Experimenting with flickr and GoodReads

As a part of migrating to an external host and also experimenting with an approach of choosing and using software and services for the sake of convenience rather than idealism, etc., I've been experimenting with using a few 3rd party social media applications, and thought I'd describe my impressions of them.

First up is photo management. This is something that was done extremely poorly on my old self-hosted site, I had to admit. I originally had a series of scripts for producing static galleries, pulling in metadata (like descriptions and tags, etc.) from flat text files, but it didn't work particularly well because I simply didn't have the energy to do things like make the gallery look nice (not that I didn't care, but CSS, despite being a fantastic idea in principle, is such a poorly implemented and counterintuitive hack-fest that making a nice image gallery without using tables is far more difficult than it should be) or provide the metadata. After that I experimented briefly with using a third party PHP-based application which didn't require the use of a database, but it was a fairly half-hearted effort and I never got a significant number of pictures into it.

The two big "cloud based" photo management apps are, of course, Flickr and Google's Picasa. Flickr has by far the largest userbase, and looking at a few "Flickr vs Picasa" articles turned up by a quick Google suggested that Flickr is ahead in a lot of important ways, so I decided to give it a try. When I first tried to sign up and realised I would be forced to setup a Yahoo account, complete with @yahoo.com email address, I quickly lost my stomach for the idea: I have no interest in using any other Yahoo services and I don't want yet another unused and unwanted email address cluttering up my online presence. After further reflection, I decided this was a little hypocritical, since I already use a number of Google services which require a Google account: although, to be fair, Google makes this process a lot easier and much more lightweight. Anyway, I bit the bullet and got myself an account.

My experience with Flickr so far has been entirely adequate, but it's not exactly something I've fallen in love with. The interface could use a lot of work, I don't find it particularly straightforward: or rather, the stuff that everybody wants to do all of the time is not done using more obvious or prominent controls than the stuff that few people want to do some of the time. While the idea of a Photostream (a sequence of photos arranged in the order you upload them) is fine (indeed, it makes sense from the point of view of keeping up to date with the photos of friends), I don't think. I'm also really irritated that the ability to define hierarchical sets ("sets" are coherent collections of photos in Flickr, things that might be called "albums" elsewhere) is restricted to those who hold a "Pro account", which costs US$25 per year. I have no objection to the concept of a paid account status, I realise Flickr have costs to cover, but it seems like the features that you need to pay to get should be limited to things which are in some sense "special" or "excessive": increasing the number of photos you can upload per month, or the maximum size of photos, or even the number of sets you can define are sensible candidates. Such a basic and fundamental organisational tool as the hierarchical arrangement of sets should really be a freebie: it's something just about everybody is going to want. Anyway, I'll be sticking with Flickr for the forseeable future.

The second thing I've been looking at is keeping a record of which books I read and when. This is something I never even tried to implement back when I was self-hosting everything, although I've had an account at LibraryThing for some time (but rarely used it). I really like the idea of this kind of thing, but I never kept up to date with it much, mostly because the interface at LibraryThing is so terrible. For one thing, I found it genuinely hard to remember where to go to edit my metadata about the book (such as when I read it or what tags I'd like to give it), as opposed to editing site-wide metadata about the book, such as who wrote it. Worse, LibraryThing treats every separate edition of a book as a distinct entity. There is no way to say "I have read William Gibson's Neuromancer". You are forced by the system to proclaim "I have read the edition of Neuromancer published in 1984 by Ace Books". By all means, the ability to express that information if one wants to should be present, but actually forcing it is ridiculous, especially since if your particular edition is not in the system you are forced to enter incorrect data by choosing another edition. The refusal to combine all editions into a single Platonic concept of the book also really interferes with useful collaborative filtering: you can't learn things about people who also read Neuromancer, only things about people who also read your edition of Neuromancer, which is not a distinction I can imagine many people are interested in.

Frustrated with the above shortcomings, I've checked out GoodReads, an application similar in spirit to LibraryThing but which seems to lack these shortcomings. It has Platonic concepts of books, but also the ability to specify editions if one is so interested. I've loaded a small, initial subset of everything I can remember reading (basically favourites and stuff I've read recently) into my account, and the interface seems clean and intuitive. It remains to be seen how long I can keep up the habit of chronicling what I read, but it's nice to have a tool for doing so which is straightforward enough that I can't really blame it for my failures to do so.