Advice on what language to use... n00b question

[Spinach]

I'm looking to design a website and I need some advice on what language would be best to use for server/client interaction.

The site is basically going to involve two sections; one where someone says "I have free space to carry x tons of stuff from A to B" and another where someone says "I have x tons of stuff to carry from C to D"

The idea is to match the two up ie match packages to free space.

Obviously I need to have some kind of database runnning on the server to allow someone to add/query packages and free spaces.

To further complicate matters, I'll need to charge people for certain sections of the site and also have secure log in i.e. https...

I also suppose that recommendations for good servers would be nice too!

I know its a rather nebulous question but any help at all, on any section of what I'm writing, would be helpful.

EDIT:

Also postcode interaction would be helpful, e.g. how close is postcode A (where the package is) to postcode C (where the free space is leaving from). I guess you have to pay for the postcode map/distances database... again any thoughts appreciated!

[directhex]

never EVER use anything to inetract with the user other than raw HTML. really.

as for the back-end, pick a database & language. postgresql, mysql, firebird sql, ms sql server; asp, jsp, perl, php. it doesn't really matter, as long as the user never gets fed anything more complex than html.

you might want to try JSP, i like its simple user tracking facilities, for login etc

[Iain]

The PAF (Postal Address File) does have a cost, if memory serves from a previous project it's based on how many users will be served by the postcode lookup functionality. That alone won't allow you to calculate distances, that will just let you work out an address based on a postcode. What you're talking about is a complex application just for calculating distances, never mind all the other bits. I'd have a look at FedEx, UPS etc and see if you can identify what languages they are using. If it works for them then it's going to work for a new entrant into the market.

You'll need to buy certificates for use in SSL as well. If it's a large scale application then you need to be aware that SSL is a performance dog and you may need to buy a dedicated SSL Accelerator card for your server (~£1200)

Sounds like you need to do a lot more investigation as the spec at the moment is pretty undefined. Most of your answers will come down to cost though, if you don't have any/much money then use PHP and MySQL, if you've got money to burn then consider JSP or ColdFusion, I'd recommend JSP running on a Sun server with some sort of application server software, BEA maybe

[dozer]

php and mysql, free and easy options with lots of support available, codestealers welcome

[killgORE]

Woa.. thats sounds pretty large and interesting.

I am pretty much in the directhex camp... Clients should not need any flash/applets etc to download.. a little bit of form handling in JavaScript maybe.. to check that key fields are not null before wasting the servers time etc. BUT good clean HTML to the clients should ensure it works for most web browsers and therefore to more potential clients for the pay areas of the site.

Backend systems ? MySql is pretty much teh s*it.. but then I haven't used many alternatives and certainly not for very long.

If the data is modelled correctly into a database. then the rest of the system writes itself. MAKE sure that ALL data necessary <- not possibly useful in the future data -> is included in a formal structure. Agree on a naming scheme etc and THEN once the core of your site is designed. At its heart it is all about the data, never forget this. Come up with the means to access this data.

So far you have made no choices about language etc. Since the data is modelled. U should have a better idea of user numbers and how the users would interact with the database before leaping onto a specific language. This is good. if it turns out the language that is selected is wrong for the task.. or in the future some UBER language evolves.

The design and database should be transferable.. which is kinda a key goal.

Just some food for thought. GOOD LUCK with it.