Don't get me wrong, I'm not an advocate of death by a thousand specs. But I've found the most important element in creating software is design. That design may be all in your head ( mine often is ) but it had better be very big picture and very comprehensive and have inevitable change built in. The worst thing I've seen developers do is get locked into a particular approach to part of a solution mentally before the big picture has come into focus and then not be able to paint themselves out of the corner so it leads to some very bad decisions with a lot of unintended consequences.

I'm involved in a project right now that involves more people on both the business and tech side, more systems and more real-time moving parts than I'd ever imagined before and I'm seeing the parts that are working are the result of good design and the parts that are gonna get some people fired are failing because of being in such a hurry to look productive that nobody took the very short amount of time it would have taken to do some design.

I'm lucky that my team have all been on the same page ( my page ) about this and before we started investing a lot of time in writing code we knew the scope of what the code would be asked to do and what the likely change points were going to be and built something that had a very decoupled design and a lot of flexibility for the changes we knew would come fast a furious once lots of stuff came on line at once. Just the changes I made today as seven real time feeds were coming in would have been enough to sink me if we hadn't done those couple of day of really figuring out how we would handle new and changed requirements and how we would recover from very bad things happening whether they were our fault or that of some other team or some unforeseen glitch.

I'm sometimes amazed at how much software and how many business processes - even in places where you'd think they'd know better - are really badly thought out before they go into action.


>There is another wise phrase about "the paralysis of analysis." Probably the hottest trend in software development the past few years has been agile development, sometimes called iterative development. The objective is to release a system in small, fast chunks. I am not wise enough to know whether that's really the answer but I do know Agile was a reaction to development teams disappearing for months or years before users had anything they could run.
>
>>Bet when considering the value of analysis always remember the words of the wise old rabbi ( Menachim Bazian ) "The quicker you get to the keyboard the longer the app will take."
>>
>>>I could possibly analyze the performance of my application on the 3 current customers (since I have remote access to the servers). But the reality of my work/life is that I never have time to stop and think or analyze; always next feature to write or a bug to fix or a sale to make. I think I remember the joking message that goes something like "I am so behind in everything; I am afraid I will never die"
>>>Thank you.
>>>
>>>>But also remember the design of most VFP/Sql apps I've seen (or written) demand very little of server other than basic CRUD. 1 gig of RAM for that kind of backend processing is not a bottleneck unless you have already exceed the recommended number of *simultaneous* users.
>>>>
>>>>>And I understand it is limited to 10 GB size database and certain number of simultaneous users (don't know the number though).
>>>>>
>>>>>>Express is limited to a single CPU and 1 Gig RAM.
>>>>>>
>>>>>>>Thank you very much for all your input. Very helpful. I do have currently 3 customers using my VFP with SQL Server. Two of them using SQL Server 2008, I know it is not Express but not sure what exactly. The 3rd one is using SQL Server 2005 Express. Not one of the current customers complained about the speed. But it was my perception that Express will have some bottle necks (I am probably mistaken). The key, of course, is how many users they will have. I sent them an email and hopefully this will determine if Express is an option. I don't mind saving them money if they don't suffer performance.
>>>>>>>