Would you start a new project in VFP

[GriffMG]

Hmm

Just put in a proposal for a rewrite of an app I started in the year oatcake and have been enhancing and maintaining ever since. It was originally in Clipper Summer 87 I think, might have been 5.0, and migrated to xBase++ a few years ago.

Now I get to do it over, something I have avoided for years, and I seem to have a free hand (at the moment) the brief just says 'a fully windows environment'.

I'm sorely tempted to try and do it in VFP9 - I won't be able to use much of the existing gazillion lines of code, but I will be quick and comfortable.

But, would I be stitching my client up?

Regards

Griff
Keep [Smile]ing

There are 10 kinds of people in the world, those who understand binary and those who don't.

 

[MrDataGuy]

The short answer is Yes.

Now for the long answer:
-I suspect that you will get a range of answers from my Yes, to any programmer that starts a new application in VFP should be taken out, shot drawn and quartered, burned and his ashes put into a new concrete driveway.

It really depends on the application, the work environment, the developer etc. For example if the app is data centric, not so large that a database like MS-SQL is needed, no ( or limited ) interface to other applications, the developer and maintainer is well familiar with VFP, it will be a 1 to ~20 person application used in house, then I would very much consider using VFP. But I would also consider the pros and cons of doing it say in .Net with SQL-Express, or PHP and MySQL as a web app.
What happens if you somehow are out of the picture? Who will maintain it?

I do/use VFP, .Net and a number of other application languages and am right now rewriting a 12+ year old Fox DOS app into VFP 9.0 because after chatting with the client we decided it was the best thing to do in their situation.

I know that there are some developers who have threaten to take away my computer for doing so, I think their words were something like “You should be sued for professional malpractice” and I should mention that this was coming from a person who has in the past received MVP in VFP, e.g. a heavy hitter. But and the end of the day VFP is what I am using to solve a particular client need.

So yes you can use VFP to start a new application, but think about it first. If you are not a good VFP developer maybe it would be best to spend your time and energy learning/using a new language (or contact me and have me write it <smile> )



Lion Crest Software Services
Anthony L. Testi
President

http://www.LionCrestSoftwareServices.com

 

[GriffMG]

Hi Anthony,

My thoughts are not a million iles from yours, otherwise I would not have posed the question.

Am I a 'good VFP developer'? good question.

'Who will maintain it' is also good, and I have a cunning plan for that - if I am 'unavailable' then one of our other MVPs will probably inherit it - along with one or two other things... if he wants it (no pressure - I don't care, apparently I'm dead).

I'm thinking about it, as you suggest, and I am asking the community - and I'm not sure there is a better option, one with a more secure future?

I'm willing to learn something new, love learning new things, but is .net any more secure going forward?


Regards

Griff
Keep [Smile]ing

There are 10 kinds of people in the world, those who understand binary and those who don't.

 

[MrDataGuy]

.net any more secure going forward?"

Microsoft can drop it tomorrow, but I doubt it. No one has a cystal ball but my gut tells me that .Net will be supported until at least 2020.

I also think VFP will run on MS OSs until at least 2020, if only via Virtual PC. So if it takes 6 months to write it, the app is good for 8+ years.

What ever you write it in, write it using good OOD, OOP, and with good interface layers, that way you can switch/upgrade bits and pieces as needed.

A complete rewrite does give you a good excuse to try something else.

Lion Crest Software Services
Anthony L. Testi
President

http://www.LionCrestSoftwareServices.com

 

[MrDataGuy]

Looking over your user profile ( I should have done that first ) you are in the good VFPer category, so most of my comments may be best read by 3rd parties.

Maybe a question is can your ego handle other programmers laughing at you for doing a new VFP app? I say that because I have had a few laughts, and slings and arrows shot at me for doing so, but for some situtions I think VFP is still the best tool to use. I have some very happy customers that think so.

( I know I know some of you are thinking, the customers would be even happier if I was developing in X, Y or Z development language.)

Lion Crest Software Services
Anthony L. Testi
President

http://www.LionCrestSoftwareServices.com

 

[GriffMG]

8+ years would be fair to both parties.

The app being rewritten only needed to last a couple of years, and it has seen one person to retirement, and kept the other in a job for a decade - not including me, who has been looked after very well.

A rewite in a new language is ok, that's why I'm asking B-) I can't see another project on the horizon that would pay for that!

Regards

Griff
Keep [Smile]ing

There are 10 kinds of people in the world, those who understand binary and those who don't.

 

[GriffMG]

People can laugh all they like.

Been xBaseing for quarter of a century, VFPing for a long long time, none of the doomsday scenarios have proven true yet, but we are near the end of life for this tech' like it or not and touching base with the community is a good thing.

What exactly is OOP? B-)

Regards

Griff
Keep [Smile]ing

There are 10 kinds of people in the world, those who understand binary and those who don't.

 

[MrDataGuy]

A rewite in a new language is ok, that's why I'm asking I can't see another project on the horizon that would pay for that!"

IMO the first app in a new language is going to be a PAIN and will take much longer then doing it in VFP, so make sure that it is a goof investment for all. Each app in the new language will be better, but still IMO it will be a pain for awhile. Sometimes thou one just has to handle the pain.

Lion Crest Software Services
Anthony L. Testi
President

http://www.LionCrestSoftwareServices.com

 

[MrDataGuy]

OOP = Object Oriented Programming

OOD = Object Oriented Design

OOA = Object Oriented Analysis

Massively simplifying things: Classes, Objects, Polymorphism, inheritance, sub-classes, Methods, information hiding etc. etc.






Lion Crest Software Services
Anthony L. Testi
President

http://www.LionCrestSoftwareServices.com

 

[GriffMG]

You are SO right.

BTW thanks for putting me in the Good Guy column.

At the moment I am thinking VFP is probably right, the customer doesn't even have one Win 7 or Vista PC (officially), but that could change at the drop of a cheque book. Mind you they have about 4,000 PCs in the UK alone, so it's quite a big cheque book to fumble and drop.

Regards

Griff
Keep [Smile]ing

There are 10 kinds of people in the world, those who understand binary and those who don't.

 

[SBTBILL]

There are still a number of things that VFP does better then most other things. I've yet to see another language the handles relations as well as Fox when your auditing. Access seems to come close. SQL can be set up to look like it with lots of work.

So much is going to the internet now. If it is a stand alone app I'd certainly consider VFP. If it runs as an EXE I'd consider VFP. If it runs on the Web no way, I'd go with VB, which is also rumored to be on the way out. But VB looks like VBA.

 

[JRB-Bldr]

I'd suggest looking at it first from a non-developer perspective.

If it is your own application that you have SOLD to someone, then do whatever you think best - I'd personally recommend using VFP9.

Or is this an application developed for a client/customer?
If it is a customer's application, then how will they look at it? And/or will they care at all?

Will they feel that the work needs to be done in .NET so that they will be more able to find other developers down-the-road (just in case you get run over by a truck!) to support the application?

That may dictate what you use for your development environment.

And will the customer be expecting internet functionality?
While there are many ways that VFP can support this, some of it is easier with .NET tools.

Lastly, do they have a fixed budget and/or timeline?

Obviously developing an entire application in a totally different language will take a good bit longer and just changing some parts of FP/VFP code into the newest version. That longer development time will mandate a larger budget and timeline.

In summary I'd recommend approaching the decision from a business perspective first and then, after that, from a developer's perspective.

Good Luck,
JRB-Bldr

 

[GriffMG]

For what it is worth, the first pass of the proposal through the client's IT department has indicated that VFP would not be the preferred option (i.e. No) and that MS SQL should be the back-end...

We'll see how sure they are about that!

Regards

Griff
Keep [Smile]ing

There are 10 kinds of people in the world, those who understand binary and those who don't.

 

[JRB-Bldr]

MS SQL should be the back-end

OK, but MS SQL cannot, by itself, create 'front-end' applications.

So you could still have a VFP9 application that merely uses MS SQL Server as its back-end.

What I am saying is that, while the back-end is VERY important, it is quite often not the Development Language.
The way in which you are conveying the client's 'IT Dept' decision suggests that they might not be thinking fully through the whole project and its development.

Therefore, their choice has not yet defined if you will be allowed to start a new project in VFP or not.

Good Luck,
JRB-Bldr

 

[GriffMG]

I understand that JRB-Bldr.

In fact I could probably live with the back end being shifted to MS SQL, but I have priced the proposal based on VFP front and back... so I reckon it will double the cost and time if I have to go a different route.

On the other hand if they have to use someone else, they will have to double that double... as there is no written spec of any kind - just their existing system to measure against - and any other third party would need one, and would not have the benefit of nearly two decades of maintaining and developing it.

Regards

Griff
Keep [Smile]ing

There are 10 kinds of people in the world, those who understand binary and those who don't.

 

[JRB-Bldr]

If the client's IT Dept is dictating the development, then THEY need to be identified as the ones who are changing your Development costs.

So if THEY want to separately purchase MS SQL Server license(s) and have it ready for you to use, then fine, that cost is on THEIR dept and not on your Development.

One thought would be for you to do your development with a SQL Express (free) back-end and then when/if their needs grow beyond its limitations, THEY could upgrade and pay you again at that time to make any changes.

But as I indicated before, nothing they have defined so far has yet DEFINED the project's development language/environment - just its back-end.

Good Luck,
JRB-Bldr

 

[GriffMG]

No, they have now been quite explicit about the front end they would like; Visual Studi0 2010 VB.

I have no idea, yet, which way they will fall once they work out the costs.

A copy of SQL server is neither here nor there to them, or a copy of the Visual Studio 2010.

Regards

Griff
Keep [Smile]ing

There are 10 kinds of people in the world, those who understand binary and those who don't.

 

[JRB-Bldr]

they have now been quite explicit about the front end they would like; Visual Studi0 2010 VB.

OK, that wasn't mentioned before.

It pretty much explicitly answers the question: Would you start a new project in VFP

Well if you are up to doing the work for them in VB.Net, fine - just modify your development costs and justify it by the fact that THEY have changed the project requirements and those specific requirements were not in your original proposal.

Good Luck,
JRB-Bldr

 

[GriffMG]

I think it actually answers the question 'would they develop in VFP?' which is slightly different.

I am HOPING they chose cost over environment, but I personally would chose to shift it to an Android based app and to heck with the cost.

That probably is not going to happen. They are not going to pay me enough to move it to VFP-SQL never mind VB-SQL (never mind Android).

I should find a tool to count the lines of code...

Regards

Griff
Keep [Smile]ing

There are 10 kinds of people in the world, those who understand binary and those who don't.

 

[GriffMG]

Meeting today.

Client IT not keen.

There are cost consequences, but there is also some risk.

If it works out to VFP->.dbfs that is good
If it works out to VFP->SQL that is good (not easy)
If it works out to .Net->SQL that is good (hard work)

If project is binned, that is a shame

Regards

Griff
Keep [Smile]ing

There are 10 kinds of people in the world, those who understand binary and those who don't.