Subject | Re: [Firebird-general] Embedded Engine |
---|---|
Author | Paul Schmidt |
Post date | 2004-06-20T00:59:07Z |
On Sat, 2004-06-19 at 07:56, Lester Caine wrote:
benefits of having multiple engines. This could be done two ways:
1) Binaries, package everything together, pick a default install, say
Superserver, however give the installer the option of picking Classic or
embedded, so that experience folks can do whatever they want. There is
nothing wrong with installing the classic version, but disabled, and the
embedded as well as client, so that by changing a configuration file or
a couple of registry entries you can switch.
2) Sources, building from sources, should be easier, this could even be
as simple as make, which builds the binaries installer version. I know
this means that some components would need to be compiled 4 times, but
so what? It's not like you compile it six times a week.
Of course there could be options to build more specifically, and
experienced users could fine tune this process.
One thing that is needed, is a good set of documentation, relying on an
ancient Interbase beta documents and some readme files is more likely to
scare off a potential user then picking which version of the engine to
build.
Paul
> Dariusz Zelichowski wrote:Personally, I think we can make it easier to market, without loosing the
>
> > Wouldn't it be too hard to decide which setup to leave? How
> > can you decide for me whether I need CS or SS, or embedded
> > without risking that I loose my interest in the database.
> > Each one of the versions seems to have a benefit the others
> > don't, but also suffers from "features" the others are free
> > of.
>
> The fact that you can ask that question means that you are well on the
> way to understanding Firebird ;)
>
> The 'Beginners Guide' should be able to say - Install X and do this and
> get a new user interested. THEN the details can be dealt with :)
benefits of having multiple engines. This could be done two ways:
1) Binaries, package everything together, pick a default install, say
Superserver, however give the installer the option of picking Classic or
embedded, so that experience folks can do whatever they want. There is
nothing wrong with installing the classic version, but disabled, and the
embedded as well as client, so that by changing a configuration file or
a couple of registry entries you can switch.
2) Sources, building from sources, should be easier, this could even be
as simple as make, which builds the binaries installer version. I know
this means that some components would need to be compiled 4 times, but
so what? It's not like you compile it six times a week.
Of course there could be options to build more specifically, and
experienced users could fine tune this process.
One thing that is needed, is a good set of documentation, relying on an
ancient Interbase beta documents and some readme files is more likely to
scare off a potential user then picking which version of the engine to
build.
Paul
>
> > I think that it would be much more beneficial to FB
> > community if a major Linux distro included it by default,
> > and that we might want concentrate our efforts around that
> > goal. If Fedora, or Mandrake or others include Firebird
> > then everybody will suddenly loose their confusion ad
> > FireBird Foundation can not only keep the current versions,
> > but even add new ones, including FireBlue for boys and
> > FirePink for gals.
>
> Now that the confusion with 'the other Firebird' has been resolved, the
> reasons for not doing it can be squashed, so yes - every distribution
> should have at least FB1.5 on it.