syncany-team team mailing list archive

[Samúel Jón Gunnarsson <sammi@xxxxxxxxxxxxxxx>]

On 06/27/2011 01:44 PM, Philipp Heckel wrote:
> Hi Samuel,
>
> in the syncany.sh file, the "-Xms512m -Xmx512m" parameters dont really
> make sense. They mean maximum memory and minimum memory, right?
> Shouldn't it be just a minimum of -Xms256m or something like that? Or
> is this a Mac thing?
After I switched to using JarBundler for creating Syncany.app it is no 
longer necessary to change syncany.sh so you can revert the change I 
made to the script.  The reason for the change was Mac specific. I will 
take a closer look at the max/min heap size settings when I get home and 
adjust minimum heap size.
> Cheers,
> Philipp
>
> 2011/6/27 Philipp Heckel<philipp.heckel@xxxxxxxxx>:
> > Hi Samuel,
> >
> > that sounds great.
> >
> > > I pushed the changes I've made to the project at lp:sammi/syncany/mac . I
> > > have added:
> > >
> > > Basic notification support with Growl (http://growl.info)
> > > Ant target that bundles all relevant files into Syncany.app with JarBundler
> > > (http://informagen.com/JarBundler/).
> > > Changes to BufferedWatcher.java so that is portable between operating
> > > systems.
> > >
> > > I have tested Syncany.app on other osx machine and it works out of the box.
> > I'll merge your branch into my "newsync" branch. Once I decide to make
> > the "newsync" branch the new trunk, it'll be available for everyone.
> >
> > > I looked at the documentation for jpathwatch and had to remove the BSD
> > > ENTRY_RENAME_FROM and ENTRY_RENAME_TO events since they are not supported in
> > > BSDPathWatchService.java. I have now made modification to the
> > > BufferedWatcher.java so that I checks for which operating system is running
> > > and assigns the right KIND instead of hardcoding it. The code that I changed
> > > outside of contrib directory was just for better read and is something that
> > > I did not intend to change.  I do agree that It is probably best to create
> > > MacBufferedWatcher and other inheritance where we have to deal with mac
> > > specific issues. But for now I think that this small modification I've made
> > > to the BufferedWatcher.java does not need to be in a subbclass of
> > > BufferedWatcher.
> > Good choice. I probably would have done the same thing :-D
> >
> > Cheers,
> > Philipp