I just learned the difference between legacy code and LEGACY CODE. #200wordsTIL
Buckle up this is gruesome
Most startups define legacy as "that stuff I don't like working with anymore"
We define legacy as "the person who wrote it is literally dead"
To me, legacy is code I wrote a while ago and now hate. Or have a hard time updating.
For example: We just updated Sidekiq from some old version to the latest and greatest. Worked swell in testing and QA, deployed to production, everything died.
Like our Ruby and Sidekiq code is so old that a gem we use became unmaintained and deleted their documentation 2 years ago. Just to discourage people from using it.
Now we're both using it and can't read the docs. Thanks ...
Both @mike_julian and Law of http://fuzzy-logic.org (dont know his twitter) laughed at me. Laughed at me hard.
Some examples _they_ shared ?
racks upon racks of 12+ year old servers, a couple AS400s in the corner, and "that one app that runs on PHP, but not any framework-based one, and oh by-the-by it's PHP3"
What's an AS400s?
a client that runs Rails 2.1 + Ruby 1.8.7, and all on baremetal
they've written a whole bunch of code that re-implements standard features of later Rails
And then @wesgarrison finished us all off with
Here’s where I jump in with a :kai-story: about the pre-rails-1 reporting app I just retired last year, running on Windows.
So yeah don't get in a pissing contest about legacy code. You will lose.