eol with hg 1.4.x?

[Martin Geisler]

Mark Hammond <skippy.hammond at gmail.com> writes:

> On 6/01/2010 11:23 AM, Martin Geisler wrote:
>> Brett Cannon<brett at python.org>  writes:
> ...
>>> I say we make sure the extension is rock-solid and then Python makes
>>> the switch regardless of what others want in terms of new features.
>>> The people who brought up this issue have had quite a while to
>>> follow the work and be active participants.
>>
>> Well said :-)
>
> I agree with that - the set of *features* required is becoming fairly
> well defined. What isn't clear yet though is how well things actually
> work.

Yeah, I've not been good at updating the wiki page. I don't like how it
has become a big dumping ground for various ideas instead of an actual
plan that tells us what to implement.

But please look at the tests. That is really where the best
specification is right now. Clone the repository and run

  make test-eol-clone TESTFLAGS=--debug

to see how a session looks like. Similarly for the other tests:

  http://bitbucket.org/mg/hg-eol/src/tip/tests/

> As a result, I think setting a specific date for the switchover is
> premature - unless it is being suggested the switchover will take
> place on that date *regardless* of how well it works in practice - but
> I don't think anyone would really suggest that.
>
>> I guess we have a sort of chicken-and-egg problem here. Nobody wants
>> to use the extension before it's done, and it wont be done until
>> people have used it... So I'm unsure about how to best move along,
>> but I guess people should try it out some more.
>
> Yeah, this is indeed a problem. The last time I had a play I fell over
> at the first few hurdles. I'll try and have another play in the next
> couple of weeks, but it *is* hard to be motivated when it seems no one
> else is doing the same.

I feel you should be motivated by your *need* for this feature. If you
or anybody else want it, then join in and help us make it work. Again
and again I feel like I'm trying to develop a feature that people say is
important, but nobody is really that much interested anyway.

> What about the idea which has been expressed a few times: make a copy
> of a 'well known' repo using \r\n line endings and ask a few
> non-windows people to attempt to productively use that clone with the
> EOL extension mapping back to \n line endings locally? If nothing
> else, it would help break-down the perception that only Windows users
> can help move this forward.

Well, I've taken the first step today by switching my thesis to CRLF in
the repository! So far nothing has broken :-)

> Taking this further, we could create a 'trial' Python repository in
> this format, and ask the Python community to "sign off" on the hg
> functionality while it is in this format (as no one should even notice
> this is the case once EOL is enabled). The "real" repository
> conversion would then obviously choose \n line endings, but this first
> step will probably weed out a number of practical problems.

I think that is a good idea, both to test the conversion itself and to
test the eol extension.

-- 
Martin Geisler

VIFF (Virtual Ideal Functionality Framework) brings easy and efficient
SMPC (Secure Multiparty Computation) to Python. See: http://viff.dk/.