On Mon, Sep 11, 2017 at 10:50 AM, Bryan Richter <br...@snowdrift.coop> wrote:
On 09/11/2017 03:16 PM, Stephen Michel wrote:

On September 11, 2017 5:19:29 AM EDT, Bryan Richter <br...@snowdrift.coop> wrote:
 == Last week.

 Code changes: some updates and fixes to the snowdrift project page,
 thanks to Iko.

 == This month.

 My goal for this month is to finish the website and crowdmatch
 mechanism for Alpha. Maybe a bit optimistic. :) But we seem close.
 The mechanism is functional; it just needs testing. The website is
functional; it just needs testing. What is mainly missing is a little bit of practice: practice restoring backups, practice analyzing logs,
 practice deploying breaking changes -- that sort of thing.

 There's a lot more we *could* do, but there always will be.

 == This week.

- Start practicing and figuring out where things are the most broken
  or flimsy.

- Do some system administration work. Hard drives are filling up, and
  not much is documented.

- Coordinate with other components of the project. Make sure we have
  the bones of a post-Alpha plan.

 - Encourage people to prod the website and report UI problems.

 == "What can I do?"

 - Write tests! Or ask me how. It needs documenting, anyway.

 - Get the haskell-stripe packages back into Stackage.

 - Ask questions about open issues.


 When we reach alpha, maybe it's time to run our first crowdmatch?
 It'll be below our normal minimum, thus we'll be paying much more in
 stripe fees than we'd like, but the point would be to find pain
 points in the process, not recieve significant sums of money.

My answer to your question is both "yes" and "no". Yes we should do
the first crowdmatch before October 1. But we won't charge any fees.

Recall that there are two processes: crowdmatching and payment
processing. A crowdmatch transaction updates the amount that a
patron "owes" to a project. Separately, we use Stripe to settle up
between patrons and projects. This is just another way of saying that
donations roll over to the next month if the amount owed isn't enough
to minimize fees.

This raises an important point: the significance of these numbers needs
to be made clear on the website: current pledge value, current amount
"owed", and total amount paid to the project to date.

Now, should we do special-case payment processing that ignores our
promise about maximum fees? I definitely think not. :) Right now we'd
be talking about a 30 cent charge on a 6 cent donation. Bleh

Ah, let me update my verbiage: I meant to propose that we run a crowdmatch for all months that people have been pledging with real credit card numbers, then run a payout on that total. So it'd be more like a 35 cent charge on a 45 cent donation (for me anyway, newly pledged patrons will be closer to what you said).

With 67 patrons that's just under $25 in fees. So, let's not do that, but maybe let's trigger a payout for just a handful of accounts (those of team members), to verify that it will work in The Real World (tm), rather than running into issues later that might affect people with less stake in the project. :)
Dev mailing list

Reply via email to