These are quick notes taken by Sean B. Palmer on the Semantic Web, Python and Javascript programming, history and antiquarianism, linguistics and conlanging, typography, and other related matters. To receive these bits of dreck regularly, subscribe to the feed. To browse other months, check the contents. This file was generated from plain text source, for convenience of posting, so apologies for all the in-your-face URIs.
Idea: create a microlanguage that uses features that are really against my aesthetic for a language. For example, a kind of polynesian orthography and texture, agglutination, a huge gender system, and so on.
This sort of thing isn't new... isn't Klingon basically along those lines? But what I'm testing out is whether creating a language feels similar to learning a language. When I learn bits of languages that on the surface don't appear all that cool to me, the "awesome, I'm learning this" reflex tends to take over and I enjoy it. I wonder if a similar thing will happen with a language that I'm designing on purpose to be, to me, ugly.
Some things I'm thinking about:
The /whits/ thing is mainly done, but I'm a little worried about moving the feed and putting a redirect in on the old one... I'm not sure how many feed agents will break on that.
Okay, I'm going to move Whits to its new home. There are probably only three or so people subscribed anyway, so I can just send a communique to the Swhack mailing list. As well as huffmanising the URIs, I'm also updating the software and so on.
Since bia went down, this is all being done on manxome, which is quite nifty in testing that everything still works across the machines. Having two friends' servers to use for this kind of thing is really nifty!
Okay, the transition should be complete. For some reason, Google Reader won't load the new feed in place, so I'll just have to hope that the redirect works; when these posts go up tomorrow we should find out!
I've been working on documenting some patterns, stuff like "ContentRichDesign" and "EverExpandingScope". At first I just listed the names in a single text file, and I was thinking about enumerating them in a TiddlyWiki, but eventually I decided to parse them out into separate HTML files, which I'll then run through a templating CGI.
Anyway, I did that, but due to the naivity of the thing I used to parse out each of the minilists into separate HTML files, I ended up with a lot of files that had simply '<ul>\n</ul>\n' in them. To filter these out, I just did:
$ for fn in $(grep -L '<li' *.html) > echo >! $fn
Now I'm going to go through and document them minimally...
http://polaris.gseis.ucla.edu/jrichardson/dis237/letterforms.jpg
- Oh man, a typography joke comic strip
It's been an afternoon of studying art nouveau. It eventually morphed into more abstract forms as art deco, but I think that the original seeds of the movement especially in Mucha could have a lot more mileage. The few bits of contemporary art nouveau that I found are intereting in that they look nothing like art nouveau to me: they have a kind of harshness and cynicism that just isn't present in the organic forms of the originals.
http://d8uv.org/flog/2007-01#t1167972122
- Cody has The Worst Day Ever. Ouch, man.
I'm about as good as Eeyore at cheering people up, but all the same I think that some figlet might be just what's called for here:
__
_ / /
(_) |
_| |
(_) |
\_\
Also, at least his internet connection must've come back (and bia's back too, which is almost as essential as an internet connection).
Before I start moaning about dircproxy, I should probably moan about trying to check out OLPC first. I'd read this CNN article, likely via Slashdot:
http://www.cnn.com/2007/TECH/01/02/hundred.dollarlaptop.ap/index.html
- Novel software drives '$100 laptop'
And the bit which really caught my attention was "the XO machines are organized around a 'journal,' an automatically generated log of everything the user has done on the laptop. Students can review their journals to see their work and retrieve files created or altered in those sessions." I'm always moaning about the current file/open/save paradigm in computers, so this alternative seemed interesting.
Now, I'd seen demos of the OLPC user interface, Sugar, before, and Javier had even been talking about installing the OLPC drive images and getting it to work, but I hadn't paid much attention before this. But this time I got on the trolley:
http://wiki.laptop.org/go/OS_images -> http://wiki.laptop.org/go/OS_images_for_emulation -> http://www.kberg.ch/qemu/ -> http://olpc.download.redhat.com/olpc/streams/development/LATEST-STABLE-BUILD
In other words, I read the intstructions for trying out the OS images on OS X, got QEMU, the emulator for the images, installed that, got an image, and then tried the image in QEMU. That version of QEMU didn't work, so I downloaded the one that was actually linked to on the page...
http://kju-app.org/
- Q.app
This worked to some extent, except when it got to the login phase: for some reason I couldn't interact with the mouse or the keyboard, even though up until it wanted me to put in a username it had been working fine. I wondered if that was related to this problem, noted on the OLPCWiki:
"Their documentation says something about using the USB Tablet to emulate the
mouse, no idea what this is about. The result is that I do not get a keyboard
in OLPC on my G4 Mac (USB Keyboard, obviously). I assume User:bert got it to
work using his PowerBook because PCMCIA emulation works?"
- http://wiki.laptop.org/go/OS_images_for_emulation
At any rate, I figured that I didn't want to bother debugging that and that burning one of the LiveCD images would be better. So I got one of those, and read about burning an ISO on OS X, since this is the first time I've done that:
http://www.macosxhints.com/article.php?story=20060619181010389
- How to burn ISO disc images
Then I took it to my Windows box since I didn't want to reboot, but it came up as saying "Invalid or unsupported executable format", which I found out was a strange GRUB error. I'd been using the old stable LiveCD image, so I tried using the development one instead, and burning it on Windows instead of OS X, but I got the same error from GRUB.
So instead I followed these instructions for Windows:
http://tuttlesvc.teacherhosting.com/wordpress/?p=251
- Emulating the OLPC XO On Windows
Installing the VMware Player and getting the zipped image, and that actually worked for a change (hurrah!) so I got to play with it... but there's no sign of journalling in that image so it was a bit of a let down after all!
On the other hand I did get to play with Sugar, and since I'd pretty much forgotten all of the details of the walkthrough I got to basically learn it from scratch like the children using the laptops would have to do. I found that it was rather easy to learn it (obviously I have many years' experience on computers in general to leverage off of, but still), though there were a few problems:
On the other hand, there's probably no harm in the facelessness of everything since it's going to be just as steep a learning curve no matter what. The main important thing is to keep it simple, and by goodness they've hit the nail on the head there. I really like the simplicity... you can chat, you can write stuff, and you can browse the web. And when journalling gets put in (this was an old image; perhaps it's already in the latest development versions, though the documentation for the feature seems sparse anyway), it'll probably be even niftier.
Some information about journalling:
Sorry, you'll have to stick the URIs together again yourself. I should work on some way to fix that... Using < and > would probably work. I'll jot the delimiters around there for now and then plan to add them to the Whits parser.
So bia, the server that Christopher Schmidt kindly lets me use, went down for some arcane and involved reason I won't get into here. It's the server that I run dircproxy off of, so when bia went down, so did dircproxy. I always just join #d8uv.org meanwhile until it's fixed. Sadly, when bia came back, dircproxy did not.
I keep getting this error:
17:12 *** Looking up your hostname... 17:12 *** Got your hostname. 17:12 *** Looking up irc.freenode.net... 17:12 *** Connecting to irc.freenode.net port 6667 17:12 *** Connection failed: Invalid argument
This was with dircproxy 1.0.5, so I decided to download and install dircproxy 1.2.0 beta2 from here:
http://dircproxy.securiweb.net/pub/1.2/dircproxy-1.2.0-beta2.tar.bz
And install it into ~/local and see if that works. I had to update my config file (chan_log_timestamp is now just log_timestamp, chan_log_copydir is gone, and other_* seems to have become server_*), but then I got the same error messages from it.
Chris told me that "Somewhere between creating the network socket and connecting to it, something goes bad with it, and I can't figure out why." The debug information that dircproxy was spewing was:
local/bin/dircproxy: connect(irc.freenode.net) failed: Invalid argument local/bin/dircproxy: net_close() failed: bad socket provided
Beyond that, I've no idea what could be up with dircproxy since it was of course working fine before the server crash... On the other hand, connecting to Freenode using irssi on bia works fine. Bit of a mystery.
I'd try out dircproxy on manxome, the server that Aaron kindly lets me use, but that's down too due to a disc space problem, and it'll probably be a little while until it's fixed since it's Saturday (poor Aaron; it always seems to go down on a Saturday).
So anyway, I'm just hanging out in #d8uv.org until somebody figures it out.
I'd really like my Google Search History to be downloadable, a la Gmail and its POP3 interface. Even if it was just a raw dump of the whole history in CSV or something, and even if I had to pay a pound or two for it... damn, just gimme my data!
Obviously they've got a million other things to do, and the more obvious solution would be to use a proxy myself and log all of my URIs... I keep meaning to work on Slogger, updating that awesome but sadly broken extension and enrichening it with even more awesome, but it'd be quite a task.
I've been looking for a decent light pollution filter. It seems that Neodymium filters are where it's at...
http://www.nezumi.demon.co.uk/nonad/spectra.htm
- Street Lamp Spectra / Light Pollution Filters
http://www.nezumi.demon.co.uk/nonad/nonad.htm
- NoNaD - A useful filter for Sodium light
And there's a decent looking Neodymium filter available here:
http://www.swoptics.co.uk/view.asp?KEY=1133
- Baader Planetarium Neodymium Moon and Skyglow Filter 31.7mm
And it's only 26 quid, but I'm not really sure how good it'll be; I'd mainly want to use it for naked eye observing, actually. There are some reviews:
The latter saying that "it does a good job, not spectacular", which seems fair to expect from such a cheap filter. I suppose I'd really like to try before I buy, all the same. Perhaps I should just buy one from some place that has a decent returns policy.
Once again, the funniest stuff in my aggregator this morning is from Patrick Hall. There's just some way he has about looking at stuff which is like in the uncanny valley of linguistic humour... he's all about quite serious nitty gritty stuff, and yet completely batshittedly off the wall at the same time:
http://ruphus.com/blog/2007/01/06/could-lol-ever-become-a-real-live-word/
- Could "lol" ever become a real live word?
http://blogamundo.net/dev/2007/01/06/the-truth-about-cuneiform/
- The Truth about Cuneiform
The truth about cuneiform is especially awesome (lol). I just don't know how he does it. Inkscape warped his mind, perhaps?
For whatever reason, dircproxy would not be fixed. I installed it on manxome instead, and now all is well again. Still, weird.
So I've been tinkering with a conlang for a while now, and I have a few observations on the making thereof:
I'm still also thinking about creating a language which uses choices that I don't like, mainly to see whether creating a language feels the same as learning a language.
Idioms, fossils, and sub-word morphemes show that at the lexical to semantic mapping level, words aren't the be all and end all of the matter. That's kinda cool, but it means that actually documenting your own language can be extremely difficult in some ways. In trying to be as descriptivist as possible, I'm finding that traditional ways to describe a language just aren't working for me (for example, traditional parts-of-speech categories).
So, which will be deciphered first: the Khipu, or Linear A? I asked Pat and Chris this last night, and they kinda felt that Khipu would fall first, because Linear A has been worked on first for ages with not much progress, but also that it'd depend on the corpus size most likely.
Ooh, the place where a language's speakers originally came from is called the "urheimat": http://en.wikipedia.org/wiki/Urheimat
I've been thinking about using Hangul in my language-of-unchoices, since one of the major choices that I'd make is of using the latin alphabet. And actually, Hangul is very cool indeed... it's still basically alphabetic and yet it doesn't look like it because it's syllable based. Plus it'd be easier than making up my own writing system, and it'd make it easier for me to read Korean should I ever want to learn it.
http://www.langintro.com/kintro/first.htm
- An Introduction to the Korean Alphabet
By J. David Eisenberg. It's licensed under CC by-nd-nc, it's simple, straightforward, and absolutely excellent. Hangul turns out to be even cooler than I'd initially thought!
Two criticisms:
Apart from that though, it's very good. It asks to provide feedback to the author after the first few lessons, so I've already done that.
This is good design, in zsh:
$ ls -al ... -rw------- 1 sbp sbp 1387 Aug 22 03:46 mimulus-icon.png -rw-r--r-- 1 sbp sbp 735 Aug 22 03:59 mimulus-off.png -rw-r--r-- 1 sbp sbp 699 Aug 22 03:59 mimulus-on.png ... $ chmod 644 mimulus<tab>
And it automatically completes it with "mimulus-icon.png". Awesome.
Current things on the go:
http://chatlogs.planetrdf.com/swig/2007-01-08.html#T19-44-00
- SWIG chat on some followups to Whits inspired topics
DanC found Whits whilst trying to find out how far I'd got on Notation3 grammar work, or GRDDL and EARL. Things we ended up discussing include pattern mining, huffmanising URIs and organisational strategies surrounding that, OLPC, some timblian typos and honours, the tab-completion thing in zsh, the evils of open and save, and doing simple operations such as uploading screenshots to the web.
On the latter, Kragen Sitaker and I were discussing how U+0283 LATIN SMALL LETTER ESH should display, and tried to upload some screenshots to show one another how it looked to us by default. It took Kragen 135 seconds and me 94, using fairly similar methods:
http://swhack.com/logs/2006-12-27#T20-23-50
- The Great Screenshotting Race
I introduced DanC to http://imageshack.us/ which is something I used in my 94 seconds, and I found that an old URI that deltab pasted on Swhack back on 2004-05-22 is still working:
http://img14.imageshack.us/img14/6002/postal2.png
- USPS: Those Are My Shoes
So that gives me hopes over the persistence, though I have been thinking about trying to slurp back all of the things that I uploaded and mirroring them on my own site just in case.
Also, due to publically discussing Whits, Kevin Reid noted that I should have told readers of my other weblogs, miscoranda and so on, that Whits exists now and is where I do much of my writing. The only excuse that I have for not doing so is that I suck... well, in the case of miscoranda the problem is that I want to subsume the DOCUMENT_ROOT into inamidst.com's DOCUMENT_ROOT so that I get the inamidst value (my changeset system, the super-mirroring etc.) from it, and to make it easier to upload to, but I haven't got around to asking crschmidt about that yet.
So I was finding out about the sam editor, and the Wikipedia entry on it links to the following quick reference chart:
http://plan9.bell-labs.com/sources/contrib/steve/doc/sam-refcard.pdf
Only, it's a zero length file. So I went Googling for the filename to see if anywhere else had it, and there are only 36 results for it, all merely mentions of the name, not the file itself. The original file is still in the Google cache so it can't have gone long, but I was unable to find any current file corresponding to it on the web.
A search for [steve sam-refcard] was a little more productive as it led me to <http://9fans.net/archive/2003/09/127> which led me to the original homepage for the refcard, <http://homepage.ntlworld.com/savan/plan9/>. From there I was able to find a .tbl version, but no .pdf, of the refcard from the Web Archive:
http://web.archive.org/web/20031215163521/
→ homepage.ntlworld.com/savan/plan9/sam-refcard.tbl
Having then found this tip on using groff:
http://www.macosxhints.com/article.php?story=20051020122911920
- Use groff to easily create PDF pages
It was then a simple matter of this command line to recreate the PDF:
$ tbl sam-refcard.tbl | groff | pstopdf -i -o sam-refcard.pdf
And I've posted both of the involved files here:
Victory!
The main function of Whits is to remind me what I was working on, and what level of progress I made. Its secondary function is to let others know what things I'm working on. It doesn't have a tertiary function, and I'm not even sure what the rank adjective is for four: quaternary, perhaps?
Javier complained that Whits is getting "too chatty", which is a fair complaint except that it doesn't obviate fulfillment of the functions listed above; so by mentioning his complaint I've turned this into a chatty whit just to thwart him. I guess that's the tertiary function of Whits... thwarting (a lovely word, derived from the Proto-Germanic "thwerkhaz").
"You are devolving into tech-talk-I-had-a-sandwich-blogging."
- Javier Candeira
So I think I just came up with the best method ever for transferring a small amount of text from one computer to another. In this case, I had been searching for an old album review from 1999, and when I found it on my hard drive I was eventually able to find it in the Web Archive. I wanted to transfer the URI for that to my other box... But how to do it?
The problem with the first two is that they require logging into accounts, and opening a client. The problem with the third is that it requires me creating a pastebin.
Then I remembered that I have an IRC bot constantly running which monitors referers to my website. Now the problem is that it doesn't monitor all hits, just those that have come from external sources. So how could I get the small bit of text into a referer?
The answer: Google for [$text OR inamidst], then hit the top link.
Unorthodox, but it worked!
http://inamidst.com/proj/hangul/bookmarklet
- Hangul Annotation Bookmarklet
This takes a Hangul character in a page, and decomposes it into its constituent sounds or Jamo, all using Javascript. It might be useful for Hangul learners, and maybe as a unicode curiosity. Enjoy!
A while ago I complained about not being able to export my Google Search History, but it turns out that it's already possible, though not very well publicised. This article has the nitty gritty:
http://persistent.info/archives/2006/02/24/search-history-rss
- Google Search History as RSS
Well, it has the nitty, but not the gritty. The article shows you how to get an RSS feed of particular searches over your search history, so the first thing that I had to work out was how to get it to export the whole history: leaving out the ?q= query component.
Then I went through starting from ?start=0 and ?num=2000 and incrementing to ?start=2000, 4000, 6000, et seq. Eventually I'd exported about 40,000 RSS items, which are a mix of search queries (things you type into the search engine), and search results (links you click on in the results).
The next step was to try to parse these into a non-RSS-2.0, indeed non-XML, format. I opted for "$date $type $link? $title" where $type is either Q for search queries or R for search results, and $link only appears if the type is Q. But the RSS was malformed... Expat wouldn't parse it because of character encoding errors.
To find out what the character encoding errors were, I tried using pcregrep but for some reason it didn't support -o even though the man page said it did; and grep -P doesn't work on my system either, so I ended up writing a quick and dirty Python script. All of the hi-byte sequences (\x80-\xff) were one byte long, so it seems a good bet that the feeds were actually cp1252 encoded. I went through them all with nano, correcting the XML declarations; I couldn't use emacs because nxml-mode wouldn't let me save with encoding="cp1252".
So then I was able to write a parser. Here it is:
http://inamidst.com/stuff/2007/gooshparse.py
- Google Search History RSS Parser
I converted each RSS file into my plain text format, then ran it through a sort -u (uniqifying) filter. I've made about 42,000 queries and link clicks overall, and about 23,000 queries. The most frequent search that I make is "encnorm", which is for my encoding normaliser tool; I eventually added a link to it on my start page.
Interestingly, the feed exporter doesn't adhere to ?num=2000 all that well, exporting between about 1900 and 2100 each time, except on the last feed which only had about 1700 items in it. I may have missed a few due to incrementing the same amount each time rather than the amount that I actually got.
So that's the gritty. But it means that rather than just the HTML interface, or even the 13MB RSS export, I've got a nice compact 3MB file of my searches. Per nitty et gritty ad nifty!
Today, I got to speak to my favourite cartoonist of all time. He wrote a strip that, even though it ran in a national newspaper for years, didn't get nearly as much recognition as it deserved, and I wondered what happened to it and what was going on. For my meagre question I got an amazing, forty minute involved discussion about the past and future of the strip, interspersed with chatting on the history of cartooning in general, theories on humour, and all sorts of stuff. It was really kind of him to give me so much of his time, and it's because he's clearly very passionate about his work as well as being considerate to his fans.
I read this strip as I was growing up, so it was really fun hearing some of the details on where the characters came from and so on, even though I haven't read it for years since it hasn't been around. Hopefully that will change in the near future, because he's planning on going online.
The strip was called Bogart, and the author is Peter Plant. It's almost not worth mentioning the details at the moment, because on the web it's like he doesn't even exist. There's practically no information on him or his strips, which is an absolute travesty. I think I'll probably set up a kind of quick information page about it before he puts his own materials online, because really the more people that know about all this, the better.
So anyway, the dude is seriously funny, and I hope to be able to report more details on the saga in the not too distant future! Watch this space, &c.
Things I've been doing or thinking about:
http://www.compulink.co.uk/~shutters/reviews/98050.htm
- Ken Campbell's Pidgin Macbeth: A Review
In the language itself, "Makbed blong Willum Sekspia". Bagarap is an awesome word because, if what I remember from reading Wikipedia ages ago on pidgin langauges is correct, its etymology is "bugger up".
Quicksilver, the application launcher for OS X, just changed its menu bar icon from the natty Mercury symbol to a big "Q". That's kinda a shame!
Yesterday morning I played about with Inkscape, learning its features. It's almost the object oriented equivalent of The Gimp's assembly. I was inspired in part by some corporate logo parodies, via Cody:
http://flickr.com/photo_zoom.gne?id=198311028&size=o
- Corporate World Meet Web 2.0
And partly via the mad Inkscapin' skills of Pat Hall, who showed me a few of the techniques, such as making a deck of cards via mass duplication and then even distribution.
One of the more annoying things is the gradient interface. I'm starting to get used to it, but it's still one of those kind of "just play with the sliders until it looks less crap than at the moment" things. It's a bit counter intuitive: for example, why do you have to add a stop to be able to edit its balance? Why can't you add two and edit their balances?
Also I ran into a slight problem for a while trying to paste in a Unicode character. I was using Command+V, but since it's an X11 application it requires me to use Ctrl+V. It also doesn't support Unicode input, probably because Alt is mapped to other things in X11.
Other than that, it's great. I really like the logic operations on shapes, and I've actually got the hang of gradients enough to be able to get reflections working quite well. Here's an example of what I produced:
http://inamidst.com/trove/glossy
- Glossy inamidst logo
I tried it with a slanted reflection first of all, but that screwed the perspective up no end: I think I would've needed to slant the logo itself too for that to look right. It's actually somewhere between glossy and hazy... I tried doing the "cut haze" trick that most people seem to use to get a gloss on an object, but for some reason I couldn't make it look right. All the same, I rather like the glosshaze.
Sheep really do look quite small from that altitude. It is comical.
As I noted over a week ago, I've been working on Shakespearean things. In fact, I've been so happily engaged in Shakespearean research and stuff that I've not been posting here as a result. I'm not sure whether anything that I've been doing will escape to the web, but basically I've been reading about dozens of separate subjects and taking notes. I've thought of several things that I might do on the web as a result of it, such as SHAKSPER posts that I'm noodling on, and I'll report stuff here as it comes up.
Apropos of a discussion on IRC, I present a new topic:
http://inamidst.com/topic/written
- The Written Word
Just 600 words of babble at the moment, but I rather like the concept of a Pathetic Renaissance. I should probably put it in the design patterns directory I've been trying to maintain. That's a thankless task... coining a pattern is a really easy thing to do, but then if you have to give 300 words of documentation to 200 patterns, that's going to take a while.
http://www.uvm.edu/histpres/HPJ/streetlights/index.html
- Shedding Light on Alienation in the Public Realm
A wonderful little talk by Alfred Holden on why using incandescent, full spectrum, lighting in public areas is beneficial for many reasons. Brought about by me researching city design patterns... Milton Keynes is 40.
Linked to via another interesting site:
http://www.carfree.com/library.html
- Carfree Cities: City Design Library
Its pieces on boulevards and parks and so on are pretty good: basically series of example photos with reviews and critiques on what elements of them are good and bad.
So I just bought a fragment of Dho 908, a lunar meteorite. I worked out that it's valued at the equivalent of $2591.60/g, which is actually fairly middling for this class of meteorite. Martian meteorites reach the craziest prices, especially those with wacky minerals in them. At any rate, Dho 908 is a bit of anorthositic impact-melt breccia, or as one paper by R. L. Korotev says, "normatively, the meteorite is a magnesian troctolitic anorthosite."
It's so interesting that I might try to cobble a page together about it when it arrives. Apparently, the lunar meteorites are actually of scientific value, so small is the amount of material that we have from the moon from the manned and unmanned missions. Dhofar 908, for example, "gives credence to the hypothesis that that [sic] surface material of the FHT is more mafic than deeper material", the FHT being the Feldspathic Highlands Terrane, or "the lighter bits" as they're more commonly known.
Street of the day:
Whip-Ma-Whop-Ma-Gate is one of the smallest streets in York, if not the smallest. It is between Colliergate and Fossgate and intersects The Pavement and The Stonebow in the York city centre. It is 35 metres long and was once the location of the city's pillory, hence the name. On older maps, the street was called "Whitnourwhatnourgate", "Whit nour what nour" meaning "neither one thing, nor the other" in Anglo-Saxon.
There were a lot of bickering birds this morning, mainly crows and starlings infighting. It was almost an aviamachia.
The Surgeons and the Barbers:
"The origins of the College lie in the union in 1540 by Henry VIII of England
of the Worshipful Company of Barbers (incorporated 1462) and the Guild of
Surgeons to form the Company of Barber-Surgeons. In 1745 the surgeons broke
away from the barbers to form the Company of Surgeons."
- http://en.wikipedia.org/wiki/Company_of_Barber-Surgeons
Whoo, my Dho 908 fragment arrived!
"The Afrikaans (and sometimes English) word for gnu is wildebeest, which is Dutch, but the Dutch word for wildebeest is gnoe (pronounced [xnu])." - John Cowan, on #d8uv.org
A Google Translation, French -> English, on my name gives "Sean B. Micrometer caliper". I was telling Javier this when my "f" key got all sticky, probably from a crumblet getting stuck underneath it, and so that of course ended up with him asking me to spell "forfiffimo" with long-s instead of f, and...
Long story short, I hate the hooks on the back of long-s that some fonts have, including Lucida Grande, my IRC font. I'd really like to be able to create a custom font, taking glyphs from lots of different fonts and having them scale appropriately using some fuzzy scaling algorithm. Getting the consistency right would be difficult, but I'd still rather have a slightly inconsistent long-s sans hook than a consistent one avec.
Here's an example of what I mean by a long-s with hook:
http://inamidst.com/stuff/images/esh
- Esh/Long-S IRC Screenshot
It's also hosted at http://img124.imageshack.us/img124/3078/eshvu7.png but for some reason Imageshack hasn't been working for me for a few days now.
The previous post had "beenn" as a typo for "been", now corrected, which reminded me that yesterday I came up with a possible name for a pub should I ever become a landlord: Pubb! It's so totally modern and trendy, eschewing the old "Dog and Biscuit" tradition for making memorable signs and whatnot; but I think the short and sweet name thing is a good new tradition to nurture. "Pubb!" is somewhat based on the Seatbelts' song "Tank!".
(I should note that it's not very likely I'll become a pub landlord, so you probably won't be able to go to Pubb! soon, unless someone else takes up the name. Please feel free to use it, and let me know if you do so...)
Today I got some coins from The Shire Post: http://www.shirepost.com/ They sell "fantasy coinage", i.e. hand pressed coins designed with things in mind such as J.R.R. Tolkien's legendarium. Now, I don't usually go in for this kind of thing. I like coins, but I'm not really into medallions or tokens or anything of that sort, even when they deal with real world concerns. With fantasy stuff, it's even worse.
But these coins are totally superb. For $10, I got these five:
They come in excellent little plastic presentation pockets, with cards explaining about the coins. I'd actually ordered the 1405 Hay-Penny, not the Penny (which is more expensive), but I guess either they didn't have it or they made a mistake since they gave me a Penny instead; and I also didn't order the Mathom but they threw that in for free and sent a very polite handwritten letter explaining that.
There's a little more about the coinage here:
http://www.shirepost.com/ShireCoinage.html
- Tolkien-inspired Coinage Page
I'd originally thought about getting some way back on 2006-05-08, when Mark Shoulsdon showed me the page:
http://swhack.com/logs/2006-05-08#T17-02-26
The thing about tokens and medallions and fantasy coinage is that usually the aesthetic quality of the pieces are extremely low. These are starkly different, with really beautiful and intricate designs, which is why I was attracted to them: it really brought to mind the kind of origins of coinage, stamping bits of metal to guarantee their weights. These coins are somehow primitive (because they hark to mythological worlds and are hand pressed), and yet very modern (due to their exquisite designs and quality presentation), so it's the best of both worlds to some extent.
So yeah, if anyone's looking at the website and thinking "those are beautiful, but I'm not sure about this", I totally recommend going ahead, at least to get a few of these cheap coins--what's $10, after all? If you don't like them, then surely you're in some manifest danger not to understand them, but if you do then perhaps, like me, you'll be considering buying some more; especially of the silver pennies, the dozen of which Sam et al. used to buy Bill the Pony.
Sean B. Palmer, inamidst.com
