strange phases order (public < draft < secret)

[anatoly techtonik]

On Fri, Jan 3, 2014 at 11:52 PM, Pierre-Yves David
<pierre-yves.david at ens-lyon.org> wrote:
> On 01/03/2014 12:31 PM, anatoly techtonik wrote:
>>
>> On Fri, Jan 3, 2014 at 9:47 PM, Matt Mackall <mpm at selenic.com> wrote:
>>>
>>> On Fri, 2014-01-03 at 20:33 +0300, anatoly techtonik wrote:
>>>>
>>>> I really don't get why changeset evolution phases
>>>> are inverted in Mercurial. You start with secret and
>>>> draft changesets, which then are evolved into public.
>>>>
>>>> So, logicall the "public" is "higher" on the evolution
>>>> chain. Why is it the reverse?
>>>
>>> Just about every display of changesets shows more recent changes above
>>> older ones. Also, earlier revisions have lower numbers. So down = lower
>>> = older = public is the natural direction.
>>
>> I kind of can accept that, but it is still confusing to hear that
>> public is older
>> than secret and draft.
>>
>>> Originally, the analogy for phases was frozen vs liquid, and thus there
>>> was an obvious temperature component to the analogy as well.
>>
>> The analogy that public is more hot is cool. =)
>>
>>> Evolution is a concept that came later (and is not even quite yet part
>>> of hg, while phases have been a core feature for years).
>>>
>>> $ hg help phases | grep evolution
>>
>> Sorry, Windows here. No grep, and even less is not possible if you
>> don't know more.
>>
>> Is there any desire to invert the phase of phase terminology?
>
>
> There is no desire to have the termilogie inverted from the implementation.
> It does not bring real good and will bring much confusion to people writting
> hook and extension.

Ok. I'd just avoid the names higher and lower, because the picture is correct
from evolution point of view,:

   (public < draft < secret)

is the same as:

   (secret > draft > public)

Just need to overload "less/more" operator with "evolution" arrow.