Convert Linux internal time format to Windows 1601 epoch.

Convert Linux internal time format to Windows 1601 epoch.

This is only used on Linux, Mac has separate time code that will have to be
changed in a similar fashion.  Although we represent time internally starting
from 1601, there are still things like time explosion that will not work before
the year 1900.  This limitation is the same as it was previously.

BUG=14734
Closed by mark on 2009-10-24.  This was rolled into http://codereview.chromium.org/173296, landed by brettw at r24417 on 2009-08-25.

[willchan no longer on Chromium, 2009-06-30 14:02:22]

On 2009/06/30 10:19:16, Dean McNamee wrote:
> 

Didn't Brett say we had to write conversion code for the databases?  If you
commit the change as is, what will happen when we read the databases from disk? 
Will the history have really funky dates from the 1600s?

[willchan no longer on Chromium, 2009-06-30 14:04:59]

On 2009/06/30 14:02:22, willchan wrote:
> On 2009/06/30 10:19:16, Dean McNamee wrote:
> > 
> 
> Didn't Brett say we had to write conversion code for the databases?  If you
> commit the change as is, what will happen when we read the databases from
disk? 
> Will the history have really funky dates from the 1600s?

Ah woops, I just read the bug thread now.  I'm not sure if we should just let
the profile be messed up.  I defer to Brett.

[Dean McNamee, 2009-06-30 14:06:12]

The conversion code is related, and probably still needs to be done.  However,
we need to make sure that there isn't other problems, and that this change would
work.  I was hoping for a review of this, running it through all of the tests,
and then trying to figure out the profile upgrade.

On 2009/06/30 14:04:59, willchan wrote:
> On 2009/06/30 14:02:22, willchan wrote:
> > On 2009/06/30 10:19:16, Dean McNamee wrote:
> > > 
> > 
> > Didn't Brett say we had to write conversion code for the databases?  If you
> > commit the change as is, what will happen when we read the databases from
> disk? 
> > Will the history have really funky dates from the 1600s?
> 
> Ah woops, I just read the bug thread now.  I'm not sure if we should just let
> the profile be messed up.  I defer to Brett.

[brettw, 2009-06-30 14:25:25]

This seems fine so far. I do think we should write migration code for Cookies
and History to avoid messing people up. Chrome could always migrate history from
the earliest versions during its development, and we have many more Linux & Mac
users now than we had Windows users then.

[Dean McNamee, 2009-06-30 14:26:50]

Ok, I'll coordinate with getting the mac-side of the change and making sure the
tests all still pass.

Any takers for the migration code?

On 2009/06/30 14:25:25, brettw wrote:
> This seems fine so far. I do think we should write migration code for Cookies
> and History to avoid messing people up. Chrome could always migrate history
from
> the earliest versions during its development, and we have many more Linux &
Mac
> users now than we had Windows users then.

[Mark Mentovai, 2009-06-30 15:36:17]

Early unsolicited fyi feedback because I'm friendly and helpful

I know it's "in progress" but the stuff you've already done looks good.

http://codereview.chromium.org/150097/diff/1/2
File base/time_posix.cc (right):

http://codereview.chromium.org/150097/diff/1/2#newcode166
Line 166: #error No usable tick clock function on this platform.
You'll trip this on the Mac now.

Can you just leave this function as it was (and put its #include back)?

[willchan no longer on Chromium, 2009-06-30 16:28:47]

Can you also add a test to CookieMonsterTest_TestCookieDeletion that tests with
expires=Thu, 01 Jan 1970 00:00:00 GMT?

As far as the database upgrading, doesn't that have to be done in the same
changelist?

http://codereview.chromium.org/150097/diff/1/3
File base/time_unittest.cc (right):

http://codereview.chromium.org/150097/diff/1/3#newcode152
Line 152: exploded.day_of_week = 0;  // Should be unusued.
s/unusued/unused/

[Dean McNamee, 2009-07-02 08:51:23]

Any news on the Mac side?  Linux seems to be working fine, and passes all tests.

On 2009/06/30 16:28:47, willchan wrote:
> Can you also add a test to CookieMonsterTest_TestCookieDeletion that tests
with
> expires=Thu, 01 Jan 1970 00:00:00 GMT?
> 
> As far as the database upgrading, doesn't that have to be done in the same
> changelist?
> 
> http://codereview.chromium.org/150097/diff/1/3
> File base/time_unittest.cc (right):
> 
> http://codereview.chromium.org/150097/diff/1/3#newcode152
> Line 152: exploded.day_of_week = 0;  // Should be unusued.
> s/unusued/unused/

[Dean McNamee, 2009-07-06 11:33:11]

Amanda, can you give this a shot on the Mac side?

[Amanda Walker, 2009-07-06 11:50:12]

On 2009/07/06 11:33:11, Dean McNamee wrote:
> Amanda, can you give this a shot on the Mac side?

Yes, I can give this a shot later this morning.

[Dean McNamee, 2009-07-13 14:51:56]

ping

On 2009/07/06 11:50:12, Amanda Walker wrote:
> On 2009/07/06 11:33:11, Dean McNamee wrote:
> > Amanda, can you give this a shot on the Mac side?
> 
> Yes, I can give this a shot later this morning.

[Amanda Walker, 2009-07-13 15:07:13]

Oops, this fell off my radar.  I'll take a look at it in a few.

On Mon, Jul 13, 2009 at 10:51 AM, <deanm@chromium.org> wrote:
> ping
>
> On 2009/07/06 11:50:12, Amanda Walker wrote:
>>
>> On 2009/07/06 11:33:11, Dean McNamee wrote:
>> > Amanda, can you give this a shot on the Mac side?
>
>> Yes, I can give this a shot later this morning.
>
>
>
> http://codereview.chromium.org/150097
>

[Dean McNamee, 2009-07-14 14:33:06]

Now with Amanda's patch for the Mac.

Whose up for the database upgrading?

On 2009/07/13 15:07:13, Amanda Walker wrote:
> Oops, this fell off my radar.  I'll take a look at it in a few.
> 
> On Mon, Jul 13, 2009 at 10:51 AM, <deanm@chromium.org> wrote:
> > ping
> >
> > On 2009/07/06 11:50:12, Amanda Walker wrote:
> >>
> >> On 2009/07/06 11:33:11, Dean McNamee wrote:
> >> > Amanda, can you give this a shot on the Mac side?
> >
> >> Yes, I can give this a shot later this morning.
> >
> >
> >
> > http://codereview.chromium.org/150097
> >

[Dean McNamee, 2009-07-22 12:02:10]

What's the verdict, shall I commit?

On 2009/07/14 14:33:06, Dean McNamee wrote:
> Now with Amanda's patch for the Mac.
> 
> Whose up for the database upgrading?
> 
> On 2009/07/13 15:07:13, Amanda Walker wrote:
> > Oops, this fell off my radar.  I'll take a look at it in a few.
> > 
> > On Mon, Jul 13, 2009 at 10:51 AM, <deanm@chromium.org> wrote:
> > > ping
> > >
> > > On 2009/07/06 11:50:12, Amanda Walker wrote:
> > >>
> > >> On 2009/07/06 11:33:11, Dean McNamee wrote:
> > >> > Amanda, can you give this a shot on the Mac side?
> > >
> > >> Yes, I can give this a shot later this morning.
> > >
> > >
> > >
> > > http://codereview.chromium.org/150097
> > >

[brettw, 2009-08-13 05:03:09]

I added some comments to the bug:
  http://code.google.com/p/chromium/issues/detail?id=14734
I don't think this should be accepted without migration.