Being for the Month of 2007-03
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.
Once, I was staying somewhere that was prone to fog, and looked out of the window one morning to see the sea fog had rolled in up the valley but... it was only about five feet deep at most. You could see the trees sticking out of it. It was surreal because you couldn't tell where the sea ended and the land began. Sadly I didn't get a chance to go out walking in it, but I wonder what that would've been like.
The beginning of the Whits month seems to be font time. I've been looking at using some nicer fonts in things that I do that require typesetting, and I found an awesome collection of top five font lists, transcribed them all, and found the most popular. Here are those that got mentioned three times or more:
Whilst I was checking out some of the fonts I hadn't previously heard of, I found out that I'm not very interested in sans-serifs anymore. Some of the new semi-serifs are kinda interesting, but serifs are where it's at. So I derived a list of interesting serifs that I wanted to check out, and here are the notes that I jotted down about them:
As you might be able to tell from my sly little mention of it, I'm still really into ITC Galliard for printed body text, but perhaps one of these will entice me to diversify a little.
(And as you might be able to tell from the linked text, I'm using Avocet now to write these notes. It's going well so far, and I've even made a little test suite to make sure that it's fairly robust, but I'm still trialing it.)
Some guys gave a talk called Web Typography Sucks at SXSW 2007, and I was leafing through the slides the other day. They describe this technique called "vertical rhythm" which is absolutely awesome. Basically you overlay a lot of evenly spaced horizontal lines going down a webpage, corresponding with the line-height, and if the text consistently matches the lines then it has vertical rhythm. If not, then it doesn't. You wouldn't think that without the lines it'd be noticable, but when you see the pages sans lines, it's still obvious because the page with rhythm looks clean, balanced, and professional, and the one without it looks cramped, jerky, and awful.
So I've been tinkering with their technique, and I think I've got it sussed. The principle is not particularly difficult, but of course the details are another matter. I've implemented it on Whits, and I think it looks much better than the previous design. I'd tried switching to sans-serif font to get it looking sleek, but all along I could've achieved a similar effect just by getting the layout rhythmical. What a great technique.
I was adding a few more content-blacklist items to my feedback script, which I \xHH encode using Python so that the content doesn't get into the web, and I accidentally typed "exit" rather than "exit()" to quit, as I often do. That got me wondering about the nature of the exit object, because it does this when you screw up:
>>> exit Use exit() or Ctrl-D (i.e. EOF) to exit
So I went investigating...
>>> repr(exit)
'Use exit() or Ctrl-D (i.e. EOF) to exit'
>>> vars(exit)
{'name': 'exit'}
>>> type(exit)
<class 'site.Quitter'>
>>> vars(site)
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
NameError: name 'site' is not defined
>>> import site
>>> site.__file__
'/usr/local/lib/python2.5/site.pyc'
>>> myexit = site.Quitter()
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
AttributeError: 'module' object has no attribute 'Quitter'
>>> exit.__call__()
This is a reduced version of what I found in site.py:
class Quitter(object):
def __init__(self, name):
self.name = name
def __repr__(self):
return 'Use %s() or %s to exit' % (self.name, eof)
def __call__(self, code=None):
try: sys.stdin.close()
except: pass
raise SystemExit(code)
__builtin__.quit = Quitter('quit')
__builtin__.exit = Quitter('exit')
Quite cool. For some reason, I was expecting it to be an object derived from str only with an added __call__ method...
>>> class Quitter(str):
... def __call__(self):
... print 'You would have quit now if this were loaded'
...
>>> q = Quitter('Try calling this object instead')
>>> q
'Try calling this object instead'
>>> q()
You would have quit now if this were loaded
>>>
Yeah, I think that's cooler.
I went to the Post Office to exchange a ten pound note for ten pound coins, and the assistant said "first class or second?". For a while I was wondering just what first and second class pound coins would let you do... perhaps you can only buy first class seats on the train with first class pound coins? Usually I get stamps there, so you'd think that would explain the mistake, but I never ever get second class ones so the mystery is as yet unexplained.
Another bug ironed out in Avocet: it wouldn't let you put ** around links for strong emphasis (regular emphasis didn't work either). Doing the actually- use-the-dang-thing trial seems to be rather a good idea. I'm also thinking about having an -n argument for the note script, which would let you edit the current entry sheet without making a new entry, because I often go back to edit old stuff and just type "note" and delete the heading currently to do it. It's a lot easier to type -n than to delete a heading.
I'm thinking of switching from Firefox to Safari. Firefox is slow and takes up an extraordinary amount of memory, something that's becoming more annoying over time as processor speeds and memory lag behind software bloat. Previously I hadn't considered the switch because I thought that Safari was lacking features that I rely on in Firefox, but I've been reassessing that and am pleasantly surprised.
The main problem was the lack of Undo Close Tab. I use this dozens of times a day, pretty much as a feature rather than an "oops" thing. When I went looking for something that adds this to Safari some weeks ago, I didn't find anything; but this time I found ForgetMeNot. The comments on the page are quite new, so I guess that the feature is new, which is useful because it really is a core essential for me.
Of course, for Undo Close Tab to matter you have to have tabbed browsing to start with. This is turned of by default in Safari, for some unknown and presumably bizarre reason, but is easily turned on and works almost the same as in Firefox... except that favicons don't display on the tabs. This is something that I rely on again since I often have literally over a hundred tabs open at any one time, and finding the tab I need when there's no title visible comes down largely to identifying the favicons.
I've only just started looking around for a solution, but it looks as though SafariStand might be just what I'm looking for and more. It gives you a kind of tabs thumbnail sidebar, which is even better than the favicon orientation really. I haven't tried this out yet though, and I'm also wondering about what happens when you have so many tabs that the tab bar overflows... Firefox handles it quite nicely now, after some initial teething problems.
As for other extensions, I don't have many installed in Firefox, surprisingly. I used to, but I really only rely on a few. Here's what I've got:
Of course, there may be things in Safari that aren't in Firefox that I come to rely on too. Stop-or-Reload points to some decent innovation therein, but given the popularity of Firefox, it's likely that anything good in Safari will quickly be cloned.
Spring, the sweet spring!
Cuckoo, jug-jug, pu-wee, tu-wit ta-woo!
Lots of stuff going on today. The first thing is that, of course, it's spring—Germinal in the French Revolutionary Calendary—and the equinox, which I just learned essentially means "equal night" even though practically speaking, the sun's diameter means that night and day aren't equal on the equinox.
Of course, I take spring as starting from the 1st March anyway. It makes sense that the spring equinox should be the middle of spring, since it's exactly half way between the solstices. That'd make the beginning of spring to be 365/8 days (half a season) before the spring equinox, which seems to usually fall on 20th March though this year it was seven minutes after midnight on the 21st. So that's about 46 days, which puts it at 2nd February.
But the other thing is that the earth lags in being warmed by the sun, of course—the oceans lag by a couple of seasons in their warmth so I've heard. In any case, the 1st March is about 19 days before the equinox, which is just under a quarter of a season... I wonder how well that accords with the lag? It'd be nice if it was roughly correct.
Stuff in the news today that's interested me includes a big theft of Fabergé eggs in Wales, though not of course of the originals (eight of which are anyway missing); an awesome Map of Science (cf. a, b); and a new dinosaur called Oryctodromeus cubicularis. My immediate thought on the latter was to check Wikipedia but there's no entry for it yet. I'm going to wait to see how long it takes rather than adding it myself (I realise that this strategy does not scale).
I also like Shorpy, a blog of 100 year old images (though it'd be much better if it had UK images), and the New York Times's article on giant snowflakes. Snowflakes are pretty. I found a page on snowflake photography which notes that it's a pretty expensive craft, which is a shame. But then it's no different to, say, astronomy—you get what you pay for.
I'm having trouble getting used to the Avocet idiom of putting the URI last in a link... it used to be the other way round in Pwyky. The Avocet order is better though.
Several days ago, I made a Unicode Tool to look up codepoint names and characters. For example, you can search for interro or ‽. (Also, I'm quite impressed that Avocet manages to grok the {U+203D} unicode syntax in link text without me even having a test for it. It does the right thing! I'll make a test for it.)
So I signed up to twitter as "sbp", and the funniest thing so far is that when I entered my location as the United Kingdom of Great Britain and Northern Ireland, it cut it off at "United Kingdom of Great Britai". How dare it truncate the name of my country! I can't help it if I live somewhere that has an extraordinarily long name... it's not like I'm the only one, either. Wikipedia says that there were 60,609,152 others just last year.
Apart from that, and the fact that I used the weird voynich gherkin as my signup image, it's unimpressive. As usual, Aaron Swartz got there first and did it better anyway.
I've moved my dock to the left and made it a bit bigger. Having a widescreen, I decided it made sense to use the extra width because most application such as browsers don't need extra horizontal space but do need extra vertical space. It makes it a bit hard for a while because having the bar and dock at the bottom is something I've been used to for many years now, but since it's also let me make the dock bigger giving me bigger application targets (a common UI rule: make frequently accessed controls larger and thus easier to hit), the cumulative benefit will hopefully be quite large.
It would be nice if there were a quantitative way to adjust the size of the dock, though, rather than just having a slider.
Transcribing Will-o'-the-Wisp articles is fun. I did a few more today, and announced them to my mysterylights group, which as Morbus observed is so old-school that the signup email addresses were still saying "egroups" until recently. The articles transcribed were by Henry Duncan, James Motley, and the awesome Thomas Lamb Phipson. I really like the whole idea of commenting on ancient writers as if they were just writing yesterday, so I kinda did a bit of that too.
As well as that, I restored uwimp.com for William, since for some reason it went AWOL and all the files got lost. I used a thing called Warrick, a perl tool written by Frank McCown, to restore it: it looks through Google, MSN, Yahoo!, and the Web Archive to get your files back. I thought I had a local archive of uwimp.com lying about, but if so I haven't found it yet; there may yet be more files lying around.
The creator of Warrick asks to send him an email letting him know if you've recovered a site using the tool, so I'll do that now.
Somebody emailed me today offering to translate bits of the site into Polish, which is great. At first I was worried it was just a kind of "I'd like to have content on my site to plaster ads all over", but the offer actually seems genuine, and I'm going to be hosting the resulting translations on inamidst; he's even offered to buy Polish domain names to redirect. Wonderful!
In other news, the moving of my dock from the bottom to the left of my screen is actually going quite well. It's not so hard to adjust to it as I thought... a lot of application switching is performed simply by clicking on the other window if it's visible; you only need to use the dock for hidden applications.
I've been getting some weird spam in my feedback form recently where people just paste in stock phrases like "You have made nice web!!" and use fake email addresses. They're clearly spam because they're repeated extremely frequently from a variety of IP addresses, and use those fake return addresses.
So I've been adding the IP addresses to a blacklist, but I haven't tested whether it works or not, and as I was outside today I was thinking that I could simply check the list for duplicates as one test that it's working.
That got me to thinking about a simple duplicate testing algorithm in Python. I was going for clarity and simplicity over performance, and I came up with this:
>>> items = ['p', 'q', 'p', 'r', 's', 'r', 't', 'r'] >>> iset = frozenset(items) >>> if len(items) != len(iset): ... for item in iset: ... items.remove(item) ... print frozenset(items) ... frozenset(['p', 'r']) >>>
Obviously one limitation is that it doesn't show the frequency of the duplicates, but that wouldn't be too difficult to add, and isn't required for my use case anyway. (And it turns out that there are no duplicates in my blacklist.)
Phew, I've done an amazing amount of writing today. Also a smidgenous bit of coding, including fixing my scrape.py Yahoo! Groups Archives mirroring script to use mechanize instead of the now outdated ClientCookie. Why they had to change the thing I dunno, but thankfully the only change that had to be made was "import ClientCookie" to "import mechanize as ClientCookie". People should really think about their interfaces more before releasing them—and after for that matter! The rdflib instability is another fairly recent example of the same thing. Still, I'm sure that I've done the same many times... perhaps there's something that can be changed in the nature of the programming language or environments to help instead.
Applying the vertical rhythm technique to miscoranda seems like it might be a good idea. It looks quite dapper kitted out in its extra whitespacing, placed on a field of grey with a hint of green. When the links mirror (though darker) the colour of the fleur-de-lis too, the whole effect is quite becalming and professional looking.
As to the content of the site, I'm not sure what to do with it really. I have wondered about changing the \d+ scheme to names as in What Planet is This?, but whether that'd make me think about the site any differently is hard to say. The tradition of it is technology with a few other things thrown in. I had a little spurt at the end of 2006 of being interested in technology, and was thinking about setting up yet another technology blog (with quite a good name idea), but my interest in technology again faded, and it's still languishing. Yesterday, after I wrote "I've done an amazing amount of writing today", I wrote an extra 4500 words on the state of the Semantic Web, its problems and possible improvements. But just criticising something isn't all that fun, and I've spent a lot of effort trying to fix it already; it's really quite hosed.
All the same, I think that the Semantic Web in some form will emerge, probably through completely different sources. It's a simple concept, and just takes a good initial direction and then a lot of grassroots effort and a few big companies on board.
I've been listening to a lot of Rock and Roll from the '50s, and inevitably I got drawn into wondering what the first Rock and Roll song was. Wikipedia says Rocket 88, but going back and listening to an extraordinary amount of stuff from about 1938 up to 1955 it's clear that the picture is a lot more mixed and nuanced than it first appears.
For a start, Rocket 88 (1951) is based very closely on Cadillac Boogie (1947), with the fuzzy guitar being the main change in instrumentation bringing it closer to Rock and Roll. I think that the non-honking sax in Rocket 88 actually drags it back away from Roll and Roll, though.
Later in 1951, Bill Haley and the Saddlemen (as they were then called) got turned onto Rock and Roll by the Treniers, and chose Rocket 88 as their first new recording. Their version is sometimes called the first Rock and Roll song, but it's also derivative of that earlier Cadillac Boogie in its structure, so you have to look at the instrumentation and performance. It's in a way more rocky than Ike Turner and his band's version, but in a way less too; last night I came up with a points scheme to rank how many Rock and Roll elements are in a song, and both versions of Rocket 88 got 24/30.
Earlier songs which seem very Rock and Roll to me include Rock Awhile (1949) by Goree Carter, and Rock the Joint (1949) by Jimmy Preston. The former is a kind of Texas style boogie, and the guitar is very much in the tradition that Chuck Berry capitalised on. The latter is hoppin', and it mentions the Hucklebuck. The so called first Rock and Roll concert, the Moondog Coronation Ball on 21st March 1592 set up by Alan Freed only had one person playing before it got shut down, and that was Paul "Hucklebuck" Williams. Perhaps it was The Hucklebuck that he played, though I haven't been able to confirm that yet.
Anyway, if Freed was able to make a Rock and Roll concert in 1952, Rock and Roll must have been around by then. But it's a spectrum rather than a big leap, and you can't really draw the line, but you can say "what was the biggest jump along the spectrum?" I'd say that Rock Awhile is as good a contender as Rocket 88, and Rock the Joint is up there too. Another interesting one that deserves a mention is Saturday Night Fish Fry (1949) by Louis Jordan. The refrain for it goes "it was a rocking" and then has an excellent guitar break, all of which is repeated a lot. If you except just those bits, which make up like 30% of the song, and played it to anyone who was reasonably well versed in '50s Rock and Roll, I'm sure they'd say it was definitely Rock and Roll. But it's only a part of the song; the rest is jump blues.
One funny thing is that between the big inception of Rock and Roll in 1949 or 1951, and the first Rock Wave in 1954-5, there's usually only one person that people mention: Bill Haley. I'm sure he wasn't the only person doing Rock and Roll records in that time, but he seems to have the monopoly on people's minds, and it's actually not hard to see why... if Rock and Roll really requires the melding of boogie, blues, and country and western, then Bill Haley was the first person to give it the final mix, even if it's Chuck Berry who would perfect it in a slightly different strain (the Texas double pentatonic stops being pretty much missing from Haley).
The biggest thing that impresses me with Bill Haley with Haley's Comets, as they were called on this record, is Crazy Man, Crazy (1953). It was written by Bill Haley and Marshall Lytle, so it's one of the first (i.e. not just a buzzed up copy of an earlier song) Rock and Roll songs to be written. Not that it didn't have its derivative elements: Wikipedia says that the "go, go, go!" part is reminiscent of "Go, Go, Go!" by the Treniers (who introduced Haley to Rock and Roll) "which music historian Jim Dawson suggests may have been an influence". Well, I listened to both of those songs before I read Wikipedia and already put two and two together myself; it was extremely obvious. Another influence that Wikipedia doesn't yet mention is Nervous, Man, Nervous (1953) by Big Jay McNeely. It sounds very reminiscent, but I'm not sure which of the songs was recorded first. Crazy Man, Crazy was early in the year though, so it might be that McNeely derived his.
If so, that just underlines the absolute weirdness of Bill Haley and the Comets. They were just so far ahead of their time that it's more than uncanny. They didn't just take these amazing proto-rock songs and do dilapidated safe white versions of the songs, they actually injected more pace into them. And yet Bill Haley himself was a yodeling middle-aged ex-country and western guy with a family and a curl. He speaks so clearly on the records... and then you have all this backbeat and guitar work and sax and stuff in there, and you really wonder what on earth's going on.
Well, when Blackboard Jungle popularised Rock Around the Clock in March 1955, that's probably what got the new wave going. Elvis Presley, Chuck Berry, Buddy Holly, Little Richard, and Bo Diddley were all getting going at about that time; I think that Elvis was the first to take it up, about a year before Blackboard Jungle even. The Sun Sessions of Elvis are just astounding... Scotty Moore and Bill Black with Elvis put together a stunning sound.
So I'm not confidently sure what the first Rock and Roll song is, but if you can call some of the proto-rock songs like Saturday Night Fish Fry Rock and Roll, then fair enough; Chuck Berry says that the first person he heard playing Rock and Roll was Louis Jordan. And Rocket 88 was certainly a leap, though it might not have been the biggest. Due to that fuzzy guitar, perhaps you can call it the first Rock and Roll recording, but even then you can't ignore Rock Awhile. As for the first song, Crazy Man, Crazy seems to have a really strong claim to me. Or if you want to go for when Rock and Roll was perfected, the first 30/30 song, perhaps you can go for Maybellene, which actually had Willie Dixon thumping away on the bass.
Whatever we make of the music and its progression, the music itself is cut and it's not going to change, and the purpose of Rock and Roll is to get on down and freaking well enjoy it. And get on down and freaking well enjoy it we can still do, even with the early proto-rock stuff: one of the most pleasant surprises of getting into all this has been finding that it's really good!
The blackbird has a wonderful song. I'd forgotten how much I missed it, given that it's been so long since one's been around. According to BBC News, the number of blackbirds has fallen by 25% in a year, whereas the number of song thrushes has fallen by 65%. There's a garden near here that has a load of trees and bird feeders and all kinds of bird attractive things, and that's where I saw the blackbird this morning, sitting on top of the porch at the end of the garden, singing boldly away. In fact I can hear one out the back of here now from time to time; perhaps it's the same one.
According to the same BBC article, the three most common birds nowadays are the house sparrow, the starling, and the blue tit. I always considered the blue tit a rare treat, so that's surprising; but even more surprising is the ascent of the house sparrow, which seems startlingly rare here. There actually seem to be more magpies than sparrows. The most common bird in Scotland's gardens, on the other hand, is the chaffinch, which I'm not sure exists around here. There were a few interesting thrushes late in the previous spring. Hedge accentors are awesome, but I'm not sure that there are many of those around, either. Wikipedia says that their "song is thin and tinkling".
The way that Tolkien felt about "cellar doors" I feel about "hedge accentor". That name for the dunnock is actually quite recent though, dating back to 1825, and apparently being an East Anglian term. It comes from the Latin for "one who sings with another". The OED says that it's first mentioned in The Vocabulary of East Anglia by Robert Forby, the only edition of which I can find was published in 1830 rather than 1825, but is ascribed to "the late Rev. Robert Forby", hence the OED's "a1825". The only bit that the OED quotes is "Hedge-accentor, the hedge-sparrow.", which is a shame because the full quote is wonderful:
“HEDGE-ACCENTOR, s. the hedge-sparrow. This quaint and seemingly affected term of East-Anglian ornithology, has not the air of having been invented by the vulgar, or of being likely to be used by them. Many naturalists have thought it proper that this bird should be separated from its nominal connection with the common house-sparrow; and somebody has done so, by this fanciful and far-fetched name. If it can be said to have any meaning at all, can it be that, as in every choir there is a precentor to lead, and in some a succentor to follow the chaunt, so this feeble warbler is an accentor, by adding its simple twittering notes to the great chorus of Nature?” (pp.155–6)
After considering it for ages, I've actually got around to writing republical.py, a script that prints out the current day in the French Revolutionary Calendar. I've used the Romme method for calculating which years are leap years since it essentially matches the Gregorian Calendar, which is handy.
$ ./republical.py Bouleau, Germinal CCXV (Romme)
So today is the day of the Birch Tree, in the month of the Seed.
Whereas tag soup is what makes the web work; and whereas the better part of tag soup interpretation is liberal parsing; and whereas the better part of liberal parsing is the ignoring of errors to the greatest extent possible, it seems that error recovery should play a greater part in format specifications of the future. I wonder if there'll be a kind of paradigm shift at some point where the model for parsing and recovery takes a larger rôle than the actual definition of the structure of the format itself?
At the moment, tag soup is seen post facto, implemented in browsers as a necessity of economics (people won't use the browser if it doesn't parse all available web pages!), and hasn't even been formally captured except in yet more code: HTML Tidy is sometimes mentioned as a function, htmltidy(), to magic away the tag soup problem, and yet HTML Tidy is a complex system.
So perhaps the core of the problem is that error recovery on this kind of level is just too complex to be specified, and therefore isn't seen as a stringent system. That there are some guiding principles or rules in such systems is more than probable though, so the isolation and study of those is presumably what the TAG are interested in in their TagSoupIntegration-54 issue.
One of the main reasons for publishing stuff to the Gallimaufry of Whits is so that I and others can find the details again on search engines. But it's odd that for some reason it's getting a PR of 0, and has been for some time now. I suspect that one of two possibilities is at play here:
So to attempt to remedy this, I've taken out the old redirect for now. I haven't removed the links, because the links are useful. If this still doesn't work, I might move it to a different directory or something, but I don't really think that ought to be necessary. I've also made sure that my sitemap includes the pages, though I doubt that does much.
An interesting point to note is that even though the pages have PR of 0, they still come up quite high in searches, e.g. for "gallimaufry" or "whits". I guess that PR isn't the only factor influencing the rating of a site; it must in part be due to the performance of the site overall. Either that or the PR reported for this site is out of date or otherwise erroneous, a distinct possibility given the fact that I fixed the redirect problem quite recently.
When most of the United Kingdom of Great Britain and Northern Ireland, including my automatic clocks and computers, switched over to BST on Sunday, my faithful trusty wristwatch did not. In defiant civil protest of the stupidity of BST, therefore, I have decided to thus far not change my wristwatch to show anything other than the approximate Greenwich Mean Time (okay, UTC) that it does now. I wonder if I'll be able to keep this up until the 28th October? Converting between the dates is already proving to be a little bit of a nuisance, but it does make me feel rather more settled as to what time it is than most people who are in their period of annual thrownoffness by the time shift.
Note also that entries to the Gallimaufry of Whits are still in UTC, though that's something which I've been doing pretty much universally for all of my weblogs anyway, so not a new innovation.
This morning I was wondering about some kind of What Planet derivative site focussing on smaller posts, just because it'd be fun. Even if I only did it for a certain period before abandoning it, that'd be okay; it's almost a workflow idiom to set up a "periodical" inspired site, work on it for a few "volumes", and then admire the results. On the other hand, there'd be no set term, so it'd be quite laid back.
There's no shortage of potential topics. As for the software, I think I'd combine Azimuth and the code that's running the Gallimaufry of Whits. I'd like titled and named posts as in Azimuth, but I'd like to use Avocet to write the material. As for styling, keeping it sans-serif and clean and beboppy would be great; though perhaps a semi-serif like Optima would be better. Of course, Optima only works on the Mac anyway.
I'm also wondering, as with What Planet, about binding it up as a printed matter periodical, and having ISBN'd published volumes of the materials after months or years if it spans such milestones. The main remit of the topics would be discoveries and histories, antiquarianism and the history of science, an inquiry into amateur activities and figures, middle of the road topics like Shakespeareana and post-Romano-British history, and modern topics such as astronomical and etymological gambols. Basically, the study of vectors over resultants.
Further to my report yesterday on such birds as the chaffinch, "which I'm not sure exists around here", today I saw two of them. One was in a tree, and the other was on the ground in front of a gorse bush. It's amazing what you can see if you just look out for it! Chaffinches are rather pretty. I also saw a songthrush, some skylarks, crows, magpies, sparrows, blackbirds, and most excitingly a stonechat. It was a european stonechat or Saxicola rubicola, and it was sitting in the top of a rather low bush as is quite characteristic for the species. It's black and white on top with a white collar, brown and dun coloured on the breast, and quite small.
Didn't see a hedge accentor, but apparently they tend to start appearing again at about this time of the year. Heard some coal tits, which have a very striking but repetetive tune, but didn't see them. They seem to manage to work themselves into hedges quite well.
It's a wonderful, real spring day today. Cloudy to the point of being slightly misty, and mild but yet not cold, such that you don't know whether it's going to get colder or warmer... perhaps it'll even rain? And then it does rain, but it only sprinkles a few small mild drops over the landscape before abating again. Good stuff.
The vertical rhythm technique is proving to be most useful once again. I've devised a style using it for the new periodical that I have planned, the name for which will almost certainly be Lo and Behold!
Yesterday I was quite busy with the designing and lots of other things, but I suppose there must be two kinds of busyness since sometimes that impels me to write up lots of things here, but yesterday I was just too involved in what I was doing. One thing I found was some relatively recent reports of the Will-o '-the-wisp, from early in the 20th century.
Another busy day, with the huge endeavour of writing the opening article for Lo and Behold! being mainly done now. I pretty much just need a title for it, and then it can be posted. Of course, I have to write the site software too... at the moment I'm using Avocet and SmartyPants to do all the formatting for me, which is working nicely.
Lo and Behold! is now much bepublishèd! I've managed to find a whole slew of antedatings for two awesome words, detailed in the first couple of articles. Since I'm currently wondering about publishing the whole thing in print quarterly, I might apply for an ISSN. The design is still something that I'm really delighted with, and a little surprised that I managed to make it that great, especially after the nice clean style of What Planet is This? The style for Lo and Behold! takes the best bebop aspects of its predecessor whilst tacking on some antiquarian feel, and then moudling the whole thing into something that looks a lot more professional. Designing on LCD helps.
The main task for today had been to write the software. Yesterday evening I settled on /lo/ as the pathname, over such contenders as /loandb/, /behold/, /origins/, /matters/, /sublunary/, /lab/, /lobe/, and /loandbehold/. I'd also already finished the bulk of the writing of the introductory article and some of the supporting material. The supporting material isn't entirely finished yet, but it's in a good enough state to move out of development anyway.
For the software, I modified Azimuth, so that it's a bit leaner and more suited to the job, uses perhaps more lightweight templating, ordered dispatch mappings, and all kinds of stuff. It's also a lot more simpler, with a lot of the optional features for comments and the suchlike taken out (rather, I started from scratch pasting elements of the old code in as they were needed). The post masters have .avo extensions since they're Avocet files, and they're built to .txt files which are really HTML fragments. Supporting material is also mastered in Avocet, but built to .html so that Azimuth can differentiate between the two.
The process of building a post is automated, but not quite complete yet. I want to just be able to use a "lo" script, but at the moment there's basically an edit and build process. Anyway, that's all good enough to get it running.
As for the direction of the whole thing itself, I hope that I manage to do some things that are on the same order of interesting as the What Planet is This? stuff; it's difficult because not only is it rooted in the mysteries of creativity, but also discovery, which is why a few bits of What Planet is This? were actually recycled—especially the bit on Hamlet and Jabberwocky from miscoranda, for example. Anyway, if the first couple of articles are anything to go by, it should be quite fun.
