two seven

The Challenge

I’ve got this friend who is responsible for a corporate front-end web environment with almost a dozen browser-facing web applications. Pretty much all he can do at the moment is change static content in the antiquated and wildly inappropriate CMS. Changing anything significant requires implementation of an expensive one-off SDLC waterfall-type project with a business case, requirements-gathering, PM, BA, dedicated test-resource – the whole box and dice. These projects often under-deliver, with scope being reduced en route to avoid budget and deadline blowout.

He’s not happy with the way things are.

He’s got it into his head that it would be far better to create an easily-manageable front-end, with a unified, standardised UI under the control of his front-end web team. Key aspects are simplicity, speed, cost-efficiency, and trust – none of which can be used to describe the current state of affairs.

Some ideas

I was talking to my friend, and he said that at a high level, he’d like to abstract the various applications from the UI, where possible, by means of API/Web Services/etc. On the front-end would be a web application framework – He’s thinking Symfony or similar. He believes he has sufficient developer resource on his team to build/maintain/support this.

Some other ideas he’s been tossing around, in no particular order:

• Source repository. (He doesn’t have one at the moment). svn or git? Stable trunk policy?
• Continuous integration. (thanks Mike!)
• Test-driven development.
• Automated processes.
• Security.
• Content management.
• Performance. Code-efficiency, caching, etc. (Although he’s heard it said that performance shouldn’t become an issue until performance becomes an issue)
• The database. Does CouchDB lend itself to supporting a content-driven web application?
• Will it blend scale?

More ideas…

… are welcome. He needs all the help he can get. While it’s all very bluesky (with pie) at the moment, he needs to turn it into a watertight, bulletproof, business case. And soon.