Disk cache: Reference CL for the implementation of file format version 3.

Disk cache: Reference CL for the implementation of file format version 3.

This is not to be commited.

[gavinp, 2013-06-03 20:45:05]

This is really difficult to examine; this patch is a mixture of things that have
already landed in trunk and the work on the v3 cache.

[rvargas (doing something else), 2013-06-03 21:16:51]

On 2013/06/03 20:45:05, gavinp wrote:
> This is really difficult to examine; this patch is a mixture of things that
have
> already landed in trunk and the work on the v3 cache.

This is the (almost) current state of V3. It doesn't make sense to update the
baseline before https://chromiumcodereview.appspot.com/14991008/ lands.

[gavinp, 2013-06-04 00:51:22]

On 2013/06/03 21:16:51, rvargas wrote:
> On 2013/06/03 20:45:05, gavinp wrote:
> > This is really difficult to examine; this patch is a mixture of things that
> have
> > already landed in trunk and the work on the v3 cache.
> 
> This is the (almost) current state of V3. It doesn't make sense to update the
> baseline before https://chromiumcodereview.appspot.com/14991008/ lands.

OK, so while reviewing https://chromiumcodereview.appspot.com/14991008/ , I have
come to this CL, and I found it was difficult to understand for these reasons.

If you can't update it before that lands, that's fine. I'll create a private
version that contains updates, and compare to it. I will share that with you to
confirm I dealt with conflicts correctly.

Thanks.

[rvargas (doing something else), 2013-06-04 02:18:51]

On 2013/06/04 00:51:22, gavinp wrote:
> On 2013/06/03 21:16:51, rvargas wrote:
> > On 2013/06/03 20:45:05, gavinp wrote:
> > > This is really difficult to examine; this patch is a mixture of things
that
> > have
> > > already landed in trunk and the work on the v3 cache.
> > 
> > This is the (almost) current state of V3. It doesn't make sense to update
the
> > baseline before https://chromiumcodereview.appspot.com/14991008/ lands.
> 
> OK, so while reviewing https://chromiumcodereview.appspot.com/14991008/ , I
have
> come to this CL, and I found it was difficult to understand for these reasons.
> 
> If you can't update it before that lands, that's fine. I'll create a private
> version that contains updates, and compare to it. I will share that with you
to
> confirm I dealt with conflicts correctly.
> 
> Thanks.

I don't think there's a need to pay too much attention to the details of this CL
in order to go through 14991008. After all, the whole point of making that other
CL without modifying code is avoid reviewing everything in one pass. In fact,
the idea of posting a CL with the whole thing was meant to help landing
subsequent changes, not really 14991008. To be totally honest, I was actually
thinking on TBRing that cl from the start given that it meant a purely
mechanical review and had zero impact in the current code but I decided to use a
regular review given that there was no rush to land anything.

As for seeing stuff here that has already landed, only two very isolated pieces
have landed: tracing and stats, and they are not part of 14991008 and are
included here only for completeness.

[gavinp, 2013-07-20 02:41:40]

Thanks for the upload updating this Ricardo.

Is this upload a rebase only, or does it reflect changes in the v3 goal?

[Randy Smith (Not in Mondays), 2013-07-20 16:33:01]

Ricardo: Just so I can create a checkout with this patched in if I feel the urge
while reading it, what trunk revision is this based on?

[Randy Smith (Not in Mondays), 2013-07-21 01:06:31]

On 2013/07/20 16:33:01, rdsmith wrote:
> Ricardo: Just so I can create a checkout with this patched in if I feel the
urge
> while reading it, what trunk revision is this based on?

Whoops, never mind--I didn't realize you used SVN so that the revisions were
there in the raw diff.