Fuddland

Entries which have associated advertising provided by Google Adsense.

This category is a subcategory of Indexed.

Here are a couple of small problems I had whilst upgrading to Movable Type 3.3, the solutions of which I’m putting here in case anyone else has similar troubles:

• During the database upgrade process, the error message “Categories must exist in the same blog” means that, for some reason, one of your subcategories has a parent category in a different weblog—the most likely explanation for this is that your database has gotten a bit screwed up. The solution is to either:

  • restore your backed-up database, then use the previous MT installation to change the parent category;

  • edit the database table mt_category directly, using phpMyAdmin or a similar tool if your host provides it.

• If, after rebuilding your weblog, you find that all your weblog entry pages are giving a 500 Internal Server Error message, it may be due to MT setting their permissions to 666. To remedy this, delete all MT-generated content, add the following lines to your mt-config.cgi file and rebuild:

  UploadUmask 0022
UploadPerms 0666
HTMLPerms 0666
HTMLUmask 0022
DirUmask 0022
DBUmask 0022

Today I needed to transfer some of my hard-earned cash to my UK bank account via the Bank of China, so I thought I’d list what information one needs to provide in order to fill out their form, for anyone else who might be searching for help beforehand:

• A photocopy of your passport

• Your bank’s registered address

• Your bank’s SWIFT/BIC [Bank Identifier Code] number [this is a unique code of the form XXXX XX99 assigned to each bank for easy identification], which you write instead of your bank’s usual name on the form

• Your IBAN [International Bank Account Number, which should by EU law appear somewhere on your statement]

Make sure you leave enough money in your pocket to pay the transfer fee—how much you will be charged depends on how much you are transferring, but it can run in a few hundred yuan for transfers of a couple of hundred pounds.

coComment is a new service that allows you to keep track of any comments you make on weblogs; it’s currently in the early stages of development and needs a bit more tweaking to make it work perfectly with the most common browsers and weblogging platforms. Here’s how to automatically allow anyone signed up to its services to keep track of comments they make on MT-based weblogs.

Read the rest of “Integrate coComment with a default Movable Type installation”…

Filed under “I shoulda sorted this out ages ago but it became a priority for me so I finally got around to doing it” was finding an easy way to guarantee that I’m always sending emails over a secure connection, regardless of how I’m connected to the network [whether directly dialled-up to my ISP, plugged in to someone else’s network, or over an unsecure wireless connection].

Perhaps it’s not common knowledge, but generally, when connecting to one’s email service provider using a client such as Thunderbird, Eudora or Outlook, the usual and default set-up is to send the username and password in plain, unencrypted form—anyone listening-in could simply read off your details and have complete access to your account. This is very Not a Good Thing. So the ability to always guarantee a secure connection, at least when sending messages, is quite a handy thing to have.

Why can’t you rely on your own ISP to provide you with a secure connection when sending emails? Most ISPs claim that sending is “secure” because you can only send messages via their [SMTP] servers if you have connected directly to them. But this is no use whatsoever if you’re not dialled-up to them, such as on an unsecure wireless connection, but you still want to send emails.

My solution: use Google’s secure Gmail server. All you need is a Gmail account and to be using an email client capable of establishing secure connections—Opera’s M2 mail offers this facility, as do Thunderbird and Outlook, and probably almost every other modern piece of email software. [It might be called “SSL” or “TLS”, or just a plain old “secure connection” in the settings.]

Read the rest of “Secure sending of emails, wherever you are”…

At some point, presumably around the time Audioscrobbler became one with Last.fm, the format of the RSS feed changed, so it’s time for an updated version of my original feed parser. [Requires a PHP-based site.]

If it’s something you’ll find useful, just:

• install MagpieRSS if you don’t already have it on your server

• download the new version of the script and change the settings at the beginning

See it in action on this very site [of course I’ve added an extra bit to the code which filters out all the Spice Girls tracks I listen to].

Thanks to Robert for making me aware of the change—check out his alternative script if this one doesn’t float your boat.

Prompted by Gordon’s Laziest-of-the-Lazyweb-pleas plea, I’ve made a quick modification to my original Flickr-feed script, so that it now applies various effects [including greyscaling] to the grabbed images. [For example, mine are currently sporting a hint of a bevelled edge.]

You will need:

• The script [rename to flickr.php]

• MagpieRSS

• The GD2+ Thumbnail Class Distro

You’ll also need to ensure you have the GD Graphic Library installed—I’m not sure how you go about checking if it’s there [beyond the script failing]; it appeared to be readily available on my server so I’m assuming it’s fairly common for it to be installed.

See the original entry for more detail.

Read the rest of “Flickr feed parser update”…

Cutting to the chase: I’ve knocked up a PHP script which takes any Flickr photostream RSS or Atom feed and makes a Flickr “badge” out of it, with some extra bits thrown in to make it all worthwhile.

Download the script. Updated version available. Edit it and fill in the required details. Upload it to your server. Cross fingers. [You’ll also need MagpieRSS. Grab it, extract it, stick it somewhere on your server. No need to change any of its settings. Just remember where you put it so you can tell the Flickr script. Oh, and create a writable directory for the cached feed.]

Read the rest of “Use a Flickr feed to include any photostream on your site via PHP and MagpieRSS”…

Although I’d read about it earlier in the month, Lyle reminded me today about Odeon Cinema’s complete lack of understanding regarding accessibility issues when it comes to website design.

The backstory goes like this: Matthew Somerville coded a very nice, accessible portal to the Odeon website, which provided cinema times and the like but without all the unnecessary Flash and other scripty rubbish. He’s done the same with the UK National Rail timetables and a couple of other sites, and they’ve all been lauded for their obvious benefits.

Sadly Odeon’s lawyers didn’t see it that way, and threatened Somerville with legal action unless he took his site down; fair enough he was using copyrighted images, trademarks, etc., but they claimed he was illegally [or, at least, immorally] storing users’ data. Somerville states that he did no such thing and merely passed the data straight on to Odeon’s own database through their usual channels. Nevertheless, Odeon wanted the site down and so it’s gone. I don’t think that’s a particularly clever way to go about their business—or to put it in Lyle’s more apt words, I was keffed off with this bunch of arse—so I dashed off the following email to their contact address.

Read the rest of “Access this, Odeon Cinemas”…

A few weeks back Gordon mentioned a new web-based service called Audioscrobbler, which sounded pretty interesting to me.

Essentially it builds up a profile of your musical tastes—not by asking you questions about which genres or artists you prefer, but by logging exactly what songs you listen to on a day-to-day basis, using a plugin for your media player [a wide range of plugins have been developed]. This information is then compared to other users’, and that’s where the fun begins.

You can quickly find other people who are listening to the same things as you are, and then check out what other artists they also like, building up a network of “friends” whose tastes you respect; you can see which songs by a particular artist are most listened-to, and which users are listening to that artist or song; you can create or join groups dedicated to your favourite performers and join in with discussions about them.

One feature still in development is automatic recommendations—kind of like Amazon Recommendations, but because it would be based on what one actually listens to instead of purchases or on user-entered preferences, it should be more accurate.

They’re also planning to provide a variety of RSS feeds of collective and individuals’ data; currently the only one on offer is for a user’s last ten songs played. Inspired by Richard’s del.ic.ious feed PHP-based integrator, I’ve knocked together some code that produces a nicely marked-up list of the last ten songs in the form of a definition list:

<dl id="audioscrobbler10">
	<dt>Song title</dt>
		<dd>Artist name</dd>
		<dd>[Album title]</dd>
.
.
.
</dl>

You can format this however you like using CSS—for example:

#audioscrobbler10 dl{
border: 2px solid #000;
background-color: #fff;
.
.
.
}

#audiosrobbler10 dl dt, #audiosrobbler10 dl dd{
margin-left: 1em;
.
.
.
}

Read the rest of “Audioscrobbler”…

I’ve long been an admirer of Lump’s commenting system, which allows people to edit their comment for up to 15 minutes after it was originally posted—it allows links that have been messed-up to be corrected or typos to be addressed [some of us could have done with this a couple of weeks ago over at D4D!]. So I was pretty happy when Mark Jaquith guest-posted on Scriptygoddess a technique for MT-based blogs to add this very function.

It took a bit of time to iron out a couple of general bugs, and then some fairly heavy editing of the script to get it outputting things specific to my layout [and to make that output XHTML 1.0 Strict-compliant], but I think it should be working now. I’m not completely sure it will work across all timezones, but time will tell I suppose.

I’ve set the time-limit of the life of the “Edit this comment” link to 20 minutes and the final time-out to be 30 minutes, so if you choose to edit your comment you’ll have a minimum of 10 minutes and maximum of about 30 minutes in which to make your changes, and you can edit it as many times as you like within that time. Note that MT still sends me an email containing the text of the original comment, so if someone drastically changes the sentiment of their comment, I’ll know about it. [Ooh, that sounds a bit like a threat doesn’t it?]

To further enhance the dynamic category pages, I wanted to take care of situations where a category was defined but no entries in that category had yet been made; ordinarily the page would be output with just a blank space where the entries were supposed to be, which isn’t particularly helpful to the user [it could easily be mistaken for a mis-loaded page, and refreshed a few times before they gave up and went elsewhere].

What was needed was a compliment to the <MTEntries> tag: an <MTIfNoEntries> tag. MT doesn’t have one as standard; a few googles later and I found that Jay Allen had wondered why, and the solution [but not the answer] had been left in his comments. I found nothing further anywhere else, so I just took the code Lummox JR had suggested, saved it as a file called ifnoentries.pl, uploaded to the plugins directory and bingo! I had a working <MTIfNoEntries> tag.

Read the rest of “MTIfNoEntries plugin”…

In one of the weirdest duets ever, for the Man on the Moon soundtrack Michael Stipe performs with Jim Carrey reprising his role as Andy Kaufman [and Kaufman’s alter-ego Tony Clifton], covering Fabian’s This Friendly World, and you can’t help but smile [especially when they sing alternate words during one of the verses].

With the skies so full of stars and the river so full of song
Every heart should be so thankful

[…]

The world is such a wonderful place to wander through
When you’ve got someone you love to wander along with you

With so much interlinking between weblogs it seems natural to try and form some kind of network beyond “he links to them, she links to me, I link to those people”—generally there’s no indication of what relationship, if any, the linker has with the “linkee”.

A relatively new approach to tackling this issue is the XFN [XHTML Friends Network] protocol, and it’s beautifully simple: all it asks is the addition of the attribute rel to your links, with values selected from a small list of options, and it’s aimed in particular at weblogs.

Read the rest of “This friendly, friendly world”…

I’ve never really figured out the best thing to do with entry categories. There are some categories for which I want to gather all the relevant entries into one archives—for example, photo-based or music-related entries—but MT has this “all-or-nothing” approach to category archives: you can either create archives of all your different categories, or have none at all. The former seemed to me like quite a waste of server space [some categories would probably never get visited], so I never bothered having true category archives, and just made bespoke templates utilising the SubCategories and CatEntries plugins. However, rebuilding each one manually was a bit of a pain, and setting them all to update automatically seemed like unnecessarily adding to rebuild time—I really wanted a better solution.

The method I devised is described below, and is probably only of interest to MT-geeks. The rest of the world can just skip to the category archive list to see the fruits of my labours.

Read the rest of “Dynamically-generated MT category archives”…

Last week Yahoo! officially unveiled their new search engine—previously their results had been powered by Google, but now they have their own spiders crawling the interweb.

Since their algorithms for indexing and ranking are not the same, it’s highly unlikely search results will match precisely across both engines. There’s an interesting tool for comparing the search results called, appropriately enough, “Yahoo! vs Google”: it draws lines connecting identical hits in the first 100 search results for both engines. I’ve invented a game based on this tool: it’s called YaGoogle!™, and there are several objectives.

• Find a search that whose results form a series of vertical lines—that is, the results are ranked exactly the same by both engines at some point [or points]. The more consecutive vertical lines, and the closer the group is to the top of the results, the higher the score.

• Find searches where a sequence of results for one engine are the inverse of the other—for example, result 20 in Google is result 80 in Yahoo!, result 21 is result 79,…, up to result 30 in Google [matching result 70 in Yahoo!]. Maximum points for all 100 results fulfilling this criterion.

• Searches that create interesting geometical patterns with the connecting lines earn bonus points.

• Find searches with a single common result [I suppose you’d have to call that a “YaGooglewhack”].

• Find searches with no common results whatsoever.

I’m sure YaGoogle! could become as popular as Googlewhacking, if it weren’t for one small problem: I’ve not managed to get anywhere remotely near any of the objectives. I’ve found the Yahoo! vs Google tool itself is a little buggy, but still: getting any more than a couple of vertical lines seems to be very tricky, and it seems to me that would be the easiest of those four objectives; as the saying I’m about to make up goes: just because it’s the easiest, doesn’t make it easy.

When I wrote the previous entry, I noticed that the link to Daisy’s comment did not necessarily make the comment appear at the top of the page, because there was not enough text below it. In order to make it a little clearer to which comment you’ve just jumped, I knocked together a quick bit of JavaScript which alters the background colour of the relevant comment. [I know others must have done this already, but after four fruitless seconds of googling, I decided it’d be quicker to do it myself.]

Read the rest of “JavaScript highlighter”…

This is the kind of thing only I would worry about. I could try and blame Richard and his throwaway comment about ordered lists, or mrtn for very kindly buying me Eats, Shoots & Leaves, but if I’m honest I’d been thinking about this anyway.

I’m always interested in discussions about correct and valid HTML or XHTML mark-up, and am getting mildly obsessive about making sure I use the right tags for the right job. Only recently I’ve got into the habit of putting <cite> tags around any book titles I mention, because this helps search engines when they’re indexing your site.

Similarly, using <abbr> and <acronym> tags gives important clues to search engine spiders about the topic of the text it is indexing. For example, if a page referring to “bugs in IE” is actually talking about bugs in Internet Explorer, marking-up the first instance of “IE” as <abbr title="Internet Explorer">IE<abbr> makes all the difference.

This week I are been mostly thinking about lists.

Read the rest of “Punctuating unordered lists”…