[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: FTBS owncloud-client



Hey,

I've got reponse from the owncloud-client team [1]:

"I would be willing to take patches, but without our explicit support for that 
for the next time.

About the options:
  * a null watcher is no option because we rely on it for the sync 
functioality
  * QFileSystemWatcher is also not a good solution
  * A file watcher implementation for $OS is what we should have. Please 
provide pull requests or patches."

Regards,

sandro

[1] https://github.com/owncloud/mirall/issues/1060#issuecomment-25803748
--

Am Dienstag, 1. Oktober 2013, 19:59:49 schrieb Pino Toscano:
> Hi,
> 
> (CC'ing -bsd@ too, since basically the same email for owncloud-client
> has been sent there too.)
> 
> In data domenica 29 settembre 2013 17:48:26, Sandro Knauß ha scritto:
> > many thanks Svante and Pino for the hurd patches for ocsync [1] and
> > qtkeychain. The both packages now complile successfully for hurd.
> > 
> > But unfortunatelly owncloud-client now fails to compile [2]:
> > error: invalid use of incomplete type 'class
> > Mirall::FolderWatcherPrivate'
> > 
> > Maybe you could help me another time to create a patch to build
> > owncloud- client for hurd.
> 
> The problem affecting both kFreeBSD and Hurd is the implementation of
> FolderWatcher (in src/mirall/folderwatcher*): it seems owncloud-client
> provides own implementations for some OSes (Linux, Mac, Windows),
> leaving a compile error on other OSes.
> 
> As I see it, there are 3 different ways to deal with situation:
> 1) implement a null folder watcher to be used when there is no native
>    one available
> 2) implement a folder watcher using QFileSystemWatcher to be used when
>    there is no native one available
> 3) require the implementation of a folder watcher for $OS (on *BSD using
>    kqueue, on Hurd using the file/dir notifications, etc), not building
>    on that $OS until there is native folder watcher implemented
> You should ask upstream which way they prefer/want, so eventual porting
> efforts could go according to that.
> 
> Note that I did not inspect owncloud-client further for other prorting
> issues.

Attachment: signature.asc
Description: This is a digitally signed message part.


Reply to: