The fine people at EllisLab have a quite interesting post on how they see feature requests. Generally speaking, I like their voice. Unlike many web 2.0 companies, they're not delivering the usual pandering bullshit: "it's all about you the users and the comm-you-nih-tee." Yeah right, like we can't figure out the part about UGC economies of scale in your pitch to VCs.

EllisLab is saying, look we do appreciate your feedback, but it's our job to maintain the product's sense of purpose, stay the course and deliver. It's refreshing, it's honest, and it's the right thing to do. You need the courage to say no to a lot of requests, whether externally or within a company with your internal users by the way. You also need to execute on some of those ideas, or people will just see the whole thing as an exercise in futility. Maintaining a transparent dialog about why feature requests make or don't make the cut is key. Easier said than done as sometimes you'd rather just sit down and ship product.

Salesforce is doing a great job with its IdeaExchange (here's their blog about it). There are ASPs (Oops, sorry for the 90's wording, surely it's more hip to say SaaS) such as BrightIdea too, and of course, a blog dedicated to idea management systems (what topic doesn't have its blog this days, carrot juice fetishism maybe?). If you can educate internal and external users about trade-offs and cost/benefit decision making, I think given enough scale (i.e. you need the manpower to handle the firehose) these systems have real potential.

I'm a fan of services neatly packaged as products that answer a very well defined need. A good example is psd2html: you know what you put in the box, and you know what you're going to get at the other end. Great way to complement your own resources. We're coming to a point where our quite informal internal QA is showing its limits, and we need to reinforce our development methodology to include proper regression tests and some level of automation. What should we use, JMeter? Zend? Empirix?(Yeah I know those are apples and oranges to some degree.) Application health monitoring (especially for transactions) is also something we need to investigate as low level monitors such as CPU or RAM consumption don't give the whole picture. Many of the tools in this space are too expensive, and more importantly, too cumbersome for our organization size and project scope. I need to research leads such as this QaTraq lite approach. Byrne Reese's Test Run seems targeted for people like us.

Now, tools is only part of the answer. Even though we're growing our in-house development resources, we're still a publisher at heart and at our scale, I can't justify a full-time QA person. What I'd like to find is an external resource we can use when we need it, with a very predictable output and cost. Poking around the web, there's uTest, but they're working off their own environment. I need someone to work within our Trac system so that we don't multiply systems and tools with overlapping purposes.

Please email any advice or suggestion as my comments are still broken (sigh). I really need to overhaul my whole blog.

Update: there doesn't seem to be much that integrates with Trac (Test Run and QaTraq don't), except for a beta plugin that's not very user friendly.

02/03/08 update: TestLink 1.7.3 was released last month with Trac integration. Now I need to look how well this integration is done. I love how Trac tickets and Subversion changesets can very easily be linked together (you'll have to scroll down to Vidar Hokstad's entry since LinkedIn hasn't implemented comment permalinks yet -- sheesh why does every site that rolls out community features have to make the same design mistakes again and again?) thanks to the common wiki syntax.

03/28/08 update: Oracle buying Web app test tools from Empirix (another successful software company coming from outside the silicon valley).

Why, now that all these topics that I posted feverishly about for years are making front page news, am I almost mute? Jeffrey Zeldman perfectly captures everything that got me sick of the whole web 2.0 noise explosion. I'm busy building our own businesses and I hardly even keep up with what's new. Oh wait, there's actually preciously few innovations around, everybody is busy trying to pass me-too features for original products, and bland products for actual businesses.

My interest for online things sold to Google (or whomever else) that not only did zero dollar of revenue so far, but obviously scream for a lack of any business model is absolutely nil. More power to you if you can make a quick buck in the meantime, if I can grab some silly money I wouldn't spit on it either. But let's drop the attitude and stop pretending there's anything actually exciting going on there.