Subject | Re: AW: [firebird-support] Embedded Firebird Libary using Firebird-Server |
---|---|
Author | Artur Anjos |
Post date | 2004-05-13T23:20:20Z |
Hi Rafael,
What I use is embedded for mono-users and demos, and if they want to try
a server installation, I give them instructions to do that. This use to
work very well, and the end-user will have from the start a small notion
of client/server.
I'm arguing against automatic removing the server. I'm arguing against
having registry keys to identify Firebird applications.
With embedded it's easy to create demo versions. That's the ones that
are unninstalled more often.
Client/Server applications need some care. Always. Someplace where there
is a server (any server - mail server, file server, sql server...) you
will need someone to take care of it. There is nothing you can do to
avoid this.
You must teach your clients that they can not unplugged the server. That
they should use a UPS. That the UPS should turn off the server after a
period. That they should backup it. That they should check the backups.
You can not automatize all the server tasks.
That's the cost of having a server.
Users use to understand this. Really. If they don't, you will (maybe)
sell your product, but you will have some deep problems if the future:
when the server is full of data, and they need your application to make
there jobs done.
a feature request that we all need.
Just let the user choose if they want to use what they already have.
It's there server anyway. ;-)
If there are too much people already having this problem, maybe they can
just get togheter and sponsor this. The Firebird Foundation could easily
help on this move.
But this will not fix the question: should we remove the server or not?
But for now, with more developers comming to Firebird, we should make
them understand that they don't own the server. :-)
Artur
[ I agree with Set:
Fulda and Bamsemums belongs together - almost like Firebird and Firebird
Foundation (which you by the way can join at
http://www.firebirdsql.org/ff/foundation). ]
> Currently I am working on convincing my company to switch from MSDE to FB as aThat's an easy one. :-))
> backend for one of our products.
> As you can imagine, if we don'tThat's my market, also. :-)
> play nice and nuke somebodys FB install in order to install our application, we
> will make a lot of unahppy developers :-) On the other hand if we dont do it
> then we shoot ourselfs in the foot, bacause the customer won't be able to use
> our software and will switch to the competition. (and guess what a for-profit
> company will do :-)?)
What I use is embedded for mono-users and demos, and if they want to try
a server installation, I give them instructions to do that. This use to
work very well, and the end-user will have from the start a small notion
of client/server.
I'm arguing against automatic removing the server. I'm arguing against
having registry keys to identify Firebird applications.
With embedded it's easy to create demo versions. That's the ones that
are unninstalled more often.
Client/Server applications need some care. Always. Someplace where there
is a server (any server - mail server, file server, sql server...) you
will need someone to take care of it. There is nothing you can do to
avoid this.
You must teach your clients that they can not unplugged the server. That
they should use a UPS. That the UPS should turn off the server after a
period. That they should backup it. That they should check the backups.
You can not automatize all the server tasks.
That's the cost of having a server.
Users use to understand this. Really. If they don't, you will (maybe)
sell your product, but you will have some deep problems if the future:
when the server is full of data, and they need your application to make
there jobs done.
> well, yes you were the first one to install FB on this computer, but you alsoLocal users should be related to the database, I agree with you. That's
> changed the SYSDBA password. You did tell the user the new password, but he
> forgot it as quickly as he clicked OK on that message box. Now what? The only
> thing I can do is to kill your security.fdb and hope that the customer needs my
> software more then he needs yours.
a feature request that we all need.
Just let the user choose if they want to use what they already have.
It's there server anyway. ;-)
> Is there a sollution to this problem? I don't know, maybe the security should beSecurity will be tied to databases. In the future.
> tied to teh database files and not the server, maybe there are other solutions
> but one thing is for sure: a lot of people are in the same boat as Thomas and
> are getting really nervous as more and more developers flock to FB.
If there are too much people already having this problem, maybe they can
just get togheter and sponsor this. The Firebird Foundation could easily
help on this move.
But this will not fix the question: should we remove the server or not?
But for now, with more developers comming to Firebird, we should make
them understand that they don't own the server. :-)
Artur
[ I agree with Set:
Fulda and Bamsemums belongs together - almost like Firebird and Firebird
Foundation (which you by the way can join at
http://www.firebirdsql.org/ff/foundation). ]