Should I buy my CTO's idea of upgrading to Python 3?

Answered by,
   Jamirul molla,B.Tech

I am going to agree with some others that starting a new project in the last year on Python 3 in 80% of cases was not a good idea.  There are exceptions, such as when you are distributing your code to customer "clients" and therefore want to support the default Python version of the target Linux distros (2.6x for Centos/RHEL, 2.7x for Debian/Ubuntu, 3.XX for Ubuntu 16.04+ (possibly)).  If you are just doing a REST API on your own servers, then not using 3 was most likely a mistake.

To that end, the changes in 3 can provide productivity gains.  I won't get into specifics, but I recently ran into a problem involving ASCII/Raw string escaping in 2.7x that lead me down a rabbit hole trying to fix.  This would not have been a problem at all in 3.

So to agree somewhat with your CTO's plan to move to 3, sooner is definitely better than later and your technical debt is just going to grow the longer you are on 2.

Some others have pointed out that after taking investment, you should focus on revenue...and that is true.  So I guess I will just say that if your situation allows you to upgrade to 3, then you absolutely should.  If you absolutely can't...well you should start writing Python 3 complaint 2.7X code from here out to at least minimize the technical debt being accrued.