a work on process

Testing PHP apps with Ruby tools

2008-11-10T13:00:42Z

As I’ve mentioned here before, when working on web applications built with PHP, whether custom-rolled or drupal-driven, I often find myself missing various tools from the ruby kit. I’ve talked before about using capistrano with non-ruby code, but lately it’s been rspec and its stories that I’ve been craving.

I’m aware of PHPSpec and have played with it from time to time, but the lack of a compelling way to work with mocks/stubs has slowed my adoption, and last time I checked it didn’t offer anything for high level user stories. So this week I set out to harness cucumber and webrat to write some simple stories.

It turns out to be pretty easy. There’s no nice simple support for test environments, fixtures, mocks or stubs, but if you just want to make sure that a few pages load correctly, and have the right elements, or that logging in works as you expected, then it’ll do the job.

I’ve not done any packaging up of the code, mainly because there’s so little to it. My folder structure is:

specs/
  Rakefile
  features/
    admin_articles.feature
    steps/
      admin_steps.rb

(click on the links to see sample files)

I simply set up those files, go into the folder and type ‘rake features’ to put your site through its paces.

Recommend this post:

New home for Rails ‘geo plugin comparison’

2008-10-22T09:40:07Z

About eighteen months ago I compiled a series of reviews of Ruby on Rails plugins concerned with geography. I put together a comparison chart and posted it on this blog. It subsequently found a new home on a wiki, but lately that wiki has rarely been accessible so I decided it was time to move it all back into this site.

You can now find the comparison chart at: http://jystewart.net/process/resources/rails-geo-plugins/

A few updates have been lost along the way as they were solely made on the wiki, but hopefully it’s still of use. Since I published the original reviews and chart my attention has wandered a little from the geo plugin scene, so please do flag up any new plugins, changes in features or fixes that I may have missed. I’m going to be trying to check through all the existing listings to update them but that may take a while, so comments here may well encourage me to focus more quickly.

Recommend this post:

Play and Social Media Training

2008-10-20T10:41:58Z

Over at Netsquared.org Amy Sample Ward has posted another of their regular ThinkTank questions. This time around it’s:

What are the key questions nonprofit orgs should ask to help them determine how to prioritize social media training and experimentation as they do their technology and organization-strengthening planning?

I’m coming in a bit late. There are some good responses appearing, such as those from Ashley Messick and Beth Kanter which offer a number of key questions to consider when developing a strategy for your organisation. The responses to date are summarised on the netsquared site.

The element I’d like to emphasise is finding out who in your team already plays with these things and how. There’s an increasing chance that there will be people in your organisation who already devote some of their free time (and maybe even work time) to activities that fall under ’social media’, whether or not they recognise it: staff members with facebook profiles, a keen photographer who meets others on flickr, or someone who throws the odd video up on youtube. There are a lot of people who have used ’social media’ in some way shape or form. It may take some careful facilitation to turn those playful experiences into strategic ideas for your organisation, but they should not be overlooked.

In a strong sense, most of your audience’s interactions with social media will fall under the category of play. These websites and media are all about social interactions and free-form creativity, and it’s vital to maintain some of that in your training and your approach to engagement. You don’t have to turn everything into a game (though the power of games is well worth considering) but people are much more likely to remain engaged and your staff are going to remain energised if there’s a sense of playfulness about how you use social media.

Recommend this post:

acts_as_amazon_product

2008-10-15T14:56:16Z

A couple of years ago I wrote and released a Ruby on Rails plugin called loads_from_amazon. It made it relatively simple to populate a model with data based on an amazon search, and was very helpful in the project I was then working on.

That project ended and I’ve not had time to maintain the plugin since. It was based on a clunky amazon ECS library and I kept meaning to rewrite it to sit on something more up to date, like amazon-ecs, but the time never materialised.

Today I stumbled upon acts_as_amazon_product which looks like it does everything my plugin did, and more. If you’re looking for that functionality, it seems like that’s the place to go.

Recommend this post:

James Stewart says: Find Me

2008-10-12T17:03:56Z

So far as I’m aware, there’s not been much to link a small street near Victoria with Lesotho, South Africa, and 47 countries. Until now.

Thanks to a collaboration between photographer James Nachtwey/XDRTB.org and moblog.net one may emerge in the next few hours.

Gaining (maybe even seizing) attention is key to any campaign. Connecting causes with play (tastefully, of course) is a great way to attract it and (if you can pull it off) mystery is a potent extra component. I’m excited to see what develops in the next few hours.

Recommend this post:

Route Blogging

2008-10-12T17:07:09Z

At dConstruct in Brighton last month Steven Johnson talked at length about his startup, outside.in, that collects together place-blogs to display aggregates commentary and information on a block, neighbourhood and city level. The site is US-only at the moment, which made it a curious presentation for a largely non-US audience. Their toolset for extracting geographical information from blog entries is impressive, but a number of us were talking afterwards about what the real value of such aggregation is for those who might already live in an information-rich or tightly knit neighbourhood.

I find outside.in handy for keeping up with news from Grand Rapids now that I no longer live there, but even if it were available in the UK I suspect its main use would be for identifying other local people rather than learning what’s going on. And even then, there are existing community discussion groups and other fora that work well.

What would seem really compelling is a way to connect with the places that only get our partial attention. A couple of times a week I take a bus past a local leisure centre. It’s not somewhere that I’m likely to use at the moment, but with big life changes coming up it may be somewhere I’m more interested in before long. It wasn’t until after the fact that I saw that it was the centre of some controversial planning decisions, far too late to engage in the debate.

We all travel along a variety of paths all the time. Often only a few of the places on our routes get much of our attention, but most of them are of some importance to us. I’d really like to have a site that was aware of my most common (or preferred) routes and helped me engage with them more deeply. Presumably it wouldn’t be much more work than that outside.in does to assemble place-blogs. That geo-data just needs to be mapped to a trajectory rather than a simple locale.

Recommend this post:

Slow Burn, Quick Conversions

2008-10-11T13:25:28Z

In the midst of the ongoing financial crisis it’s been satisfying to turn to a project I announced here a few months back for commentary and interpretation. As Ann Pettifor has been called upon for commentary by numerous conventional media outlets her blog, Debtonation, has really come into its own.

It’s a good case study of how the slow-burn of building up a body of content about a key but under-represented topic can suddenly turn into a serious number of comments and traffic. The commentary on the blog, based on many years of public commentary, was relevant as soon as it started appearing, but now that it covers the biggest news story of the past month it gets the attention it deserves. It’s still mainstream media appearances that drive much of that traffic (links from the Guardian, BBC, abc.net.au, et al rank high in the referers), but it’s the option of Stumbleupon and co. that allow that new audience to act as amplifiers and sustain that traffic. While the biggest days have come from radio appearances, the impact on traffic from a couple of stumbles is remarkably close to that from mass broadcast media.

Steve, Carl and I were talking at Social Media Café yesterday about the local peaks in attention that occur for musicians as a result of external events (press coverage, an opening slot with a big name, radio play) versus the steady growth in sustained attention that will come if you really engage with your audience. That’s not limited to music. It’s precisely what we’re seeing with this blog.

The content is distinctive and both timely and built up slowly. Arguably the key contribution a commentator can make at the moment is to ensure that considered and diverse voices are being heard by a large audience, but tying one voice’s contributions together on a blog allows the deeper engagement that will come with discussion and the space to make more nuanced arguments. The blog provides a forum to engage and to convert casual attention to sustained attention; RSS and email subscriptions provide a way to further engage and potentially mobilise them.

But perhaps what’s most helpful to remember here is that for the first few months traffic to the site was low. Clearly there was the beginning of an audience early on, but what has made this blog work so far has been the ability to quickly harness and convert a casual audience, turning it into subscribers, and allowing it to further amplify the coverage you’ve been handed.

Recommend this post:

Book Review: Learning Drupal 6 Module Development

2008-09-03T09:11:54Z

Last autumn’s release of Pro Drupal Development was a significant moment in the history of the popular CMS, providing for the first time a relatively comprehensive guide for those wanting to do more than simply manage and skin a drupal site. A number of books have followed it but few have delved as deeply or been such a definitive guide.

Like most of the more recent books, Learning Drupal 6 Module Development focusses on a quite specific area of drupal development, but its a key one for any serious developer and touches every other area of the system. Experienced PHP developers may find that this book (in conjunction with some time for experimentation) will serve as a solid introduction to how they might build applications on top of drupal.

The book focusses on a single project—a website providing biographies of philosophers—and builds the modules it needs, introducing the various available tools and techniques along the way. Much of the time is devoted to generating custom content types, but there’s very solid coverage of the hooks, filters, and actions that let modules really take integrate with the rest of the framework. Theming your output, using AJAX, and working with web services all get some time and illustrate how your app can be part of the wider web and keep up with its prevailing trends. A number of times I found myself reaching for code I’d written over the past few months to make amendments based on examples in the book.

I’ve been quite critical of books from packt lately and some of my criticisms apply here—he book itself feels flimsy and the print quality is poor—but this is also evidence that if an author and editor put the work in they do occasionally produce quality material despite the publisher. There are a number of asides that clarify language which demonstrate an attention to detail too often missing. It would have been good to see clearer signposts as to which features are new in Drupal 6 and which were available previously, but the online API docs can provide most of that.

When I reviewed Pro Drupal Development I noted some disappointment that the book didn’t devote any time to automated testing of drupal code, and that criticism applies here too. Along with staged deployments, automated testing remains one of the least considered aspects of drupal and that’s a serious concern for those looking to build robust well-managed applications on top of it. There are a few good articles online about how to test drupal code, but it would be good to see it taken more seriously as a core part of the module development process.

This book is likely to sit alongside Pro Drupal Development on my desk whenever I’m working on a drupal project and is a worthwhile investment for anyone who spends much time building drupal modules. There are clear areas for improved coverage, but it is as comprehensive an account as you’ll find of how to build modules that take full advantage of the facilities Drupal provides.

Disclaimer: I was sent a copy of this book for review by the publisher. You can find it at amazon US, amazon UK and all sorts of other places.

Recommend this post:

Greenbelt Social Media: What was different this year?

2008-09-03T09:09:19Z

Yesterday, responding to a post Steve wrote on our Social Media efforts at Greenbelt I noted that it’s important to remember that this wasn’t the first year we’d worked with social media at the festival. Flickr has been our most prominent outlet, with the festival’s tags being some of the most visible in the week following the festival for several years now. But as I’ve written about here in the past (from a fairly techie perspective), we’ve made efforts to aggregate content from multiple blogs, social bookmarking services, and the like a few times previously. So what was different this year?

As Steve points out, video is a significantly different medium to photos or text and it has its own set of hooks. This wasn’t the first festival video to be posted online—a few videos had snuck onto youtube in previous years—but it was the first time tools like qik were available to allow live streaming. As I noted a few days ago live streaming currently benefits from its novelty: “this is streaming live on the internet” is a great hook for drawing in guests and viewers. That may well not last, just as blogging has lost much of its mystique over the past six years, but this year it served us well.

The “embeddability” of the content is a very important factor. We’re all pretty used to embedded youtube videos at this point, but it’s only been in the past few months that its become the majority of media storage sites that have offered facilities along the lines of what Dan Hill dubbed ‘tear-off’ content. That’s significant in a number of ways. We didn’t have time to really develop the platform for what we were doing (we’d wondered about using Alfie’s moblog platform but ran out of time) but we knew that if we used qik we could not only export the video later, but we could very quickly embed widgets into blogs and other sites to promote the content. That freedom from worrying too much about platform is liberating, but for achieving attention in a festival environment it’s the ease of embedding that’s key.

Twitter was, of course, a vital component of our strategy. Just as there was no time to build up a platform for aggregating the content, we didn’t have time or budget to do any real promotion, and since this was a very experimental approach we didn’t even have time to build it into the editorial content of the festival’s own website. But we’ve all got relatively large personal networks on twitter (and for some of us our twitter posts are syndicated into facebook) and we’ve been cultivating a Greenbelt twitter account and it was easy enough to post notes there. Whether posting automatically (”I’m streaming live on qik …”) or personally, we saw a very good response and were able to receive some quick feedback. Twitter works really well as a glue between pieces of content you’re generating around the web, acting as a hub for a network that will follow link and engage with content hosted in a variety of locations.

Perhaps the key non-techie reason that things felt different this year was that there was concerted effort from a team. The real turning point for our flickr presence was when we started posting the festival’s official photos there—it gave it a certain kudos for those festivalgoers who may have been reticent and meant we were promoting flickr heavily in our editorial—and similarly having a group of people establishing a body of content provided something resembling a critical mass. Since our online networks intersect fairly heavily there was some reinforcement (”oh, X and Y have both mentioned this, I should check it out…”) but there’s enough distinction that the message went wider than any one of our personal networks. As a team we were also able to exchange skills and discoveries through the weekend which helped enormously when we had so little time to get up and running.

In purely numeric terms flickr is still where the vast majority of social media attention around the festival rests, with views of the photos being an order of magnitude greater than of the videos. Much of the conversation is taking place among blogs, with many scattered posts picking up a few comments. It’ll be interesting to see whether video capture at the festival follows in the footsteps of flickr and attracts a much larger group of producers or whether it remains an activity of a fairly small group. Either way, we’re very pleased with how it worked out this year.

(photo above is by Jon McKay, from his ‘So What Do You Think?‘ project)

Recommend this post:

Lessons from the Greenbelt Social Media project

2008-09-03T09:09:39Z

This post is a follow-up to my initial thoughts on our Social Media efforts at this year’s Greenbelt.

Going into Greenbelt I’d made some fairly naive assumptions, primarily that it would be easy enough to just capture conversations we were having anyway and events we were attending. For people whose sole responsibility at the festival was reporting that might have been possible, but for those of us who were already deeply committed to other activities it’s not quite that simple. While Steve, Lisa and Mike were able to gather a lot of great material, and made the capture their primary focus, I was more distracted and my efforts are much thinner on the ground, and decidedly patchier.

Steve was using a Nokia N95 and the rest of us had N82s. We’d heard good things about the N82, and I like the form factor of it quite a bit, but for live streaming the N95 was far and away the better device. It seemed to get stronger 3G signals and had considerably better battery life. Both devices took quite a while to get set up, given that we only received them the day the festival started, and so we went into it cold. If we were to repeat the project with current technology we’d definitely push to get N95s for everyone and more time beforehand to set up, learn the tools, etc.

So, a few lessons for next time:

Have one team member whose primary role is logistics, not reporting. While it was possible for all of us to get some content, and one properly prepared person could get a lot of content, if there’s any chance of having someone who can manage liaison with the rest of the event, and other logistics, that’s ideal. That person may be able to do some reporting, but don’t count on it. If time allows, that person could also manage aggregation and promotion of your content.
Get to know the tools in advance. By the end of the festival we all knew our way around the phones, but there were settings we hadn’t had a chance to explore that could have affected quality. Getting off to a quick start is important for confidence, and promotion. I suspect that the sketchy quality of some of my early videos may have put some people off from watching the others, particularly viewers who didn’t know what we were doing.
Phone companies don’t make your lives easy. I bought a new Orange pay-as-you-go SIM for the festival and ended up spending 40 minutes on the phone getting it activated, and used over four pounds of credit on the call. Once that was done, their “7 days’ unlimited data for £5″ deal worked out well, but these things always take longer than you’d like.
Build in time to review the material you’re gathering. This is particularly important if you haven’t produced content in this way before, but is good practice either way. Reviewing the content is the only way to work out how to improve, and the review process is another chance to identify particularly successful videos and promote them. Steve did a good job of watching his and other videos and blogging about them as the event went on, which significantly increased his audience. In many festival programmes there will be lulls at certain points in the day and that can be a good chance to find a quiet spot with wifi.
Talk about what you’re doing. For now, the idea of live streaming to the web is novel and is a strong hook to get people engaged and start conversations. That led to some good interviews, to other people joining in and to ongoing conversations that have spilled outside of the festival.

(photo above is by Jon McKay, from his ‘So What Do You Think?‘ project)