The Occasional Masthead

If you’re reading this with Internet Explorer (and you’re on the front page), you’ll find a badge in the right hand column which leads you off to a page explaining briefly why you should consider using something other than Internet Explorer.

Tim Bray and Peter-Paul Koch talk about the problems better than I can right now, and I strongly recommend you read what they have to say.

This is not simple Microsoft bashing. All browsers are buggy, but some are more buggy than others. And buggy browsers are one reason why developing good quality web sites is expensive and frustrating. The problem is that Explorer has a massive market share, is more buggy than all of it’s serious competitors, and will no longer be fixed unless you pay for a new version of Windows.

None of this is in any way good for the consumer. So. Send a message. Use Safari if you’re running Mac OS X, or one of the Mozilla clones like Camino. If you come across a website that doesn’t “work” without Explorer, contact the author and tell them their site is broken.

Driving home from the QLHF meeting yesterday, before going over to Steven’s for dinner, I had a thought about how to keep my desired blog entry layout for well behaved browsers and something kind of OK for Internet Explorer.

If you peek in the source code for this page, you’ll notice that in the header there’s a snippet of JavaScript. It checks if the browser is Explorer, or not, and writes an appropriate bit of CSS for each case. This works because the structure of an entry is roughly:

    <div class = entry>
        <div class = header>
            Title
            <div class = timestamp >
                1:50 pm
            </div>
         </div>
         <div class = entrytext >
             vague rambling.
         </div>
    </div>

The “entry” <div> is set to position:relative. For everything other than Explorer, the “timestamp” <div> is then positioned absolutely by specifying its left and bottom, which aligns it relative to the “entry” <div>. For MSIE, the timestamp is just left where it is, with a bit of padding at the top for neatness.

So the external style sheet has styles for everything except the “timestamp” <div>, and the embedded JavaScript supplies that last little bit. I’m quite happy with this solution, as I feel it’s a reasonably neat hack.

There are still some things to do. Explorer 5.2 on the Macintosh still insists on leaving some space above the body text entry, blatantly disregarding the CSS 2 recommendation. I don’t think I’ll pursue this too much further, and instead just keep recommending people upgrade to browsers that work.

I might adapt that chunk of JavaScript to put a chunk at the bottom of the main blog page explaining why MSIE is not a good thing, or at least a link pointing to the relevant pages. Then I’ll have a cup of coffee and a bit of a nap.

Since I had a “spare” few hours, I’ve made some tweaks. The ugly archive page is slightly less ugly. The more-or-less random treatment of the meme badges on the main menu is slightly less random.

All that remains in this round of home improvements is to fix the horrible things that MSIE is doing to the title and date stamps over there on the left. Oh, and suggest to people using MSIE that they don’t.

Update, 6:37 pm I’ve done some hacking of both the CSS and the MT templates to improve matters. Well, it’s not much of an improvement, but MSIE is no longer breaking it as horribly.

If everything is rendering properly—as it does in Camino and Safari—then the title should be left of the body, lined up with the top line of the body text, and the time stamp down a little. In MSIE, it still doesn’t render correctly, as the body text has a space above it where the title used to be before I shifted it left.

The problem was that I wanted the time stamp to be aligned with the bottom of the body text. The timestamp is inside a <div> that is inside a <div> with position:relative. Which means that I can position the timestamp div absolutely, and tell it bottom:0. In Camino and Safari, the spec is followed, and the bottom of the timestamp <div> is aligned with the bottom of the container <div>. Not MSIE. No sirree. For no readily explained reason, the coders interpreted the spec as “align the bottom of the child with the top of the parent.” Despite the fact this is completely and horribly wrong with respect to the spec.

So, we have this—hopefully short term–design with the timestamp down a bit from the top. Thinking is, once again, in progress.

I thought, for a brief shining moment, that Eric Meyer had come up with a solution to my problem with the business card on my contact page.

Alas, the closest result still involved some non-semantic hacks and didn’t work if the text height exceeded the image height. Thinking is in progress. After I fix the ugly archive page. (Tomorrow. I promise.)

So, like, there’s this girl-geek, see. She’s 22, has a blog, reads The Onion, SlashDot and Dilbert.

She’s also an Iraqi, doing it in Iraq. Read it. Daily. Several Times.

I guess what I’m trying to say is that most people choose to ignore the little prefix ‘re’ in the words ‘rebuild’ and ‘reconstruct’. For your information, ‘re’ is of Latin origin and generally means ‘again’ or ‘anew’. In other words—there was something there in the first place.

Thanks to William Gibson for telling the world about this column.

Russ McGuire’s article is a bit extremely hysterical and reactionary, but worth peeking at:

This past week, everyday when I opened my Wall Street Journal, I was met with a full page ad from Microsoft. This ad was dominated by three simple words “Protect your PC.” This strikes me as something akin to the Saudi government running ads in the New York Times in mid-September of 2001 saying “Protect your Tall Buildings.” First, the message comes a little late. Second, we wouldn’t have to protect our PCs if Microsoft hadn’t provided criminals with all the elements they need to terrorize us Internet citizens.

I think McGuire needs some sort of award to recognize the number of Red! Highlight! Words! in his opening paragraph, although he would have a better chance at it if he implied that Bill Gates is a cannibal. However. Articles like this do suggest that Microsoft’s responsibility for bad software is starting to be recognised by the business community.

The next year or so in the on-going saga of The Beast From Seattle That Would Not Die could be interesting.

In an article cryptically titled “Dynomite!”, John Gruber expands on his previous article, particularly responding to counter-arguments about the complexity of computers compare to toilets.

Incompetent IT professionals are genuinely convinced that PCs are inherently so complicated that they cannot be expected to perform with high-reliability and low maintenance.

Gruber has expanded on a number of issues, particularly his previous statements regarding complexity, the dangers of mono-culture, and why Outlook is a Bad Thing.

I agree with his arguments that computer systems are not, inherently, complex. The technology, apart from getting faster and smaller, has not markedly changed for the last forty years. Certainly computer systems have lots and lots of components, with complicated inter-relationships. But the core ideas, processes and technologies are not hugely complicated. Most of the complexity comes from all sorts of baroque (or rather Rococco) frills bolted to the side. Besides, even if it is complicated, there’s no excuse for IT professionals shrugging their shoulders and saying: “It’s a computer. They break. Live with it.”

Complexity is not an excuse for low expectations. We’ve strapped men into giant rockets loaded with jet fuel, propelled them into space, and landed them on the moon. That was complicated. And our expectation was that we’d get them back. Why we don’t expect our email to work is beyond me.

My stream of thought has just been de-railed by an F111 whipping past overhead that was practicing for the dump-and-burn at RiverFire. The cat didn’t turn a whisker, although she raced out of the house this morning when somebody unloaded a tray of gravel several houses up the street.

I love John Gruber’s rants. Free-form abuse hurled at stupid technological decisions. This time, he’s aiming at Outlook and Express

Imagine if the plumbing in corporate America worked with the same degree of reliability as their computer infrastructure. This would mean that individual sinks, urinals, and toilets would go out of order on a regular basis. Water from drinking fountains would turn brown, but, hey, that’s just how it is. Every few weeks, teenage pranksters from Hong Kong would overflow every toilet in the building, knocking them out of commission.

Highy recommended.

I have, over recent days, been beating my brains with responses to Key Selection Criteria. Don’t ask. Suffice to say that my brains are now oozing out of my ears, a process accelerated by my crash-course in Objective-C and Cocoa. While Objective-C can best be described as an amazingly neat hack, and seriously cool, Cocoa is definitely an application of Clarke’s Third Law

Clarke’s law sprang to mind reading the an article by Tim Bray on perceived computer speeds:

Computers are amazingly, remarkably, unbelievably fast. When you’re executing compiled code that’s running around structures in memory (and they actually fit in memory), you can do stupendous quantities of computation with no perceptible delay.

This is a very important point, and one that most computer users and a good many technologists miss: current computers, even computers that are three to five years old, are fatuously overpowered for most activities. If the computer seems slow, there are probably three explanations, in descending order of likelyhood:

1. You are asking it to do something massive and complicated that you wouldn’t have attempted a year or two ago;
2. There is a disk or, more likely, a network bottleneck, often as a result of really bone-headed database or system design;
3. You have come up against some particularly bone-headed piece of programming (the first few releases of iCal, for example).

I will go out on a limb and state that most people use their computers as expensive type-writers that do really neat underlining, italics and two-page memos with little dancing bears in the corner. There are signs that Web browsers and instant messaging might be starting to broaden the scope of usage, but I still see an awful lot of people printing web pages to read them.

So, for free, a tip for all CEOs to improve their investments in IT: send all your staff on courses to learn how to cut-and-paste, and to not include one-page memos as attachemts to email.

I digress. As usual. I was talking about magic. A particularly good piece of applied magic, from an organisation which increasingly resembles NASA: W3C At A Glance

The best thing about this magic is that it is completely invisible. Oh yes, the layout is good, and apart from the strange tree thing on the left of the page it’s almost a text-book example of good presentation. The magic is how it’s put together and fed. Peeking behind the curtain at the “about” page gives an idea:

At A Glance pulls together news from all corners of the W3C working world into a coherent, hierarchical view of the organization’s recent activities. It is an RSS 1.0 news feed aggregator that interacts with the W3C’s authorization system to provide news to all interested parties with additional news for logged in users. Users with login rights may customise which feeds appear in a custom view.

In other words, they collate internal RSS news feeds, transform them into XHTML and further useful feeds. A very nice bit of magic indeed.

One not-too-suprising reaction to the recent spate of worms and viruses has been a re-examination of the insecurities of Windows. A good number of articles, such as this one from SunSpot.net, and this one from Mercury News have focused on the relative immunity of Mac users:

As the latest Microsoft Windows infection spread across the Internet last week, knocking out thousands of PCs in homes and businesses, Macintosh users did what they usually do during a computer virus outbreak — they continued working.

Is it really true that the Mac is more secure? Or is it simply not targeted by virus and worm authors?

Historically there were very few Macintosh viruses—the best estimate is around 50—in part because of the difficulty of writing a hidden virus that would work and spread in the older versions of the Mac OS. The majority of those ‘viruses’ were actually trojan horses of one sort or another, and generally relied on human intervention to spread.

One of the main things that insulates Mac users from viruses and worms is that they simply are not binary compatible. They just won’t run on the Mac. Of course, the Mac can still spread Office based viruses, which is another reason I’m not fond of Word, Excel and Outlook.

It’s not completely true that Mac users are not affected by worms like the most recent ones. Klez, SoBig and their compatriots are still a major pain because of the massive increase in junk mail, and the way in which the e-mail addresses of Mac users, held in the address books of infected Windows machines, can be used as the apparent sender of this junk.

And, of course, worms still knock on the door if running if you’re running servers like Apache under OS X. Their are, however, two major benefits from the Unix basis of OS X. Unix has been around for a long while, and has been hammered by crackers for a long time. Most security holes have already been closed, and there exists a community of interest to rapidly patch holes as discovered, something Apple has proven itself very very responsive to. The other advantage is that security was built into Unix from the ground up. Partitioning of memory and run-time space is right there in the kernel, it’s a fundamental rule of the universe. Unless a process runs in some sort of super-user mode, it can really only hurt itself. (Ok, that’s not completely true, there are all sorts of brute-force things you can do to clobber the system as a user, but they usually result in a sysadmin responding with similar brute-force clobbering.)

No less a journal than the Washington Post has emphasised what a lot of security folk have said for a long time: Windows is insecure by design, at a fundamental level.

Even if that changed, Windows would still be an easier target. In its default setup, Windows XP on the Internet amounts to a car parked in a bad part of town, with the doors unlocked, the key in the ignition and a Post-It note on the dashboard saying, “Please don’t steal this.”

Among other things, that article points out the important difference between the Unix and Windows universes—the former, by default, does not grant access to system administration functions and calls to the user run-time domain, but Windows does.

The Unix philosophy, for most security issues is: the door is closed, unless you specifically open it. For example OS X has a firewall built in, out of the box. By default, it’s shipped with all ports firmly locked down.

Windows doesn’t have security built in from the ground up. All Windows security features are bolted on the outside. The philosophy has always been: the door is open, unless you specifically close it.

While Windows also has a built-in firewall, by default it has been turned off. In the wake of the Blaster worm, Microsoft said last week that new versions of Windows XP will ship with the firewall active by default.

While vituperative and over the top, Dvorak phrases it well:

The experts—in the industry overall and at Microsoft, specifically—are addressing none of this. Outlook sucks at protecting users, and Microsoft has done nothing. The company has also chosen to give users the wrong defaults, so file extensions are hidden unless you track down the settings to make them visible. Why? So Windows systems can appear more like Macs? Or what? I can do a whole column on dumb defaults.

Can I recomend The Devils Dictionary 2.0? A sample entry:

web standards, noun : A large stick or cudgel, used by the slightly more anal-retentive to beat the slightly less anal-retentive.

Each day presents a new, and so far very witty, definition of a technological buzzword after the style of Ambrose Bierce.

Thinking of Bierce, and germane to weblog as a medium of discourse:

Quoting: the act of repeating erroneously the words of another.

This brief entry is, briefly, rather brief and disjointed because the weekend, briefly, was completely taken over by the convention. More on that anon, when I’m more awake.

I’m preparing an invoice. I need to put the ABN (Australian Business Number) for the entitiy receiving payment in the invoice. Not knowing the most formal way of formatting this number, I punch ‘ABN format’ into Google, and get pointed straight to the “authoritative answer:”http://www.ato.gov.au/content/downloads/nat2956.pdf.

Why do they need to make me download a 56.9 Kb PDF file for a document containing approximately 1,300 words? They could have knocked it up as a web page, complete with juicy styling, and made it accessible to everyone. But no, it’s a PDF file.

Irritation factor: high.

Oh, and in case you’re wondering, the format for an ABN is: “XX XXX XXX XXX”. That’s a 44-byte statement summing up the 56.9 Kb PDF, if I have counted correctly.

Looks like my inbox will fill up with return mail from a virus again today. Sobig is currently burning around the world in a big way.

Simple computer maintenance for operators of Windows computers:

1. install, and run, a virus scanner. Run it every day. Take notice of what it says;
2. scrub Windows and install Linux, or alternately, buy a Mac.

At the very least, eradicate Express and Outlook Express from your machines. These two abominations are responsible for the spread of more worms than Word and Excel are.

Here endeth the rant.

Update, 2:16pm: Chuq comments on really stupid anti-virus software behaviour:

I have a thought. If your anti-virus is smart enough to recognize this as SoBig or Klez, why can’t it be smart enough to delete the message and not return it to someone who clearly didn’t send the damn thing, since those messages both hijack third-party addresses for their sending? It’s one thing to send back a message saying “hey! you might be infected with a virus”. It’s another to send a message saying “Hey! someone is infected with a virus, we KNOW it’s not you, but we’re sending you a message anyway for no good reason, because you won’t be able to find the infected machine, either”. Think, people. Think.

I only just noticed that MSIE 5.2 is breaking the design of this blog. Again. In a different way. I have no idea yet whether it’s breaking on other versions of Internet Explorer—I’ll have to ask around. Blast. Damnation. Irritation.

As well as incorrectly positioning the permalink and timestamp block, it’s making a complete hash out of laying out those little meme-buttons. I may have to put them in a table just to keep MSIE happy.

Via the ever-wonderful Miriabilis, the Compendium of Lost Words

The creator of this site has collected 400 rare and apparently un-used English words, as a homage to the flexibility and richness of the language. The author says that this site is probably the only place on the Web that the words exist. Let’s change that.

There are so many wonderful words here, absolutely perfect for all sorts of occasions: acrasial, blateration, diribitory, nepheliad, pocilliform, ramifactive, schismarch, sophronize, vultuous. I could spend days here. Perhaps that is something I could add to my business card — “Vultuous schismarch.”

Jakob Nielsen nailed something on the head when he wrote about misuse of PDF documents.

Don’t get me wrong—I’m a huge fan of PDF for certain things. If I want to ship a document to someone and have some guarantee it will look the way I want it to when they put it on paper, I’ll use PDF. PDF forms that you can fill out on screen then print—love them, fantastic idea, perfect melding of screen and paper. But Nielsen is right in emphasising the problems with PDF.

I just wish someone would write something similar about the proliferation of Microsoft Word as a document exchange format.

I don’t have a copy of Word. I don’t want a copy of Word. I definitely don’t want to shell out around $AU1,000 to buy Office. Word files are bloated, fat and thoroughly proprietary. So why do people insist on sending them around in email, and requiring them for submission of text?

This isn’t a “lets bash Microsoft and Word” rant. I quite like Word for it’s abilities to knock up big documents with reasonably easy styling, spell checking, grammar checking and so forth. If I had to sit down and write a technical manual (as I’ve often done), I’d probably knock it up in Word—as long as the only way I was going to distribute it for reading was as a paper copy.

No, what I really hate is the idea of using Word as a data interchange format. There are three reasons I loathe using Word as a data interchange format, and the same reasons can be applied to why I hate seeing Word files stored in documentation repositories:

1. Word files are huge. A file with a few thousand characters can easily blow out to a few hundred kilobytes. If you don’t believe me, try writing a one page memo with a table in the middle of it. Count the letters on the page, then divide by the size in bytes of the file. You’re likely to get a ratio on the order of 1:70, or worse.
2. Word files are a proprietary binary format. If someone sends me a Word file, and I haven’t got Word, I probably won’t be able to read it. If I have got Word, and it’s a different version, there’s a very good chance that it will display oddly, if it displays at all. Worse, if someone puts a Word file in a documentation repository, then retrieves it a decade later: they probably won’t be able to read it. This is a Very Very Bad Thing.
3. Word files contain cruft. One of the main mechanisms for transmitting certain viruses and worms is Word. Word files often contain bits and pieces of other files.

It’s pretty simple: if you want to send text from place to place, or store it and retrieve it thirty years from now, use plain ASCII text. If you want to mark it up with some sort of structure, use SGML or HTML. If you want to style it, use HTML with CSS. If you want to add semantic markup and what not, store it using some application of XML. At the heart of all of those, the text is still just ASCII text that you can easily get at, no matter what sort of technology is at the receiving end.

Addendum: RMS has written about some of the problems with Word, and a few ways of telling people why it is a Very Bad Thing to mail Word files. Also, Jeff Goldberg has summarised various anti-Word arguments. Both are recommended reading.

One good thing about the recent worm scare: I’m no longer getting CodeRed knocking on my (firmly closed) door every twenty minutes or so. I’d been seeing it show up in my error logs, with a fairly consistent set of source IP addresses which suggested it was another Bigpond customer. Last Tuesday it stopped: someone out there must have finally gotten around to doing a virus scan on their Windows box.

Question of the week: given that most of the viruses and worms are on Windows boxes, and Microsoft are on a security kick, why doesn’t Microsoft build a virus scan into their OS? (they’re building everything else in, after all.)

Oh well. At the least, the horribly bad security under various forms of Windows give me the opportunity to smirk smugly when I see the viruses and worms futilely hovering outside my firewall.

Artima proves to be a valuable spot to browse, frequently feeding my omnivorous reading habits. Jim Waldo has posted a little article on Knowing and Learning. It neatly discusses an aspect of the IT industry as practiced in Australia that drives me mad: the tendency to employ based on buzz-word compliance, rather than proven experience, coupled with a tendency of most IT workers to cease active learning once they have graduated:

For the best in our field, the training never ends. Serving an apprenticeship may make one a journeyman, and further experience may make one a master. But the best of the masters are constantly training themselves…

The half-life of any given buzzword is about two years. After that, it’s either proved to be another short-lived fad of little benefit, or it’s become a base standard that should be an automatic part of everyone’s repetoire. The best juniors I’ve ever had to train have been that rare handful who reach past the surface-level language issues and grok the underlying technologies and philosophies.

Will the situation improve? Probably not. Which is why I pessimistically expect XP to not succeed in this country—too many job ads which begin: “Wanted: recent graduate with 3-5 years experience in XP, .Net, XHTML 3.7….”, which gathers up the two people in the country who’ve done that at university in the last six months. To get good staff, the job ads should read: “Wanted: open-minded technophile with a willingness to wrap her head around the latest buzz-words, get the job done, then abandon the buzz-word and move on to the next problem…”

Sigh

I’ve put an anti-spam mail link on this page, and the main entry page. Which was when I discovered that the bits and pieces I did yesterday caused the site to stop validating as strict XHTML 1.0. Swings and roundabouts.

I’ve sorted out those problems, and everything is valid again. Now to sort out the ugly archive page.

A few more things knocked off the todo list. More meme-buttons than you can shake a fist at—although I was, believe it or not, restrained—and I put in the ICBM Address stuff.

All that’s really left is to put an email address on the front here, and fix the ugly archive page. Then go back and add some more themes to the main entry page. Then build one or more additional themes in here and move a version of the style switcher in…

I plan to rest on the seventh day.

From the BBC: Idi Amin dies

Only about 50 years overdue…

I’m rebuilding everything every few minutes again… This may or may not affect you, and you may or may not care. If you do, consider this an apology.

I’ve just hit the blog style sheet with a stick, so now it validates, and altered the main index template so it validates as well. Next stop, the interior pages.

All done. Normal services are now resumed.

Chuq has another article up regarding managing group dynamics. The bit which caught my attention, and drew me in, was this:

If you are managing a group of some sort, your responsibility is to make decisions in a way that when the needs/interests of the group are in conflict with the needs/interests of an individual in that group, the group’s needs/interests take precedence. There are always individuals that have trouble with that concept, assuming or demanding that they be the center of the universe, or at least be catered to. Unless the group is about you, any group of size > 2 is about building a consensus compromise among the members so things work as well as possible for as many as possible. Any individual who can’t/won’t accept that compromise isn’t really part of the group, and is a destructive force on the group.

Chuq’s comments are within the context of on-line groups, virtual communities, but I think they’re equally germane to any discussion of any voluntary group. Pure, decentralized democracy doesn’t work—it’s called anarchy—and any group ultimately needs some single individual with the responsibility for sending someone to bed without their supper.

It’s cold and grey. To cheer us up, via Miriabilis, this QuickTime VR movie: Ziggy, the interactive cat.

You could spend hours digging around this panorama site. Our favourite, so far, was the view from the top, although the various shots in New York are great, and this shot of the Opera House from the Manly Ferry (warning: large download!) is jaw dropping.

I’ve spent the day banging my head against various CSS walls, beating the MT templates into shape. The result is what you can see. Don’t bother telling me that the archives page is ugly—I already know that, and will fix it on Monday. There’s a few other things I’ll fix on Monday too, as long as I’m not in the middle of getting a job:

• move the Creative Commons icon to be stored local to this site, so as not to keep hammering their site;
• fix the currently ugly FOAF badge to be less ugly;
• add more memetic links;
• put my primary email address on this page somewhere—meanwhile you can get to it via the main contact page.

Comments and feedback thoroughly welcomed.

This via Chuq. It was noticed on September 11, 2001, and shortly afterward, that for possibly the first time ordinary folk around the world were receiving the most up-to-date news via the internet. This time around, the internet has moved even further to the center as a source of information.

I’m sure that I’ll turn on the evening news and there will be much screen time devoted to candles waving in the subways, and commuters trudging through Times Square. No time will be devoted to the initial concerns that it was a terrorist attack, that there was initial official confusion as to the causes and extent, that there was finger-pointing going on between the US and Canada.

By only dipping in and out of the on-line news and comment stream today I’ve already gotten a clearer idea of what is going on than the “professional” news organs will have, or will present.

Jeff Jarvis has the best copyright notice on his web site:

It’s mine, I tell you, mine! All mine! You can’t have it because it’s mine! You can read it (please); you can quote it (thanks); but I still own it because it’s mine! I own it and you don’t. Nya-nya-nya. So there.

Jeff’s site is worth a visit on it’s own — good design, great content. What more could a poor boy like me want.

A reply to my despairing plea from John Gruber should see the following now working:

• this is an en dash: – (–)
• this is an em dash: — (—)

The trick was, for Textile 1.1, to edit line 274 of Textile.pm from

return $smarty->($out, ‘1’);

to

return $smarty->($out, ‘2’);

Virginal goats under the light of the full moon are now safe.

Just so that I don’t forget, here’s how I put the archive files in subdirectories. I’m not using individual entry archives, so I set the monthly archive file template to:

<$MTArchiveDate format="%Y/%m/index.html">

and the daily archive file template to:

<$MTArchiveDate format="%Y/%m/%d.html">

Before I forget again, this is where the little meme buttons came from:

• The FOAF button from Ian Davis;
• W3C buttons from Jeremy Hedley;
• Others from Taylor McKnight’s big archive.

In case you didn't realise it, this blog will be going up and down like a jackhammer over the next few days while I knock the MT templates and stylesheet into shape. In particular this means the pages will be rebuilt frequently and the permalinks will shift around crazily.

It's too late in the day now, but I'll have to see how the various archiving options behave. I'd like to see all the entries for a day appear on a single page when you select from the calendar, and I'd like the calendar etc on those internal pages.

It looks like a lot of folk are having the problem I was having, whereby Safari kept retrieving an old version of my pages, but would get the right version after getting them with Camino. This page pretty well explains what's going on: the invisible proxy that Telstra are running is caching the page. Camino tells the proxy (more or less) "Really get that page when I refresh", but Safari doesn't.

I'll experiment with putting no-cache pragmas in the header, but I doubt that will achieve anything.

Alright then. I’ve added MT-Textile and SmartyPants, let’s see how they work. Feel free to ignore this entry, and the extended entry below, since all I’m doing is taking the tools for a drive around the block.

Now, according to Brad Choate, the following section should be a blockquote (and I must remember to add a style for that):

A lowercase ‘bq.’ and a space will define a blockquote block.

Brad also says that the following should be a numbered list:

1. first item
2. second item
3. third item

whereas this will be an unordered list

• an item
• another item
• the last item

Sweet! Don’t you love it when a plan comes together? No more manual entry of &lsquo; and so forth. (Please note that I could have emphasised with strong, or put in a citation markup, not that anyone really uses that anyway.)

I’m interested to see that MT Textile is using the <del> tag, which nobody uses , and all sorts of magic markup which you can go and read for yourself.

Lovely stuff. Oooh! Acronyms. Now I can use TLA to my heart’s content (assuming Safari handles them (a test reveals it doesn’t yet, as I expected, although Camino does)).

Anyway, I’m happy with just simple quotes, or should I say “simple” quotes — since nobody is handling the W3C quote entities properly yet — because all I really need to do is hammer out text and have my quotes, apostrophes and em– and en– dashes come out right.

Or are they… I’m having some trouble getting the em dashes to work, it keeps showing up as an en dash followed by a hyphen. Maybe it’s time for me to think about this in the morning instead.

Alright, thinking is in progress. Something that is really pissing me off is that Safari, or something between here and .Mac, is caching my pages and not refreshing them. I am inclined to think that it’s Safari, since Camino seems to be figuring it out.

Alright, that was weird. Hit the page in Camino, then come back and refresh in Safari, and the latter finally refreshed.

I’ve added smarty_pants=2 to MTEntryBody and MTEntryMore in two templates. Lets see what happens.

Nothing. Absolutely nothing.

This should be an en dash: – (–)
This should be an em dash: — (—)
This should be a hypen: -

Still not working…

OK. I can confirm that SmartyPants is getting called, because the quotes and en dash are being educated. So why isn’t the em dash? maybe it is? put some in by hand.

Groan: I’m getting em dashes, but not en dashes. So — is an em dash followed by a hyphen.

I’ve sent mail to John Gruber, maybe he can figure it out.

Just to clarify matters, since you’ve reached here: you should consider everything you are seeing to be a draft, and don’t rely on anything here actually working.

In particular, search and comments won’t work, and will go away as soon as I can figure out how to make them go away.

Hmm. Let’s experiment with extended entries and editing previous posts. I meant to put down a ’todo“ list:

• Add in MT-Textile, SmartyPants and similar goodies;
• build a favicon.ico;
• add in the GeoUrl;
• Adopt the BetterBrowser ideas;
• Seriously customise the style sheets, starting by storing them in external files I can edit with a proper text editor;
• Get the links over there in the meme-bar customised.

I was wondering how to go about pushing this blog to .Mac, and realized that in the short term, there was a reasonably easy solution: alter the location for publishing the files to /Volumes/rhook/Sites/blog/ (and so forth).

This may not be the final solution, as it would be nice to have a local cache as well (that’s how the rest of the site is maintained), but with the changes due to come through with the next version of Mac OS X, I would anticipate that the performance of iDisk will improve.

Alternately, I might build a “push the blog” miniature application. Why can’t I just drag and drop? Because the URI embedded in the files needs to change from something local to my machine to one relative to the final site URI.

There are at least three things I need to do now: beat the blog into my desired style, add all the various meme and other links, and get some of the plug-in goodies to aid in authoring plugged-in.

Thinking is in progress.

Naturally, for the first entry, a test entry. Based on a number of websites, soon to be annotated, I seem to have sucessfully installed MovableType. There will be a lot of things to do to get this installation finished, including, of course, making sure that I can automagically publish it to .Mac, but this is a humble and very dull start.