Re: [meta-kde][PATCH 3/3] README: update contributor list

["Peter A. Bigot" <pab@pabigot.com>]

On 03/07/2013 09:41 AM, Paul Eggleton wrote:
> On Thursday 07 March 2013 09:50:21 Philip Balister wrote:
>> On 03/07/2013 09:41 AM, Otavio Salvador wrote:
>>> On Thu, Mar 7, 2013 at 11:35 AM, Paul Eggleton
>>>
>>> <paul.eggleton@linux.intel.com> wrote:
>>>> On Thursday 07 March 2013 15:03:17 Koen Kooi wrote:
>>>>> Op 7 mrt. 2013, om 14:04 heeft Martin Jansa <martin.jansa@gmail.com> het
>>>>>
>>>>> volgende geschreven:
>>>>>> It would be nice to know yocto-1.4 release name in advance and name it
>>>>>> the same as branch in oe-core/meta-oe will be (denzil, danny, ...), but
>>>>>> I guess it can be renamed later.
>>>>> For angstrom I'm going to use 'yocto-1.4' in the branch name, I have
>>>>> trouble remembering which names maps to which release. And the Yocto
>>>>> compliance program talks about 1.3, .14 etc, not about codenames.
>>>> Wouldn't it be worth us trying to standardise rather than all doing our
>>>> own
>>>> thing and users having to figure out what matches up between different
>>>> layers? If others feel the same as you, then maybe we should all be
>>>> using that schema.>
>>> Or we use a codename or we don't.
>>>
>>> For me, codenames work fine but for users it is sometimes confusing as
>>> the website and marketing people talk about Yocto 1.3 or 1.4 while the
>>> involved people talk about codenames. So I find myself explaining it
>>> over and over again.
>> Add me to the list of people that find codenames confusing. I can't
>> reliably list releases in order by name.
> FWIW, in the layer index web app against each branch I have a field for a short
> description (e.g. denzil could have something like "old stable" or whatever is
> helpful to explain it to people) and a sort order so that they can be sorted
> correctly where listed.
>
> That doesn't take away the need to resolve this issue of branch names across
> layers, but it may help users to understand what these codenames mean if they
> continue to be used, at least when they see them in the layer index at least.
A comment on that from up here in the peanut gallery: I don't personally 
find codenames valuable, but if they're used it would be nice if they 
were selected using a policy that allowed at least their relative order 
to be determined by inspection.  That danny=1.3 sorts lexicographically 
before denzil=1.2 is confusing.

Peter