RichText

I posted over on SwirrlSpeak today about Swirrl’s adventures in Rails.

Sorry for cross-posting, but thought it might be of interest to some.

I’ve enjoyed David Heinemeier Hansson’s Myths of Rails series of blog articles. In fact Myth #1 inspired me to stop delaying and finally try out Phusion Passenger.

It really is as simple as DHH says. Literally all you have to do is install the gem, run the installer and follow the (very simple) instructions which guide you through installing any missing prerequisites and setting up your rails app. For more details on installation and configuration, see the Passenger user guide. I got this blog moved over to Passenger a few hours after first visiting Phusion’s site.

Phusion’s Ruby Enterprise Edition is also worth a look – I’ve not done any thorough measurements myself, but on first look it does seem to use less RAM than ‘normal’ ruby.

If you like Passenger as much as me, you can buy (as I did) an enterprise edition for whatever price you think it’s worth, Radiohead-stylee.

Last week I was on the Swirrl stand at the Future Of Web Apps Expo in London.

I didn’t get to attend many of the talks, because the stand was quite busy. As with last year, I actually found the business track more appealing than the developer track. This is not due to the quality of the speakers, but rather the nature of the material – it’s hard for the speakers to give any unique insights about development that can’t be gleaned by reading books or searching the web.

My favourite presentation of the event was by Tim Bray (watch it here). I didn’t actualy sit in on this one, but Swirrl’s stand was just outside the business track auditorium and there was quite a bit of sound leakage, so I could hear what he was saying fairly clearly. I got the impression that he really believed what he was saying and he was giving it to us straight. Apparently, he’d changed the content of the presentation at the last minute due to the recent economic events, and he gave some very relevant advice.

Despite the wealth of ‘famous’ (in the geek sense, at least) speakers at the event, most of the value of the conference for me, was about talking with people about Swirrl one-on-one. It was interesing and enlightening to find out what people thought of our product, and for those who had never heard of it, discovering what worked best when I tried to explain it.

It was also great to meet some other UK-based ruby developers. On Friday Alex Young (founder of Helicoid, a company producing rails-based web apps), and Simon Starr came by our stand. We found that we agreed about a few things: that the UK ruby/rails scene is somewhat lacking; that it is possible to succeed outside Silicon Valley; and that there is a worrying trend for companies popping into existence purely with the intention of being bought out for millions. There’s an infinitesimal chance of being the next Facebook or Google, but a fair possibility of making a decent living from an Internet company. I believe the expected returns can actually be higher if companies ‘Get Real’ and start thinking seriously about just making a useful application that solves a real-world problem.

Since we released Swirrl to the public (my company’s rails-based, semantic data-wiki), we’ve been using Hoptoad, the “app error app” from Thoughtbot.

As was to be expected, Swirrl had a few minor teething problems in its first few days (as people started using the app in unanticipated ways), and I don’t think we could have handled everything so smoothly without our little green friend.

Hoptoad is a breeze to set up: To get started, you just install the plugin, change a bit of config in your app, and you’re ready to hop with the toad. The readme file that comes with the plugin gives you details of how you can customize things to your liking (such as which exceptions to ignore etc).

Once you’ve got it set up correctly, the need to scour your log files for errors is completely removed. It also has major benefits over rails’ standard error notification mechanism. Hoptoad lets you “reclaim your inbox”, as instead of mailing you all the error notifications, it presents them to you in a clean and uncluttered interface (no unnecessary bells and whistles – just the way I like it!).

As you’d expect, it provides all the important information about each individual error, such as exactly when it happened, the call stack, session-data, and information about the user’s browser etc. And once you’ve dealt with an error you can mark errors as ‘resolved’ – you can filter these from your list to help maintain a state of zen-like order. If you do want to be notified as soon as an error occurs, you can opt to be sent an email, or subscribe to your personal hoptoad RSS feed.

So, by using Hoptoad, when the unexpected inevitably ocurrs (nobody’s perfect), you can quickly identify the problem, resolve it and send your users on their merry way once more!

Swirrl, the project that I’ve been working on for a while now, has recently opened its doors to the public. You can now sign-up for free and try it out.

I’ve shamelessly plugged Swirrl a couple of times before in this blog, but never really explained what it does.

Basically, it’s a hosted wiki service, but instead of just allowing users to create text entries, it also allows the creation of what we call Data Set entries using a spreadsheet-like user interface. But it’s not just another simple online spreadsheet.

Using contextual clues provided by the user, Swirrl builds up a picture of your data and it’s structure behind the scenes, allowing some powerful searching and analysis. So, it’s it easy for the user to add semantic information to their data in the context of their normal work, with no need for any manual entry of semantic mark-up.

(Swirrl follows the principles of the Resource Description Framework, or RDF, which is one of the building blocks of the semantic web. However, you don’t need to know anything about RDF or the semantic web to use Swirrl).

So what’s the point of this? Well, with Swirrl we hope to help make the information held in an organization more accessible and exploitable. We hope that our approach provides a simple and effective way to collaborate and share information. With our Data Sets feature we aim to combine the power of a database and the flexibility of spreadsheets with the collaboration potential of a wiki.

Anyway, I’d love it if readers of this blog could sign up and have a play around, and let us know what you think (email hello@swirrl.com). We’re very keen to get some feedback and comments, so that the service can be improved. Thanks.