[PATCH 0 of 6] Evolve support patches from Sean

[Sean Farley]

Thomas De Schampheleire <patrickdepinguin at gmail.com> writes:

> On Tue, Jul 28, 2015 at 4:11 PM, Mads Kiilerich <mads at kiilerich.com> wrote:
>> On 07/28/2015 02:38 PM, Thomas De Schampheleire wrote:
>>>
>>> On Tue, Jul 28, 2015 at 1:02 PM, Andrew Shadura <andrew at shadura.me> wrote:
>>>>
>>>> Hello everyone,
>>>>
>>>> I'm submitting the some of the patches Sean Farley wrote a while ago to
>>>> support changeset obsoletion and evolution. I've been using them for a
>>>> while and they don't seem to break stuff. There were more patches, but
>>>> some of them didn't work always right and also seem obsolete with the
>>>> latest versions of Mercurial.
>>
>>
>> There is not much documentation of what this really do or daily usage of
>> this. Some test coverage would be very nice.
>>
>> The biggest feature in this area must be for hiding hidden/obsoleted
>> changesets. I don't know if we already do that or if that is on the wish
>> list?
>>
>> Has this been tested with the lowest supported Mercurial version - that it
>> doesn't break and that it actually works as expected with evolve'd repos?
>>
>>> A related changeset is the ability to mark a repository as
>>> non-publishing in Kallithea:
>>>
>>> https://smf.io/kallithea/changeset/bfe813360295e1e7ff977e9c518e1b012e23d7e0
>>>
>>> but currently it has no associated checkbox, nor is it per-repository
>>> (currently global).
>>>
>>> It would be great if that patch could be improved and sent too...
>>
>>
>> That would just be the lightweight alternative to evolve, right? The bigger
>> need in this area is the server side of evolve itself.
>
> Well, not exactly: in order to use evolve successfully, the
> repositories involved actually need to be non-publishing. Otherwise,
> changesets will still be marked public and evolve will not obsolete
> them.

Yes, this is correct. The non-publishing option has to go in before
evolve, else nothing will ever be obsolete.