How do you start programming big projects?

You need to provide a little bit more information about what you're building if you want to know what would be best, what are you building?

As for frameworks there are so many of them nowadays. My recommendation, if you want people to test it out and it's for sale, is to try a company called RunAware. They let you test drive software without losing control of it.

 

For big projects I get myself a piece of paper and draw how will my site look like, what pages will it have, what functions I need and how will they work, etc. I also plan the order in which the parts of the site should be coded. Then I get to work I do the basic layout first, then begin to code the whole thing. I do the easiest functions first, and the complicated ones later. The hardes part for me is the testing process, I hate it when I realize something doesn't work as intended... :sleep:

 

If you immediately delve into the programming phase without being concrete about what you're trying to program, you're going to run into so many pitfalls, you will probably end up quitting out of frustration. What you need to do is plan out what pages will comprise your site, and what functionality will be found on each page. This way you know right off the bat how many tables you need to create, and you can find out which databases and such will best suit your needs. Then, you need to map out every relationship each page has to one another. When it comes to object oriented programming, this step is -crucial- as it essentially maps out the framework of your code before you even get started. From here, you can fine tune with other organizational strategies or go right into it starting with the index.php and branching out from there.

 

Talking about how to start encoding a large project? Here is how I have been going about it for last 6 years. Proper designing of the project layout is what comes to mind the first time you mention of programming. I first take time to peruse through the project basics, functions and catalogs. Taking a paper, I make drawings of my potential site showing its pages and its functions and how I need such sites to work. After this I compare my plan from other types of such phenomena I attempted in the past. Personally, I admit completing the simplest parts of my project first is fundamental since it gives me the potentiality to keep becoming more accurate in the hard ones.