I would point out that a framework improves throughput FOR THOSE WHO INVEST the time it takes to use a framework well! You will see ROI drop through the floor as your developers scramble to learn ZF - it is VAST.
This is not a long way of saying "And so you should not use a framework". I am just saying that if a developer is not a framework advocate and not a framework centric developer, forcing a solution on him will not immediately make him more nimble.
The most immediate thing you will notice is that the developers will pour a huge amount of time into redoing what you have already got product wise and learning their way around the framework.
The payoff will come in a few months when they have the hang of it and can more rapidly turn around features and as you staff up, when the new developers (Hired for their framework experience) take less time to ramp up and are more effective. It also makes it easier to contract out specific self-contained tasks to outside agents.
Also, as a devil's advocate, I would point out that a small team, by definition, is slower than a larger team. Switching toolsets will help long term, and yes - at the expense of performance - but at some point your expectations have to be scaled to your manpower, and that includes being very cautious about which features your audience really needs.
On Sun, May 8, 2011 at 11:57 AM, Akram Benmbarek
<[address removed]> wrote:
Thanks Alex for your prompt response. The issue is not what framework to use. We explored Zend, CakePHP and Symphony. The reall issue is whether to use a framework at all. I have resistance from a team member to that effect under the excuse that framework based apps do not scale well... At the same time, we are faced with the reality that without a framework and with a small team, you can't get things fast enough.