Integrating alerts about stuff that didn't happen into the activity stream

[Maya Mandel]

We're trying to streamline a bunch of data sources in our solution into a single, unified activity stream.

Some of the sources generate activities that are essentially alerts the system raises about lack of activity, for example - "didn't complete", "isn't updated", "hasn't received".

How would you think about doing so while staying consistent with the activity stream format - specifically, using "not"?

Thanks,

Maya.

[Erik Wilde]

hello maya.

On 2013-12-17, 9:50 , Maya Mandel wrote:
> We're trying to streamline a bunch of data sources in our solution into
> a single, unified activity stream.
> Some of the sources generate activities that are essentially alerts the

> system raises about _lack_ of activity, for example - "didn't complete",

> "isn't updated", "hasn't received".
> How would you think about doing so while staying consistent with the
> activity stream format - specifically, using "not"?

we are looking at the same kind of "activities" in our environment. but
in the end, the "not *" pattern needs to be triggered somehow, right?
otherwise you could emit such an "activity" every 50ms until something
actually happens... so my design approach so far was to treat those
activities no different that any others, just assuming that they are
triggered by some event (probably just a timer), and then should be
reported (and maybe can be acted upon, to actually complete the
incomplete activity, in which case that completion would be reported).

if there's a better way to do this, i'd be interested to hear about
this, but the simplest solution may be to just treat these things as
activities in their own right.

cheers,

dret.

--
erik wilde | mailto:dr...@berkeley.edu - tel:+1-510-2061079 |
| UC Berkeley - School of Information (ISchool) |
| http://dret.net/netdret http://twitter.com/dret |

[Maya Mandel]

Thank you Dret for your insights - this is really helpful.

So I'll give a concrete example so we can take that as a use-case.

We have an activation process that is comprised of several activities that need to be executed by the user.

If this sequence starts, but isn't completed within 15 minutes (the timer you spoke about), the system generates an activity that "<The user> hasn't completed the activation process".

Now, if I want to use your approach - what's the syntax you'd use to describe this "timeout"?  How'd you construct the sentence?

Thanks again,

Maya.

[Christopher Lakey]

I have found that it is important to separate how you detect an activity (observe, measure, sample, timer, etc) from how you describe it. Instead of thinking about activities detected using a timer as special, simply focus on the semantics of the activity itself.

I believe there are several factors at play in your user signup use-case:

• detecting absence of activity vs observing activity
• stating a negative vs stating a positive
• user as the actor vs system as the actor

Some thoughts on your user signup use-case:

• user abandon signup
• user fail signup
• user delete signup
• system delete signup (actor is actually the signup system)
• system cancel signup

Personal I prefer to think that the system cancelled the signup process due to user inactivity.  It’s also possible that the user cancelled the signup process, so you want to capture the proper actor.  Additionally, you might have different subsystems with different rules that are able to cancel the signup process.  For instance, you might decide that the user is a bot or already has an account.

Cheers,

-Christopher

--
You received this message because you are subscribed to the Google Groups "Activity Streams" group.
To unsubscribe from this group and stop receiving emails from it, send an email to activity-strea...@googlegroups.com.
To post to this group, send email to activity...@googlegroups.com.
Visit this group at http://groups.google.com/group/activity-streams.
For more options, visit https://groups.google.com/groups/opt_out.