[Biojava-l] Maintaining biojava on several machines

[Matthew Pocock]

Hi Mark,

The releases (e.g. 1.1) are stable and maintained. If you notice bugs in 
these, then please tell us and it will be fixed both in the release and 
the trunk. If you are intending to run production code using BioJava and 
1.1 has the functionality, then I would advise using this.

The main trunk is in constant flux, with lots of new features being 
added daily, but you must be aware that we make no guarantees for the 
head other than that it compiles. Of course, if you must have the latest 
shineyest features, then you can run off a localy tested trunk build and 
only replace this with another 'trusted' build. Keeping your run-time 
systems in sync with cvs is probably a very bad thing.

As far as client/server goes, we have had positive experiences writing 
servers using BioJava, so if the functionality can be thought of as 
services (database access, running sequence searches etc.), then it all 
works. Java lets you use raw sockets, corba, RMI etc. with relatively 
little pain. Of course, you will need to write a client as well, so if 
the client has BioJava dependancies, you're back to square 1.

Matthew

Schreiber, Mark wrote:

> Hi,
> 
> What are peoples experiences with maintaining biojava on several machines??
> I currently try and keep updated versions on just two machines (both
> windows). But in the near future our group may need to install biojava on
> quite a few Unix/Linux machines. Due to the rapidly (almost daily) changing
> nature of biojava is it possible (or advisable) to keep all machines up to
> date with CVS or would it be better to download the occasional snapshots
> that come out?
> 
> Also in a networked system can biojava be accessed from just one server by
> maybe RMI (I'm very much out of my depth here but suggestions are welcome).
> 
> Thanks
> 
> Mark
> _______________________________________________
> Biojava-l mailing list  -  Biojava-l@biojava.org
> http://biojava.org/mailman/listinfo/biojava-l