timestamp - does it represent the beginning, end or middle?

[garemoko]

Does the timestamp denote the start of the experience, the end of the experience or some part in the middle?

Currently we're saying that it the time when the statement was generated, but should it be more linked to the experience than that? What if we're generating statements about events that happened a while ago?

[bscSCORM]

I suggest we clarify it as the end of the experience described in the statement. That fits with the semantics of past-tense statements. Once could still note the beginning by describing the experience of starting the larger experience... "I started this". Started doesn't have to be defined in the spec for that to work, for the "end of the experience described in the statement" to be chosen, one has to understand the statement, including the verb.

[garemoko]

Just thinking some more about this, you could well have experiences that you don't know the end point of the experience at the point the statement was issued.

I'm erring more towards saying it should be the start point because you will always know that. You could use a completed verb or something similar to denote the end point.

Andrew

bscSCORM notifications@github.com wrote:

I suggest we clarify it as the end of the experience described in the statement. That fits with the semantics of past-tense statements. Once could still note the beginning by describing the experience of starting the larger experience... "I started this". Started doesn't have to be defined in the spec for that to work, for the "end of the experience described in the statement" to be chosen, one has to understand the statement, including the verb.

—
Reply to this email directly or view it on GitHub.

[Vitzkrieg]

I believe tracking the time at the end with the past-tense approach is more viable. You'll want to create statements tracking the progress through the entire object/process: Joe "opened" the ebook. Joe "read" through page 12 of the ebook. Joe "closed" the ebook. Joe "opened" the ebook. Joe "finished" the ebook. The timestamps would correlate with the creation of the statement, which is at the end of the action. Actions like "opened" and "finished" would just be single moments in time - no true start or end).

[fugu13]

I think a reasonable approach would be that the timestamp (at least as a
matter of best practice, right now there are no normative implications in
the API) represents the end of the event if there's a duration, but if
there isn't a duration it represents sometime during the span of the
duration (at least, we hope; much harder to be sure if the client didn't
send it). As Andrew says, sometimes the most convenient point to issue a
statement isn't the end of an experience, but you want to say something
about the whole experience.

On Fri, Jun 21, 2013 at 2:16 PM, Vitzkrieg notifications@github.com wrote:

I believe tracking the time at the end with the past-tense approach is
more viable. You'll want to create statements tracking the progress through
the entire object/process: Joe "opened" the ebook. Joe "read" through page
12 of the ebook. Joe "closed" the ebook. Joe "opened" the ebook. Joe
"finished" the ebook. The timestamps would correlate with the creation of
the statement, which is at the end of the action. Actions like "opened" and
"finished" would just be single moments in time - no true start or end).

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/337#issuecomment-19841288
.

[garemoko]

Consensus seems to be to that the timestamp represents the end of the experience if a duration is specified. @fugu13 's suggestion that it's considered to be less precise if a duration is not specified makes sense. I'll draft a PR now.