How much work before the CDML experts take over?

[signal]

I'd like to develop databases for clients up to, but not past the point where a CDML guru could put the database on the web. How far should we go? I want to strike a balance between the cost savings and control of doing it in-house without wasting time.

If the clients were using Filemaker only, we could design the whole thing with lots of goodies but obviously, much of that is wasted when porting to the 'net.

Right now, we are using MySQL and PHP but I find myself frustrated with the slow dev times compared to when we work in Filemaker.

What is the furthest I could take the design? Obviously, the CDML person would have to be included from the beginning somewhat .

Feedback appreciated. BTW, still looking for that CDML person.

[ March 15, 2002, 05:13 AM: Message edited by: signal ]

[Steven H. Blackwell]

The CDML guru should be involved from the start. The architecture of the solution will be influenced by whether and how it is to b deployed for access by web browsers.

Old Advance Man

[nonzerocomm]

I agree with the fact that a CDML expert should be involved early if not form the very beginning.

I also want to point that the firm hiring a CDML expert has a few things to do first with or without the CDML expert. One of them is to asses their needs in plain english and get a second, third opinion before any code get written down.

When I send a copy of job/contractual offer to my database analyst friends (not CDML people at all), they can tell from the start if the offer is viable or if it will die in infancy.

Around here (Richmond, Qu

[signal]

That's why I wrote: "Obviously, the CDML person would have to be included from the beginning somewhat".

I guess I could have asked it this way: I'm not willing to pay someone to design a FM solution but I am willing to pay them to hook it to the internet. How much can we do and at what point should we turn it over completely?

I could make my PHP people learn CDML but we have enough projects that I would rather vend some things out. Developing in FM is a faster and more complete solution than MySql. But I know that some things (like complex scripts) that work well in FM don't port to the web.

[ March 15, 2002, 06:33 AM: Message edited by: signal ]

[Keith M. Davie]

signal, you must be encouraging my input when you write, "But I know that some things (like complex scripts) that work well in FM don't port to the web."

Au contraire mon ami.

To see complex scripts in action over the www, to challenge them against themselves or each other, just visit my website. That is what it is for. It is really quite a simple example. But the scripts are complex - on purpose so they run for a while to be challenged. Sure it will take you some time to read what to do and then do it. But you can make multiple requests. How to is all there. Or get a few of your people to join in and submit from different machines. It is very clear where the scripts are submitted. Even if you manage to get a near-simultaneous hit, you will not get misinformed.

The names you register can (and should) be fictional. You will get no cookies here. You will get no spam from interacting.

Of course, running scripts is not the answer for every solution. But the ability to run them safely exists.

http://www.simplifyfm.com:591/

[ March 15, 2002, 07:53 AM: Message edited by: Keith M. Davie ]

[signal]

I don't really KNOW that FM scripts don't work on the web, just read a thread this forum about it.

What I need is a competent CDML person to tell me what I can and can't do in FM before I pay them to implement the design.

We have lots of clients and work. Just want to see if I can put Filemaker into the mix. It would enable me to design both the database and user experience while travelling.

BTW, your email is over quota.

[nonzerocomm]

Signal

When you write: "What I need is a competent CDML person to tell me what I can and can't do in FM before I pay them to implement the design."... you are right on the spot.

The next step comes form the CDML expert who will ask you the who-what-where-when-why-how questions.

Alike IBM, any consultant will want to estimate with you the following four crucial first feasability parameters: Human, Financial, Technological and Time feasability.

Since the number of computer, operating system, browser, plug-ins, human language, user interface design (and else) combinations exeeds human common sense, you will end up both agreeing to put limits on these variables.

Then, in exchange of a conditional purchase order, the consultant will handout a conditional contractual agreement with the above-mentioned limits expressed in min/max etc.

Since IBM does not code CDML, you won't get the elaborated design specsheet for free but paying a very max of %10 of your project budget will give you a solid document that you can take to any respectable programmer (CDML or not) and get the job done in a pleaseant atmosphere, on time, on budget and online.

I would do both jobs if we started with the first part first.

[Keith M. Davie]

signal, if you tried to email me from the link on this forum, that info was old and in error. I have just updated my email address in my profile. Thanks for bringing that to my attention. I've had a disruptive year.

If you tried to use the link from my site, I don't know why I would not have received that email. If that is the link you used, please let me know. Thanks.