[libvirt] [PATCH for 5.2.0] docs: Add virt-lightning app

Cole Robinson crobinso at redhat.com
Sat Apr 6 16:35:06 UTC 2019


On 4/6/19 6:28 AM, Michal Prívozník wrote:
> On 4/4/19 12:58 AM, Cole Robinson wrote:
>> On 4/3/19 5:03 PM, Ján Tomko wrote:
>>> On Wed, Apr 03, 2019 at 02:10:19PM -0400, Cole Robinson wrote:
>>>> On 4/1/19 8:19 AM, Michal Privoznik wrote:
>>>>> There was this introduction made on the users list:
>>>>>
>>>>> https://www.redhat.com/archives/libvirt-users/2019-March/msg00046.html
>>>>>
>>>>> Add the application onto the list of apps known to use libvirt.
>>>>>
>>>>> Signed-off-by: Michal Privoznik <mprivozn at redhat.com>
>>>>> ---
>>>>>  docs/apps.html.in | 6 ++++++
>>>>>  1 file changed, 6 insertions(+)
>>>>>
>>>>> diff --git a/docs/apps.html.in b/docs/apps.html.in
>>>>> index 209854b6ac..62914b575a 100644
>>>>> --- a/docs/apps.html.in
>>>>> +++ b/docs/apps.html.in
>>>>> @@ -99,6 +99,12 @@
>>>>>          machines. It is a command line tool for developers that
>>>>> makes it very
>>>>>          fast and easy to deploy and re-deploy an environment of vm's.
>>>>>        </dd>
>>>>> +      <dt><a
>>>>> href="https://github.com/virt-lightning/virt-lightning">virt-lightning</a></dt>
>>>>>
>>>>> +      <dd>
>>>>> +        Virt-Lightning uses libvirt, cloud-init and libguestfs to
>>>>> allow anyone
>>>>> +        to quickly start new VM. Very much like a container CLI
>>>>> interface, but

s/start new/start a new/

>>>>> +        locally.

'Very much like a container CLI, but locally.' I don't really understand
what this means, maybe better is 'Very much like a container CLI, but
with a virtual machine.' or similar

>>>>> +      </dd>
>>>>>      </dl>
>>>>>
>>>>>      <h2><a id="configmgmt">Configuration Management</a></h2>
>>>>>
>>>>
>>>> I don't get the point of keeping this as a static page in git. It's
>>>> always going to be out of date, or needing tweaks that IMO add noise to
>>>> the dev mailing list.
>>>
>>> The changes proposed to this page have always shown a high
>>> signal-to-noise ratio and are neligible to all the other changes made to
>>> libvirt source code.
>>>
>>
>> I didn't say it was _much_ noise :) But I take your point
>>
>>>> Can't this be a wiki page?
>>>
>>> One argument against a wiki page would be that the barrier for
>>> contributing is higher.
>>>
>>> To get your change merged in git, all you need is to send an e-mail.
>>>
>>
>> There's three cases:
>>
>> 1) contributor asks someone else to add app to the list
>> 2) new contributor does it themselves
>> 3) existing contributor does it themselves
>>
>> In both wiki and git worlds, #1 is just an email 'hey this app exists'.
>> Like the case above: someone mentioned it on the list, and michal is
>> adjusting apps.html for them
>>
>> #2 is not just an email: it's git clone, make the change, hopefully test
>> it, then send it.
>>
>> #2 for the wiki yes it's painful for drive by contributors because they
>> need to request an account, possibly more painful depending on how
>> comfortable people are with git.
>>
>> #3 for both cases is indistinguishably low effort. Except the git case
>> always requires minimum 2 mails to libvir-list.
>>
>> And every git case requires some reviewer bandwidth, CI triggering and a
>> permanent git commit.
>>
>> Anyways I'm not gonna die on this hill, I've said my piece (again ;) ),
>> if no one else is on board I'll shut up about it
> 
> Do we have a resolution here? I like 1) and 3) and I'm volunteering for
> creating the wiki page if we decide to go with 3).
> 

Dan seemed against the wiki page too in another mail. For this change
with the way up above:

Reviewed-by: Cole Robinson <crobinso at redhat.com>

But if you are like me and prefer the wiki page route then obviously I
endorse that too

Thanks,
Cole




More information about the libvir-list mailing list