Ethain (3rd November 2010)
I wouldn't suggest otherwise, but not sure parallel is the case, with all interacting in their own way with a single database, making the changes they want to make, which will have an effect of other systems also doing the same.
With a separate database, there is no requirement to have a single copy of it. Mirror it. No need to have the mirror in an identical RDBMS.
And if only some of the functionality goes down, at least you have the rest. Many things interfacing with an MIS API will only need occassional access (adding new users), so downtime of the database is no big deal.
To be honest, worrying about the technology now is a bit early. Before that is done a project needs to have done analysis and design work first, not to mention writing the first batch of test cases (so you know the system meets the requirements before you are coloured by what you have coded). Doing that might influence the technology in ways none of us has thought of yet.
Licence SIMS Software £5000
Installation on my computer £1000 (plus expenses)
SIMS SQL new software releases £150 per year
Partner support £3500 per year
Access core data + lesson / attendance data £150 per school per year and if I want all the bells and whistles that's £200
Not exactly free then, is it? I'm talking free, as in £0.00 per school per year
Small firms can be deterred from entering the market by these charges, thus stifling innovation. Those firms that do get into it simply recoup the cost by increasing the cost to the schools, making the add-ons more expensive than they need be.
If you work with the LA, the LA can request the API etc, for a grand total of £0, you can the deploy your work across the LA for a grand total of, £0.
If you want to then go outside of the LA, it will cost per school (Access core data + lesson / attendance data £150 per school per year or whatever).
We got by OK without relational before e.g. I first worked on a bigger, badder, distributed, interoperable MIS system 20 years ago that got by just fine without a proper RDB, and at a glance it fits onto the NoSQL/NoRel category, as apparently does one of IBM's from 1966. That system did eventually go RDB, but not for any pressing technical reasons i.e. the customer's consultants taught them to ask for it on tenders and you had to tick boxes. Naturally performance & resource usage went right through the floor so customers then had to pay for that tickbox by throwing much, much more significant hardware at it.noSQL is a long way off from being able to replace Postgres in the MIS environment
That's a very spurious argument - we got on without a lot of things.. doesn't mean they're not required now. I'd to see how a modern MIS would get by "just fine without a proper RDB" these days...
There are currently 1 users browsing this thread. (0 members and 1 guests)