Re: [Publicity] Blog Czar update, week of Sept 9

[Anthony PERARD <anthony.perard@xxxxxxxxxx>]

On Tue, Sep 17, 2013 at 01:05:09AM +0200, Dario Faggioli wrote:
> On lun, 2013-09-16 at 16:13 +0100, Anthony PERARD wrote:
> > I'm trying to write a post about which feature were missing from QEMU
> > compared to qemu-traditional by looking at which feature made it in 4.3,
> > which are still missing and which are planned.
> > 
> Ok, first of all, thanks for doing it and sharing the draft.
> 
> It is indeed a bit short, but that is not necessarily an issue. However,
> I'd try to think whether there isn't some more interesting and useful to
> know (for our users) content that we/you could not add...
> 
> > And here is the draft:
> > http://blog.xen.org/?p=7887&preview=true
> > 
> For instance, in the very first sentence you mention both "QEMU" and the
> "old qemu-traditional". Apart from adding the proper hyperlinking, what
> about stating once more, as clearly as possible, what the differences
> are, i.e., which one is what, which one is the default, etc.?

Yes, I can write a bit more about the difference between qemu and
qemu-trad.

> I know that piece of info is out there already, but, to me, this really
> looks like something really easy to get confused about, even for
> experienced users.
> 
> "A thing called dirty tracking of physical memory", worth expanding a
> little bit to tell people what it is and why it is important? Still in
> that sentence, I'm not sure I understand what the following means: "but
> it has also been backported to 4.2, so itâs not that new".
> 
> "Missing feature_s_?" ?

What I meant was the memory dirty tracking was not part of the 4.2
release, but appear later in the 4.2.2 release. This was an important
missing feature to be able to live-migrate a guest.

> > I can't find anything else to put in this post, so hopefully, there
> > nothing missing.
> > 
> What about some of the development that is happening right now, whether
> or not it is targeting 4.4? You've got QEMU stubdom there already, I
> know, but is that the only thing happening on the qemu side? Perhaps we
> can quickly list what we've got in George's development update that is
> qemu related with a few words of context? I'm sure there's someone
> (Fabio Fantoni?) working on improved USB controller, spice and qxl
> support... No need to have a detailed tech report, but it perhaps is
> worth at least mentioning?

Yes, I can try to list more things.

> Another thing that I think could be interesting to (re)state is what
> does it mean to achieve the goal of getting rid of qemu-traditional? Why
> is it a good thing? For instance, in 4.4 there will be config options
> for using the distro provided QEMU, instead of cloning and building our
> own (whether it is upstream or traditional), wont it? If yes, that I
> think is quite a _big_deal_ (or so I always hear from distro people!).
> 
> So... Am I talking nonsense? Do you think it's worth trying to add at
> least some of the things I mentioned?

Yes, I'll try to put something togethere.

Thanks for the review,

-- 
Anthony PERARD

_______________________________________________
Publicity mailing list
Publicity@xxxxxxxxxxxxxxxxxxxx
http://lists.xenproject.org/cgi-bin/mailman/listinfo/publicity