Build Debate: Followup on Build Naming

[Simon Schampijer]

Morgan Collett wrote:
> On Tue, Apr 8, 2008 at 9:51 AM, Simon Schampijer <simon at schampijer.de> wrote:
>> Dennis Gilmore wrote:
>>  > On Monday 07 April 2008, Michael Stone wrote:
>>  >> cjb, cscott, and I just chatted about build names. We have absolutely no
>>  >> problem announcing official-703 (when candidate-703 becomes official)
>>  >> under whatever name seems good but we have no consensus about what that
>>  >> name should be. cscott proposes '8.1' on the basis that it will be our
>>  >> first 2008 release. mstone thought we should bake a month into the name
>>  >> (perhaps 2008.04 or April-2008). Scott strongly preferred to avoid
>>  >> baking a month designator into the name because, as best I understand,
>>  >> he thinks we cannot afford to ship another release until we've made
>>  >> 'enough' improvement in at least one of our (approximately) four
>>  >> networking scenarios.
>>
>>  I like scott's proposal '8.1'. Putting a month or a
>>  season(summer/winter) there restrict us - and since we have seen that
>>  'update.1' has taken longer than expected it would be wise not to.
> 
> We need to call it something while it's proposed / under development,
> and before we know exactly when it will ship. Some distros use
> codenames while it's under development, and then the final release is
> named accordingly. (For example, what if a release we think will come
> out in late 2008 slips to 2009?)
> 
> Otherwise, strict time-based releases would be required (which I'm in
> favour of, but I don't know if that decision has been taken yet...)
> 
> Morgan

Good point, I think having a feature or a time based releases has a 
great impact on the naming. In feature based releases the naming dennis 
proposed (e.g. OLPC-2) would make sense to me.

Simon