Introducing Python

Question

The company I used to work with has two developers working fulltime, and a handful of freelancers. They're in the process of hiring a new lead developer to try to bring order and management to the development.

But, currently, one of the developers has seen the light of Django (the company has only developed in PHP to date) while the other developer is concerned that introducing a new language (Python) is a bad idea right now.

How should they approach introducing this new technology? Obviously with only one of the developers actually knowing Python, there will be no redundancy when that dev is away or leaves the company.

Should they bother to introduce Python, or should they look for PHP-only solutions until such a time when the team actually have more than one Pythonion? Without a team leader, the decisions are having to fall to them.

Answer 1

Well, python is a high level language.. its not hard to learn and if the guys already have programming knowledge it should be much easier to learn.. i like django.. i think it should be a nice try to use django ..

Answer 2

If the mandate of the new lead is to put the house in order, the current situation should likely be simplified as much as possible prior. If I had to bring things to order, I wouldn't want to have to manage an ongoing language conversion project on top of everything else, or at least I'd like some choice when initiating the project. When making your recommendation, did you think about the additional managerial complexity that coming into the middle of a conversion would entail?

Answer 3

I don't think it's a matter of a programming language as such.

What is the proficiency level of PHP in the team you're talking about? Are they doing spaghetti code or using some structured framework like Zend? If this is the first case then I absolutely understand the guy's interest in Python and Django. It this is the latter, it's just a hype.

Answer 4

@darkdog:

Using a new language in production code is about more than easy syntax and high-level capability. You want to be familiar with core APIs and feel like you can fix something through logic instead of having to comb through the documentation.

I'm not saying transitioning to Python would be a bad idea for this company, but I'm with John--keep things simple during the transition. The new lead will appreciate having a say in such decisions.

If you'd really, really, really like to introduce Python, consider writing some extensions or utilities in straight-up Python or in the framework. You won't be upsetting your core initiatives, so it will be a low/no-risk opportunity to prove the merits of a switch.

Answer 5

I think the language itself is not an issue here, as python is really nice high level language with good and easy to find, thorough documentation.

From what I've seen, the Django framework is also a great tooklit for web development, giving much the same developer performance boost Rails is touted to give.

The real issue is at the maintenance and management level.

How will this move fragment the maintenance between PHP and Python code. Is there a need to migrate existing code from one platform to another? What problems will adopting Python and Django solve that you have in your current development workflow and frameworks, etc.

Answer 6

I recently introduced Python to my company, which does consulting work for the Post Office. I did this by waiting until there was a project for which I would be the only programmer, then getting permission to do this new project in Python. I then did another small project in Python with similarly impressive results. In addition, I used Python for all of my small throwaway assignments ("can you parse the stats in these files into a CSV file organized by date and site?", etc) and had a quick turnaround time on all of them.

I also evangelized Python a bit; I went out of my way to NOT be obnoxious about it, but I'd occasionally describe why I liked it so much, talked about the personal projects I use it for in my free time and why it's awesome for me, etc.

Eventually we started another project and I convinced everyone to use Python for it. I took care to point everyone to a lot of documentation, including the specific webpages relating to what they were working on, and every time they had a question, I'd explain how to do things properly by explaining the Pythonic approach to things, etc.

This has worked really well. However, this might be somewhat different than what you're describing. In my case I started with moderately small projects and Python is only being used for new projects. Also, none of my co-workers were really Perl or PHP gurus; they all knew those languages and had been using them for awhile, but it didn't take much effort for them to become more productive in Python than they'd been before.

So if you're talking about new projects with people who currently use PHP but aren't super-experts and don't love that language, then I think switching to Python is a no-brainer. However, if you're talking about working with a large existing PHP code base with a lot of very experienced PHP programmers who are happy with their current setup, then switching languages is probably not a good idea. You're probably somewhere in between, so you'll have to weigh the tradeoffs; hopefully my answer will help you do that.

Answer 7

I love Python and Django, and use both to develop the our core webapps.

That said, it's hard to make a business case for switching at this point. Specifically:

• Any new platform is risky compared to staying with the tried and true
• You'll have the developer fragmentation you mentioned
• It's far easier to find PHP programmers than python programmers

Moreover, as other posters have mention, if the issue is more with spaghetti code than PHP itself, there are plenty of nice PHP frameworks that could be used to refactor the code.

That said, if this developer is excited about python, stopping them outright is probably demoralizing. My suggestion would be to encourage them to develop in python, but not the mission critical parts of the app. Instead they could write some utility scripts, some small internal application that needs doing, etc.

In conclusion: I don't recommend switching from PHP, but I do recommend accommodating the developer's interest in some way at work.

Answer 8

It's really all about schedules. To me the break should be with a specific project. If you decide your direction is Django then start new projects with that. Before you start a new project with a new language/framework, either make sure that you have scheduled time to get up to speed in this new direction, or get up to speed before using on new projects.

I would avoid going with a tool of the month. Make sure you want it to be your direction and commit some time/resources to learning enough to make a good decision.