It's easy for people to talk of ideal methodologies and doing things right with greenfields projects; we can easily forget that most developers will be supporting a lot of legacy code that still 'works' and keeps the client happy. Very interesting talk on the various approaches available to help build confidence in ongoing support of a legacy codebase.
Well paced talk about the practicalities and the difficulties of developing a testing framework for legacy code. Enjoyed the discussion about different approaches and ideas to the problem and the pragmatic sentiments throughout.
Loved this. It's easy to present ideals, cool new technology, and perfect code in a conference talk. What's much more valuable is to have a speaker (and even more, the company) prepared to open up and admit that yes, people write crappy code, it ends up in critical systems, and no, you can't just rewrite it all. I think the slides could be a bit improved though, the code examples were a little hard to follow in places.
Comments
Comments are closed.
Good set of tips on how to approach testing legacy code with some real world anecdotes of how this has worked with the SugarCRM project.
It's easy for people to talk of ideal methodologies and doing things right with greenfields projects; we can easily forget that most developers will be supporting a lot of legacy code that still 'works' and keeps the client happy. Very interesting talk on the various approaches available to help build confidence in ongoing support of a legacy codebase.
I loved this talk, gave me a good set of tips for testing the current legacy code that I'm working on.
Probably my lack of enthusiasm for testing, but I didn't really enjoy this topic.
Well paced talk about the practicalities and the difficulties of developing a testing framework for legacy code. Enjoyed the discussion about different approaches and ideas to the problem and the pragmatic sentiments throughout.
Well presented and the content was good.
Loved this. It's easy to present ideals, cool new technology, and perfect code in a conference talk. What's much more valuable is to have a speaker (and even more, the company) prepared to open up and admit that yes, people write crappy code, it ends up in critical systems, and no, you can't just rewrite it all. I think the slides could be a bit improved though, the code examples were a little hard to follow in places.