Re: Why do we check the base checksum so often?

On Mon, Feb 6, 2012 at 9:02 AM, C. Michael Pilato <cmpilato_at_collab.net> wrote:
> On 02/04/2012 01:10 PM, Hyrum K Wright wrote:
>> On Sat, Feb 4, 2012 at 10:59 AM, Julian Foad <julianfoad_at_btopenworld.com> wrote:
>>> I'm not quite sure I fully follow you at the moment, so I'm not sure if my reply is on the right track at all, but it's really sounding like you're up against a mis-match of responsibilities between Ev1 which sends deltas according to particular rules and Ev2 which is designed to be wrapped inside a driver-receiver pairing that knows privately how to deltify and recover to full text in any way it wants to.  The shims obviously need to convert from the Ev2 deltification back (via a full text intermediary if necessary) to what Ev1 expects.
>>
>> What's driving this discussion is this:  Up until this point in the
>> Ev2 shims we've been supplying a NULL base checksum for apply
>> textdelta, which the receivers have dutifully ignored.  However, when
>> the Ev2 shims attempt to be honest about the fact that they are using
>> the empty stream for the text base, the receivers start complaining,
>> because that's not what they expected---even though the end result is
>> the same.  In essence, all these checks are returning false positives,
>> which is extremely unpleasant.
>>
>> I don't know that there is an easy way around this, since by the time
>> we're translated from Ev2->delta-editor, we don't have the original
>> text base, or its checksum, available to us.  We have the full text,
>> which is the reason the new text base is the empty stream: it's the
>> only one we need.
>>
>> Does that make any sense?
>
> Yeah, sounds like we're in a tough spot here.  The checksums in Ev1 exist
> only as sanity checks -- they definitely are NOT the primary selection
> mechanism for the editor implementation's base text.
>
> I assume we still have a valid checksum to pass to close_file() (the
> checksum of the post-edit fulltext for that file), right?  It's just the
> pre-edit base checksum that's the problem?

Correct.

r1241097 relaxes the checks by special-casing the checksum of the
empty stream (as discussed elsethread). This addresses the immediate
issue, and I think the generalized case can be punted toward the
individual ra-layers long-term.

-Hyrum

-- 
uberSVN: Apache Subversion Made Easy
http://www.uberSVN.com/