where to report about libvirt strings

[Daniel Veillard]

On Wed, Jun 03, 2009 at 02:07:48PM +0530, Runa Bhattacharjee wrote:
> Daniel Veillard wrote:
>> On Wed, Jun 03, 2009 at 10:16:13AM +0530, Runa Bhattacharjee wrote:
>>> Hello,
>>>
>>> Is bugzilla a good way to ask for clarifications about a couple of  
>>> strings in libvirt or should they be mailed somewhere?
>>
>>   Ask, I'm listening !
>>
>
> Thanks :-)
>
> Here goes.
>
> Clarifications Required for the following:
> =========================
> #: src/test.c:1206
> #, c-format
> string: saving domain '%s' failed to allocate space for metadata

  The user tried to save a domain on disk, %s is the domain name and the
operation failed to allocate memory for the operation

> Q. Not sure how this should be translated
> -------------------------------------------------------------
> #: src/lxc_container.c:186
> string: unable to send container continue message (& similar strings with 
> 'container continue')

  the container is a domain, we are asking the kernel to continue.
this is done by writing data to an open file descriptor, and the write
failed.
  "container continue" doesn't mean anything in itself, it's just
the type of domain being handled, and the fact we try to send a continue
message.

> Q. What is 'container continue' and should it be translated?
> -------------------------------------------------------------
>
> #: src/lxc_container.c:789
> string: failed to run clone container

  Same thing we try run the clone operation on that type of domain.

> Q. Is 'clone container' a command?
> -------------------------------------------------------------
>
> #: src/qemu_driver.c:4376 src/uml_driver.c:1696
> string: cannot set autostart for transient domain

  it's a domain which disapears completely once it has finished running,
there is no permanent description for it.

> Q. Should 'transient domain' be translated?
> -------------------------------------------------------------
>
> #: src/remote_internal.c:3163
> #, c-format
> string: block peek request too large for remote protocol, %zi > %d (& 
> similar strings with 'peek request')

  Peek is the operation of reading something in that context. Here we
  try to read a block, but the block size is too large for the protocol

> Q. Is 'block peek request' a command and should 'peek request' be translated?
> -------------------------------------------------------------
>
> #: src/virsh.c:4895 src/virsh.c:4936
> string: Dettach node device its device driver before assigning to a domain.

  it's detach *from* which is missing
Basically we plan to use a device exclusively for a domain. The device
needs first to be detached from the node, i.e. its driver need to
free the device before it can be assigned and used by a given domain.

> Q. Not sure how this should be translated.
> -------------------------------------------------------------
>
> #: src/virsh.c:4935
> string: reattach node device its device driver

  reattach node device to its device driver

that's the reverse operation, the device is reattached to the node OS
and basically the node driver can make use of it again.

> Q. Not sure how this should be translated.
> -------------------------------------------------------------
>
> #: src/xs_internal.c:327
> string: adding watch @releaseDomain (& similar strings with @<value>)

  Those are name of actions, I doubt translating them would add much

> Q. How should these strings be translated?
>
> =================================
> <end of clarification list>
>
> Possible typo bug:
> ==============
>
> #: src/openvz_conf.c:198
> #, c-format
> Cound not read 'IP_ADDRESS' from config for container %d
>
> 'Cound' should read 'Could'.
> =================

 yup a bunch of then in that file,

   Fixed in upstream CVS, thanks !

Daniel


-- 
Daniel Veillard      | libxml Gnome XML XSLT toolkit  http://xmlsoft.org/
daniel at veillard.com  | Rpmfind RPM search engine http://rpmfind.net/
http://veillard.com/ | virtualization library  http://libvirt.org/