1098022 - Videos that don't start at 0 are treated as a new segment by TimestampsFuzzyEqual

Status: RESOLVED FIXED

(Core :: Audio/Video, defect)

Description

[Matt Woodrow (:mattwoodrow)]

Attachment: initial-timestamp

We should initialize mLastEndTimestamp to -1 to be able to differentiate between 'no previous data' and an actual time of 0.

Comment 1

[cajbir (:cajbir)]

We need an Option or Maybe type...

Comment 2

[Matt Woodrow (:mattwoodrow)]

http://mxr.mozilla.org/mozilla-central/source/mfbt/Maybe.h#83

Comment 3

[Anthony Jones (:ajones, :kentuckyfriedtakahe, :k17e)]

Comment on attachment 8521782 [details] [diff] [review]
initial-timestamp

Review of attachment 8521782 [details] [diff] [review]:
-----------------------------------------------------------------

I'm not super keen on -1 as the universal sentinel and would tend to use a constant like NO_TIME=-1 which is slightly less horrible. We may be able to just check if the Decoder actually has any data in it instead. I'm OK with landing it as-is if there isn't a way to check TrackBuffer being empty.

Comment 4

[cajbir (:cajbir)]

(In reply to Anthony Jones (:kentuckyfriedtakahe, :k17e) from comment #3)
> Comment on attachment 8521782 [details] [diff] [review]
> initial-timestamp
> 
> Review of attachment 8521782 [details] [diff] [review]:
> -----------------------------------------------------------------
> 
> I'm not super keen on -1 as the universal sentinel and would tend to use a
> constant like NO_TIME=-1 which is slightly less horrible. We may be able to
> just check if the Decoder actually has any data in it instead. I'm OK with
> landing it as-is if there isn't a way to check TrackBuffer being empty.

Can we use mfbt/Maybe.h?

Comment 5

[Matt Woodrow (:mattwoodrow)]

Attachment: initial-timestamp v2

Comment 6

[Matt Woodrow (:mattwoodrow)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/b52b8019d49a

Comment 7

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/b52b8019d49a