Subject | Re: firebird client for ARM platform (KinterbasDB interface) |
---|---|
Author | David Rushby |
Post date | 2006-09-24T10:34:49Z |
--- In firebird-python@yahoogroups.com, "Petr Jakes" <petr@...> wrote:
the following choices (listed in the order that I'd try them, if I
were in your position):
- Use something like SQLRelay (http://sqlrelay.sourceforge.net/ ).
There is a Python interface to SQLRelay, though I don't know whether
it works on ARM, or how compliant it is with the Python DB API 2.0.
- Put an XML-RPC, SOAP, CORBA, or Pyro (http://pyro.sourceforge.net/ )
layer between the ARM devices and the database server. This approach
could work quite well, if performance requirements aren't too
stringent. Of the four listed RPC mechanisms, I'd choose Pyro if
possible, because it's by far the most transparent and Pythonic.
- Port your application to a programming environment that already has
an FB client library implementation that runs on ARM. If Jaybird
works properly on ARM, and you were to use Jython and zxJDBC, this
might not require very much effort.
- Port the FB client library to ARM.
- Create a pure-Python implementation of the FB client library and a
Python DB API 2.0-compliant interface atop it (a huge amount of work,
no doubt).
> Se would like to connect our ARM based computers (Linksys NSLU2 withIf the FB client library hasn't been ported to ARM, I guess you have
> the DebinaSlug Linux) to the firebird database running on the server.
> Is there a way how to compile linux Firebird (client ONLY!!!) for ARM
> platform?
>
> We are using Python and KinterbasDB Firebird interface, which runs
> only, when the Firebird is installed on the machine.
the following choices (listed in the order that I'd try them, if I
were in your position):
- Use something like SQLRelay (http://sqlrelay.sourceforge.net/ ).
There is a Python interface to SQLRelay, though I don't know whether
it works on ARM, or how compliant it is with the Python DB API 2.0.
- Put an XML-RPC, SOAP, CORBA, or Pyro (http://pyro.sourceforge.net/ )
layer between the ARM devices and the database server. This approach
could work quite well, if performance requirements aren't too
stringent. Of the four listed RPC mechanisms, I'd choose Pyro if
possible, because it's by far the most transparent and Pythonic.
- Port your application to a programming environment that already has
an FB client library implementation that runs on ARM. If Jaybird
works properly on ARM, and you were to use Jython and zxJDBC, this
might not require very much effort.
- Port the FB client library to ARM.
- Create a pure-Python implementation of the FB client library and a
Python DB API 2.0-compliant interface atop it (a huge amount of work,
no doubt).