from bloglines. Same thing happened to whiterose.org's blog collection when I did this upgrade. I also took the opportunity to upgrade to the standard xml/rss and atom feeds so that I wasn't publishing broken stuff anymore, so all good!

Um. I'm disturbed to see that the LJ syndication indeed isn't working, as comment #55 suggests.

It's here. I don't know how to fix it. More to the point, I don't know what's wrong; it seems to be using our Atom feed URL, which it appears to have right.

To those reading this via the LJ feed: the feed name appears to have changed. It was "makinglight", and now it's "making_light" with an underscore. Use the LJ search function to find the new feed page and add it to your friendslist, and you'll be fine.

Looking at the profile page, I see:

Syndication Status: Last checked: 2006-10-02 01:25:55 (Too big)

And I also see on the syndication page itself that the last two items picked up were "John M. Ford, 1957-2006" and "Sign your organ donor card". Jim's great big "Mike Ford: Occasional Works (Pt. One)", which should be in between those two, isn't. So maybe LJ is choking on those giant posts. Or maybe it's something else.

Hm. I think Lee may have a better angle on it than I do. But then, I don't read Making Light thru LJ.

Another thing: The note above the comment box says that comments containing more than seven URLs will be held for approval. But I just had a couple posts held that only had two URLs each.

Something horrible appears to have happened to Padraig O Mealoid, between Ulrika O'Brien and Chad Orzel.

Hmmm, posting a test here. I was trying to preview a post with URL on the Rumsfeld thread and getting a "Connection has timed out" message 5 times in a row, at intervals.

That message and this one seem to be finally getting through the preview...

Clifton, I think it was down.
Back to the comment I was trying to post at the time....

There's a problem with some characters: see the Martian apostrophe in #58. It isn't, strictly speaking, a problem with MT. I've seen it other places, including with my favorite genealogy program. Characters with diacritical marks seem to be expecially favored for transliteration into Martian. (Assumptions about character sets, probably.)

#68 ::: P J Evans wrote:
There's a problem with some characters: see the Martian apostrophe in #58. It isn't, strictly speaking, a problem with MT. I've seen it other places, including with my favorite genealogy program. Characters with diacritical marks seem to be expecially favored for transliteration into Martian. (Assumptions about character sets, probably.)

It's a bit off topic here, but what is your favourite genealogy program?

Trying to ascertain the intent of a poster elsewhere, I discover that "view all by" leads to a blank page.

The "Martian Apostrophe" in #58 is indeed the end result of a long chain of character set mistakes.

I don't know what it was originally, but at some point it was transliterated into Unicode, but the transliteration thought it was an invalid character so converted it to U+FFFD REPLACEMENT CHARACTER. That was then stored in its UTF-8 form, 0xEF 0xBF 0xBD, and was output as such. The output stream is marked as ISO-8859-1, though, so it's now interpreted as characters 0xEF LATIN SMALL LETTER I WITH DIAERESIS, 0xBF INVERTED QUESTION MARK, 0xBD VULGAR FRACTION ONE HALF.

xeger, I've been using 'Rootsmagic' for the last few years; before that, it was 'Family Origins' - same people, different name. It's very user-friendly. The website-building routine produces nice HTML code, too. It puts the date on the backup, so you can only overwrite one day's files - nice touch!

72 ::: P J Evans replied:
xeger, I've been using 'Rootsmagic' for the last few years; before that, it was 'Family Origins' - same people, different name. It's very user-friendly. The website-building routine produces nice HTML code, too. It puts the date on the backup, so you can only overwrite one day's files - nice touch!

Ah, okay! I was thinking of the opposite end of the spectrum, for the sorts of tools that places like FreeBMD or the Later Day Saints archives use for data entry.

Xeger, I have no idea what those people are using for data entry. I just hope they're doing decent quality control. (I'm not judging them by Ancestral File, which is not as reliable as it should be and is anyway user-submitted. I've seen circular linkages in there!)

#74 ::: P J Evans wrote:
Xeger, I have no idea what those people are using for data entry. I just hope they're doing decent quality control. (I'm not judging them by Ancestral File, which is not as reliable as it should be and is anyway user-submitted. I've seen circular linkages in there!)

I don't know about the LDS (beyond "too much of it in the sixties"), but FreeBMD has pretty decent quality control based on what I know about the backend technology (no idea what people use to produce the files, however, which is why I was curious)

Claude Muncey in #70 noted that "View all by" is broken.

It appears to be broken (whether still or again, I don't know) this morning. It shows a header followed by a blank page.

Did you ever sort out the problem with the site's LJ feed, Patrick, or did it sort itself out? I ask because my blog's syndicated feed has started coughing up a 'too big' error, and I'm not sure what, if anything, I should do about it.

I don't know about your LJ feed, but your Atom feed isn't feeding to my newsreader anymore. I don't know about anyone else's, but my newsreader hasn't picked up anything since Sept. 27.

MKK

Mine isn't working either. I tried to get my wordpress feed into LJ:

http://syndicated.livejournal.com/clintwordpress/

but it says that when it checks, it is "too big":

Syndication Status: Last checked: 2007-09-26 01:25:55 (Too big)