I apparently installed both DBI and DBD::mysql fine but when I tried to access the Movable Type CGIs I was trying to run I’d only see Internal Server Errors and this in the Apache error log:

dyld: lazy symbol binding failed: Symbol not found: _mysql_init Referenced from: /Library/Perl/5.8.8/darwin-thread-multi-2level/auto/DBD/mysql/mysql.bundle Expected in: dynamic lookup

dyld: Symbol not found: _mysql_init Referenced from: /Library/Perl/5.8.8/darwin-thread-multi-2level/auto/DBD/mysql/mysql.bundle Expected in: dynamic lookup

Premature end of script headers: mt.cgi

I did a lot of Googling and found a many people with similar problems with “_mysql_init” while trying to get Ruby working, but precious little concerning Perl. The Ruby solutions pointed toward it being something to do with the installed version of MySQL being 64-bit rather than the 32-bit Ruby was expecting. (I know nothing about Ruby, so I’m waving my arms a lot here.)

Because I’d recently moved over from a G4 PowerBook to a new MacBook, using ‘Archive and Install’, my MySQL was initially a PowerPC version. First I’d tried replacing it with the packaged MySQL for “x86_64” because I had a 64-bit machine, but the same errors appeared. So I tried the plain “x86” package, assuming this would be 32-bit. The same errors.

I guessed that maybe when I originally installed DBD::mysql some configuration was set that made it think I still had the PPC version of MySQL installed. So I deleted DBD::mysql:

sudo rm /Library/Perl/5.8.8/darwin-thread-multi-2level/auto/DBD
sudo rm /Library/Perl/5.8.8/darwin-thread-multi-2level/DBD

(If you have other DBD::somethings installed you should probably delete …/DBD/mysql rather than the whole …/DBD directories.)

Then I installed DBD::mysql again and, hurrah, mt.cgi finally worked.

You should make sure you have /usr/local/mysql/bin in your $PATH — put this in your ~/.bash_profile:

export PATH=/usr/local/mysql/bin:$PATH

Then when you install DBD::mysql it picks up the appropriate settings from mysql_config, including setting which version of MySQL you’re using. So make sure you have the correct MySQL installed before installing DBD::mysql. Unlike me.

Thanks to Mike T and Brad Choate for saving my sanity and I hope this saves at least one other person an afternoon of swearing at Google.

Come on people, get a grip. I thought Ross/Brand was going to be the year’s low water mark of non-news. Leave this shit to ‘Heat’ and start being journalists. Remember investigating? Exposing? Holding people to account? It’s not like you’ve finished everything else and writing about dance competitions is the only news left. We should be proud of you, not regretting you were ever taught to write.

I know your business model is tanking and your newspapers are going to wither and die but you could at least have the decency to go out fighting so we’re sorry you’re gone, rather than breathing a national sigh of relief at your wholly deserved demise. And don’t get me started on your sorry cousins on TV and radio. At this rate, the sooner all mainstream news services are fragmented into obscurity the better for us all.

(Front pages from PressDisplay.com.)

I doubt it’s interesting enough to watch the whole thing if you weren’t on the trip, and it’s not old enough for fashions to look laughably quaint. You may, however, spot the very occasional glimpse of me with hair.

Looking back, it was an amazing trip. The main teacher in charge, Bob Burns, asked us if we’d be prepared to go on a “magical mystery tour” — pay the money, take our passports, and leave it to him to arrange the holiday. Sorry, intensive field study.

We set off from Bristol in a coach, heading for the coast. The first stop was… Bognor Regis. Butlin’s. But that was just for a night and from there we drove to Innsbruck, Berlin and Venice in, I think, ten days. We saw a lot, had a lot of fun, and even did a bit of drawing occasionally.

One of the highlights was laying out a huge square of paper in the middle of St Mark’s Square and everyone kneeling down to draw the surroundings, Bob somehow managing to convince the officials that it was all fine.

When we returned to Bristol, a large marquee was set up at college and we spent a week or two recreating some of what we’d seen inside — a merry-go-round, a Berlin wall, a little canal, etc. — for an exhibition.

Some time later Ned, Tim and Chris managed to edit all their footage of the trip and the exhibition into this film. There are plenty of indulgent studenty things in it but I think it works pretty well. Good going.

But watching it through I was almost struck more by the few shots of college corridors at the very end than I was the sunny continental locations. I love art schools and those corridors made me all nostalgic. You (or I, certainly) don’t realise quite what you have while you’re there. I really must get back to drawing more.

Facebook has always made me a little uneasy. On the one hand I enjoy having so many “friends” in one place, far more than on any other site thanks to its audience reaching beyond the usual early-adopting geeks. On the other hand I never feel I get it. I don’t feel at home there and the way people post private messages for each other in public is plain weird (even for a compulsive Twitterer).

But after reading Roo mention piping his tweets to his Facebook status I thought I’d give it a go. The result mirrored my split Facebook feelings.

At first it was pleasant to feel more in touch with all those friends who, for me, only exist online in Facebook. Many ex-work colleagues and acting friends don’t Twitter or Flickr… they’re only on Facebook. So I enjoyed getting occasional comments from these people on my automatically updated Facebook status.

However, it also felt cheap and wrong. I wasn’t spending more time on Facebook; I was blindly firing stuff in there, occasionally receiving a gratifying morsel of attention in return. Maybe I’m old fashioned but this didn’t seem fair.

This became a real problem when one or two Facebook-only friends were upset that I didn’t respond to them. I don’t understand why people send messages on Facebook rather than email and I often end up not replying. This attitude must seem extra arsey if I’m still updating my Facebook status frequently. There was an imbalance between how much I seemed to be using Facebook (a lot) and how much I was actually using it (very little). I’ve now stopped my tweets being sent to Facebook so I can more comfortably ignore the strange place and all those weird normal people.

Automatic posting to different services is a tricky thing. Well, it’s easy to do but it’s tricky to do right. Several services, such as Blip.fm and Gyminee, allow you to copy messages directly to your Twitter stream about your activity. Easy to do but liable to drive all your friends mad as the messages always read like self-promotional spam. They’re of a different nature to the usual tweets.

Now I’m also looking at aggregators like my Jaiku page and my FriendFeed and wondering if I’m doing something similar. I never use either site and yet I’m pumping oodles of feeds and services into them both. This feels like owning a holiday home in a seaside town — you occasionally make a show of being involved with the place but you’re not really part of the community. And when you’re not around the locals are probably slagging you off for ruining the place.

A different solution would be to turn things around. Rather than you aggregate all your feeds together and I have to read them all, I should be able to choose which of your feeds I want to read on Jaiku, which on FriendFeed, which on Twitter, etc. It would probably be a nightmare of an interface and a struggle for anyone to keep up to date, so I’m not sure it’s the ideal solution. But I’m not sure the current way is ideal either.

I’d love to hear from anyone who actively uses sites like these and how they find the mixed flood of friends’ feeds.

They’re back now. If you want to catch up on stuff that fell through the gap you can page through previous days on the site. If you ever notice the feeds or the site falling unusually quiet for a couple of days in future feel free to give me a prod to make sure they’re OK.

It’s a bit clearer at the larger size. Along the top are five of my own websites. The red arrows show things that I update by hand — writing weblog posts, adding Delicious links, uploading photos, etc. The grey arrows show one website automatically updating another without my intervention.

I started thinking I’d create a gorgeous graphic but soon realised it was going to be too complicated and interweaving to be as elegant as I first imagined.

A good proportion of those sites I almost never visit: Pownce, FriendFeed, Tumblr, Jaiku, Brightkite, LiveJournal and Fire Eagle (which is designed so you don’t have to visit it). They trundle along, getting updated by the ghostly fingers of APIs without me actively using them. Quite magic, almost scary, and possibly a little pointless.

On the off-chance you want to draw your own graphic — I bet some of you would have far more complicated ones — and you use OmniGraffle, here’s my original document to give you a head start with the logos etc. Do post a comment if you make your own.

I haven’t done this because I have an over-inflated sense of my importance and believe every word I type should be saved for posterity.¹ It’s more that I’m trying to slowly push this site toward an example of what I think a personal website should be these days: an aggregation of all a person’s online activity.

Once upon a time home pages were little more than static lists of your favourite things or photos of your cat. Maybe you went as far as linking to other stuff on the web, a behaviour that gradually evolved into weblogging. But for some time there still wasn’t much distributed personal activity that could be aggregated.

These days, of course, many of us are generating much more activity online but it’s spread increasingly thinly among many third-party websites. I eventually want my site to be a single collection of everything I do online in public. So far, aside from the writing that only happens here, I also aggregate links posted to Delicious, display my most recent Flickr photos, include posts from my other site, and feature my most recent Twitter. (And when Twitter fix their API I’ll archive all my Twitters here.) Aggregating offline, real world, activity such as my reading and music listening is a bonus.

There are tools to help us do this aggregation. For example, there’s Sweetcron or Movable Type’s Action Streams or FriendFeed. One problem with these, for me, is that they’re too mechanical. They treat every item as equal, whether it’s favouriting a YouTube video, Twittering a thought, posting a photo to Flickr, or writing a long essay.

Some people love this kind of aggregation. Good for them. I, however, am human and my eyes glaze over when trying to comprehend a chronological stream of equally-weighted events, a format only robots could love. This is rubbish. (Examples: Sweetcron, Action Streams, FriendFeed.)

There must be better ways of showing such “here’s what I’m up to” information. My home page isn’t the answer; the bulk of it uses an overgrown weblog format and suffers from using identical treatments for different kinds of content. But I have started to make it more human friendly. For example, if I post dozens of photos to Flickr in a day only a few appear on the site, with a link to see the rest, rather than swamp the page. And Twitters, whose value decreases rapidly with time (unlike longer essays), aren’t part of this chronological stream. But it needs, and will get, bigger changes.

However, this “here’s all the recent stuff” format isn’t the only way to aggregate things. For nearly four years I’ve been using my own “on this day” aggregation. For example, the sidebar next to this piece of writing shows Flickr pictures from the same day, a comment I posted on another site, what I was reading that day, and what music I listened to that week. This page of aggregated Delicious links is another example. Some day I’ll add the day’s Twitters and other events.

Another reason to aggregate everything as completely as possible — ie, saving the text of a comment on a friend’s site, rather than only a link to it — is that I want to physically own my own words (and pictures and…). We’re dispersing our activity around a vast number of Web 2.0 services, some of which will inevitably go bust, merge with others, lose data, stop being free, etc. I want to think further ahead and keep a record of what I’m doing and saying now that I will still be able to access in decades time.

How to provide all this accumulated information to others is another problem. I occasionally read people complaining about information overload from friends. We’re doing so much and trying to tell everyone about everything. My attempt at doing things right so far amounts to providing separate RSS feeds for each of the main kinds of data I’m aggregating, plus a single combined feed of everything together. There are probably even more flexible solutions. Be a good friend and let people choose how much of you they want to read.

I’m not saying “I’m doing everything right, all other solutions are useless”. I have a long way to go before the at-a-glance “what’s Phil up to” dashboard is as good as it could be.

And I’m not saying “you too should spend days writing code and wrestling with content management systems to do the same”. Hopefully this kind of aggregation will become easy enough that everyone can do it by default. But, so far, other mechanical aggregation solutions aren’t designed for humans.

Footnotes

1. In fact, while putting this archive of my comments together I realised how repetitive and disagreeable I can sound. One person’s helpful correction is another’s overly superior put-down. One person’s illuminating story is another’s self-obsessed conversational dead end. ↩