Tapestry

Hi all,

Despite our launch in March, Open Humans hasn't yet made a concerted effort
to invite participants in its three programs to the site. That's because it
was immediately clear that the process was clearly confusing for users and
needed improvement. In months since, I've spent a fair amount of time
creating and iterating mock-ups, understanding more about the tech itself,
and creating a standard model for how we implement data import processes.

Here is a page that summarizes the change to the flow as it would occur for
PGP Harvard:
https://personalgenomes.mybalsamiq.com/projects/update-pgp/prototype/Updating%20PGP%20Harvard%20connection?key=30090b9af29d405942c7403dec78fb37c4ff5b62

The process is designed to make the data import process – from a user's
perspective – generally the same across diverse sources. That is, the
process from a "data push" source (e.g. PGP Harvard) is made as similar as
possible to the proces for a "data pull" source (e.g. RunKeeper), despite
technical differences between the two.

I'd like to invite all PGP Harvard staff to read over the current and new
flow. Please feel free to raise concerns or suggestions!

To accomplish this, the PGP Harvard website would need to change in the
following respects:

- The wording of the page on PGP Harvard's site should change to reflect
the new process. I've suggested new wording in the mock-up for that screen
<https://personalgenomes.mybalsamiq.com/projects/update-pgp/prototype/New+-+PGP+share+with+OH?key=30090b9af29d405942c7403dec78fb37c4ff5b62>
.

- Technical process after a user finalizes on the Open Humans site
<https://personalgenomes.mybalsamiq.com/projects/update-pgp/prototype/New+-+OH+finalize+data+import?key=30090b9af29d405942c7403dec78fb37c4ff5b62>
    - (same as before) Open Humans will return the user, bearing an OAuth2
code, to the PGP Harvard's Tapestry website
    - (same as before) PGP Harvard's Tapestry website exchanges that for a
token that authorizes pushing data to Open Humans
    - (changed) PGP Harvard's site pushes that data immediately to Open
Humans without further user requests
    - (changed) PGP Harvard immediately redirects the user back to a page
on Open Humans (i.e. the user experience is never to have left the Open
Humans site)

Some of the above are technical changes that needs someone passingly
familiar with the code. Assuming the changes sound good to PGP Harvard, I'd
like to learn whether we have resources to make those changes ­Tapestry –
that is to say, whether Curoverse-associated staff (or other Curoverse
staff) could help make this change.

If it's minor and Curoverse associated folks are willing to do it for free,
then that's wonderful. But we have no desire to see Open Humans take
advantage of others when there remains funding for doing this, please let
us know if that would be helpful! Open Humans could also look into hiring
someone ourselves as a contractor (and would greatly appreciate
recommendations), as we unfortunately we don't have in house expertise for
contributing to a Rails site. (Beau has done a lot for American Gut's
Tornado-based
website <https://github.com/biocore/american-gut-web&gt;, but our technical
expertise is limited to Python and JavaScript.)

Once this flow is improved, I would love to invite all PGP Harvard
participants (using the 3rd party activities notification format) to become
Open Humans members!

Thanks,

Madeleine