commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Gary Gregory <garydgreg...@gmail.com>
Subject Re: [VFS] Feedback on VFS-422 please
Date Thu, 21 Jun 2012 21:41:48 GMT
On Thu, Jun 21, 2012 at 5:10 PM, Mark Fortner <phidias51@gmail.com> wrote:

> Gary,
> There was some talk a while back about implementing File System-specific
> Operations.  I think what Mario had in mind was supporting version control
> system functionality through VFS.  It strikes me that this might be the
> best way for implementing functionality that makes use of the JSch bells
> and whistles without breaking encapsulation.
>

Hi Mark,

I am not sure I understand what you are proposing.

Gary


> Regards,
>
> Mark
>
>
> On Thu, Jun 21, 2012 at 2:00 PM, Gary Gregory <garydgregory@gmail.com
> >wrote:
>
> > Hi All:
> >
> > WRT VFS-422: [SFTP] Allows to create other channels in SftpFileSystem (
> > https://issues.apache.org/jira/browse/VFS-422)
> >
> > I'll like to solicit feedback from the list on this issue. I've cleaned
> up
> > the proposed patch in the ticket and attached it back to the issue as
> >
> >
> https://issues.apache.org/jira/secure/attachment/12532940/vfs-422-v2-by-gg.diff
> >
> > First and foremost is whether this feature should be in VFS at all.
> >
> > Internally, VFS does not use the feature, so it is a convenience to the
> > client application.
> >
> > A simpler solution from VFS' POV would be to make the Jsch Session object
> > accessible via a public getSession() method. The client would then be
> > responsible for the resources created and be free to use all of the Jsch
> > bells and whistles.
> >
> > One can argue that this breaks encapsulation and surfaces VFS' private
> > parts (pun intended). But because the point of the feature is get to
> > another Jsch object (a Channel) through a Session, encapsulation would be
> > broken anyway, so this point is moot IMO.
> >
> > A second issue has to do with implementation of the method
> > com.jcraft.jsch.CommonsVFSChannelFactory.createChannel(AtomicLong,
> > SftpChannelType)
> >
> > - Should finalization be used?
> > - Should the counts be managed through the connect and disconnect methods
> > instead?
> > - There is a lot of repeated code in the method, should a delegate/proxy
> to
> > a channel be used instead such that there is no code duplication?
> >
> > Thank you,
> > Gary
> >
> > --
> > E-Mail: garydgregory@gmail.com | ggregory@apache.org
> > JUnit in Action, 2nd Ed: <http://goog_1249600977>http://bit.ly/ECvg0
> > Spring Batch in Action: <http://s.apache.org/HOq>http://bit.ly/bqpbCK
> > Blog: http://garygregory.wordpress.com
> > Home: http://garygregory.com/
> > Tweet! http://twitter.com/GaryGregory
> >
>



-- 
E-Mail: garydgregory@gmail.com | ggregory@apache.org
JUnit in Action, 2nd Ed: <http://goog_1249600977>http://bit.ly/ECvg0
Spring Batch in Action: <http://s.apache.org/HOq>http://bit.ly/bqpbCK
Blog: http://garygregory.wordpress.com
Home: http://garygregory.com/
Tweet! http://twitter.com/GaryGregory

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message