Fri, Aug 16
Thu, Aug 15
Embarrassing bug with an easy fix, so I'm pulling it into sprint.
Thanks for the report, @DannyS712. Wikimedia Fundraising uses this page in a very particular way - we source it as an image from the donation Thank You page (hosted on donate.wikimedia.org) so we can set a cookie on the wikipedia.org domain that we check to hide fundraising banners for about a year. So the page actually should render as a minimal image, and set a cookie depending on the querystring parameters.
OK, this is up and ready to test at https://payments.wikimedia.org/index.php?title=Special:AstroPayGateway&appeal=JimmyQuote&payment_method=bt&recurring=&uselang=en&language=en¤cy=INR&amount=140&country=IN&ffname=astropay-in
OK, the Netbanking icon is now a bit smaller, with text 'Netbanking' below.
Wed, Aug 14
Tue, Aug 13
Fri, Aug 9
I just ran two silverpop exports against the dev_civicrm database (so it wouldn't change in between runs), one with the pymysql library and one with the old mysqldb library. They produced identical files. I thing the pymysql change was a red herring, and we should be able to restore it Monday.
That RFC leaves till the last step fully blocking 3rd party images and iframes, but the report-only CSP currently being served does report on those. So if they go to step 6 (actually enforcing CSP on the main wikis) without adding exceptions to img-src and child-src they'll break the RML form.
The RFC for implementing CSP in core has a pretty clear set of steps: https://www.mediawiki.org/wiki/Requests_for_comment/Content-Security-Policy
The grandparent task for CSP in core is T28508
Thu, Aug 8
Currently running an export on frdev against the dev_civicrm and dev_drupal databases, with the pymysql library. Since those aren't moving targets, I should be able to save the file from today's export and run again without the pymysql library tomorrow and diff the exported files.
Looks like this might be nice to do on payments-wiki too, for cases when the amount box is editable
Replied to that email on July 25th with details of our hosted page customization and a sample setup API call. Matthew replied the 28th saying "myself or my colleague Patrick will follow after we further assess this."
The back end has a function that takes currency code, amount, and locale, and formats them according to the customary display of that currency in that locale, with an appropriate symbol, symbol placement, and decimal points.
Wed, Aug 7
Maybe also update the code path that imports recurring payments so that it fixes the status of any canceled recurring contribution when a new payment comes in?
So the correct fix for this is slightly more involved than just a text change - we should make that max amount a parameter and fix the places where we use it. This way it'll be correct for e.g. DLocal forms where we have a smaller max donation.
(Or rather, +2 as soon as it's moved out of draft namespace)
Yep, the note at the top of the 'Mapping from LandingPageImpression events' section looks like good clarification.
Let's document this here: https://www.mediawiki.org/wiki/Fundraising_tech/Update_Thank_You_Emails
derp, never mind, I just read the last line of the task description
Is this on donatewiki? On payments-wiki, if I don't specify an amount and currency in the URL, I can put in 11,000 USD and when I click submit the validation allows the amount (while warning on other empty fields)
Wed, Jul 31
We did all we could with scheduling - the next step is breaking up the fredge job: T227805: Break up multi-queue drush jobs
Oh shoot, I forgot '--installed' when I ran 'apt list' on civi1001. The package is not actually installed up there. @Jgreen, can we install the python-pymysql deb package on hosts configured to get the 'tools' project deployed?
Tue, Jul 30
@TSkaff This just went out along with some other Civi fixes.
Fri, Jul 26
Thu, Jul 25
@EMartin just FYI, we also have tags for each payment processor integration: FR-Adyen, FR-Ingenico, FR-Amazon, FR-dlocal, FR-Paypal, and FR-Paypal-ExpressCheckout
@Cstone can we close this?
Verified that pymysql is already installed on the cluster where we'll need it.
Heh, yeah, I'm just going back and estimating things without story points.