maria-developers team mailing list archive
Mailing list archive
Re: Is it soon 'end of this year'?
OK .. we will check this again.
But in principle we want (always) not to have any external dependencies in
our prorams/apllications at at all. We link statically everything to the
extend possible. People should download and install SQlyog with a single
click and it will just work. PERIOD! :-) That is what Windows users expect
from an installer. How do we accomplish this with the MariaDB API? Compile
it as a .lib/.dll and bundle it with our installer and let the installer
copy it into the installation directory? Or should what you describe here
happen at compile-time (linking statically)? Also please remember that
SQLyog is a Windows program and our compiler is Visual Studio (2010
Maybe you can provide *Windowssand Visual Studio pecific* guidelines?.
Also (BTW) 1- 1½ years ago I asked similar questions and the only reply I
had was (very abstractly) that MariaDB's mplementation was far superior to
Oracle's. I was actually *despearately* asking for help at that time. The
reply ("we are the best") was very little helpful and we have already lost
significant business due to this.
Anyway, I wili ask our developer team to recover as much information as
possible due to this.
On Mon, Oct 13, 2014 at 7:49 PM, Sergei Golubchik <serg@xxxxxxxxxxx> wrote:
> Hi, Peter!
> On Oct 13, Peter Laursen wrote:
> > Well .. where will I find it documented how to use it with Oracle-MySQL
> > configured for their LDAP authenticaton setup? The same code that worked
> > with oracles API does not work with the MariaDB API.
> There's only one difference that comes to my mind.
> Oracle-MySQL cleartext plugin is integral part of the client library,
> it's statically compiled into it. In MariaDB it's a separate plugin -
> which is loaded automatically on demand, but you still need to make sure
> you set the plugin-dir path correctly.
> There's nothing else I can think of, our test suite verifies that
> cleartext plugin exists and that it actually works, so I don't think
> we've broken it somehow.