How long to build a database?

[Sketchin]

I was just curious how long it would take an experienced programmer to create a database that included Contacts, Inventory, Bills of Materials, Reservation system, and Invoicing functionality.

I realize that this is extremely hard to accurately answer, but I'm not really looking for accurate answers! Just ball park me. 2 months? 1 yr?

I am also assuming that the database is pretty with all sorts of nice reports and what not.

Cheers!

 

[tehNellie]

How long is a piece of string?
Depends on the requirements, how much code I've already written that can be reused, how many times the customer changes their mind and so on.

I've spent 6 months working on a template reporting application for the place I'm at currently. Now that's written I can set a new database up in about 4 hours, maybe 2 days if it's complicated.

From my perspective the database itself is rarely the "complicated" bit, it's the application that sits on top of it that takes the time to sort out.

 

[spikepl]

The first question to ask is what is so specific in your requirements that it precludes using a proven commercial off-the-shelf system.

 

[Sketchin]

The first question to ask is what is so specific in your requirements that it precludes using a proven commercial off-the-shelf system.

This is exactly what I said to my employer, who wants a rather complicated database built using my zero Access experience. I think I have done well so far writing it from scratch in the last couple months, but am really getting hassled to finish it from the bosses.

Just wondering how on/off track I was!

Thanks for the input

 

[gemma-the-husky]

how big is the company, and how complex are your requirements

for a big system , fully bespoke you will easily get up to £50K/£100K plus for that, over time. You are describing the sorts of systems that ALWAYS keep growing. Larger firms will have in-house developers for this sort of thing.

it all depends exactly where you are prepared to compromise, as you may get systems with much of the functionality off the shelf. It depends how important it is to get the EXACT functionality in some key areas

One other thing is the number of systems that are out there. Excel say has millions of
users (probably hundreds of millions). if every user pays £1 towards development, MS have milllions available for the exercise

so - if you have a relatively small vertical market, then you may find the vendor has only sold 50 to 100 packages. if each of you are paying 20K for the package, the developer still has only £1m or so to cover his development costs. So not only do you end up paying more, the software also is rarely as slick as a packaged app from PC world, etc.

The upside is, it does exactly what you want. (or should do). You still have to source the right package/developer

Out of interest, inventory is probably the hardest app, depending on what you need to do.

 

[Lightwave]

Sketchin that's a lot of different things. Someone that is good who doesn't have the baseline changed on them every 2 weeks could do that in 6 months if and only if they were probably defining it themselves. They would have to be quite motivated though. Writing a good database is similar work to writing a good book (I think - I've never written a book but my impression is that like writing a book writing a database at some point, no matter how literate you are at writing, is just hard work). Not only that but like writing a book even if you are technically very competent somehow sometimes it just doesn't work.

Like someone said above it really takes a long time to make the user experience smooth and to make forms flow nicely. I quite often get prototypes up and running quickly or relatively quickly and spend the next bit of time making it really flow nicely and look pretty. I spent 3 hours this afternoon making an icon. I'm not happy with it yet.

 

[Sketchin]

Well, I have made it pretty far and am proud of what I've done to this point. Its certainley not as easy as the higher ups seem to think it is... "Can't you just make it do this and this?" "You SHOULD be able to do that"

Its painful listening to it sometimes, as they really have no clue how much can go into this. Especially when you have zero years experience!

 

[Lightwave]

Welcome to the world of software development.

 

[Lightwave]

I still think that management still haven't quite cottoned on to the fact that databases and systems in organisations are like the central nervous systems of organisations. Done well you will be flexible and able to use less staff. Done badly and they are worse than useless.

And sometimes the difference between a good and a bad system is remarkably small.

I think it is why software houses are always getting berated for new versions. The developers are told to iterate irrespective of whether the developers think there is a need to not. Management says you will iterate.

Suddenly the software guy goes. Oh shit I put all my good ideas in the first version.

 

[Simon_MT]

I think the best way to approach this issue is a two stage approach.

What is imperative and what would be nice. The advantage you have is this a greenfield development so you have considerable control over the development of this project.

Ignorance is sometimes a blessing. You have to project manage, do the business analysis , the design and then the implementation is at times, it must seem, challenging circumstances. This in itself creates issues but it a full life cycle development and far more rewarding that a piecemeal project.

So initially concentate on what is critical and bear in mind what else is being mooted. Provide for enhancements even if they eare not going to implemented in the first stage. Believe it or not it will be an enriching experience.

So far are being pretty that is the last thing to deal with. There can be an inverse relationship between asthetics and functionality. In other words gloss can disguise floss, functionality first. The attached screen is just the veneer and does not reflect then functionality incorporated in this solution.

Simon

 

[gemma-the-husky]

Well, I have made it pretty far and am proud of what I've done to this point. Its certainley not as easy as the higher ups seem to think it is... "Can't you just make it do this and this?" "You SHOULD be able to do that"

Its painful listening to it sometimes, as they really have no clue how much can go into this. Especially when you have zero years experience!

you never know. some things that you think will be hard turn out to be easy, some things that ought to be easy can turn out to be hard.#

the data design is vital. if the normalisation is not good, then it can become difficuut to extend the database, and to use it effectively.

and then, as well as "basic" functionality, you start getting more difficult requests, like being able to limit some users ability to do certain things. Which is nothing to do with the database, and is very hard to get right.

and then you also have to be concerned about robustness, and resilience, and your company's network configuration, and what platform to host the back-end on ....., which again is not part of the basic functionality.


out of interest, how long HAVE you spent on it, and how far along do you think you are? How much do you think it will all cost?