Hi James, I agree these types of breakages, if unintentional and not regular, should be raised elsewhere first.
Given Fae's reluctance to use private correspondence,... Is there a public wiki page which can be used to alert the relevant team to any future breakages, in the first instance? Or can this be managed through Phabricator? an existing tag? Fae, you said you have your own scripts, which you are no longer maintaining due to changes by Google. Is your code in a public repository somewhere? We do not need to use the Google apis for accessing this data. Google allows spreadsheets to be exported as csv. here is the CSV link for the Advanced Permissions data. https://docs.google.com/spreadsheets/d/1DruVc7T9ZqTcfGwFAlxQrBMR4QBSD_DtjpDtGqMAAi0/pub?output=csv With a small script, we could re-publish this dataset as csv into a git repository, and then another script could read the csv and re-publish the data as wikitext onto a Wikimedia site. On Wed, Feb 15, 2017 at 6:44 AM, James Alexander <jalexan...@wikimedia.org> wrote: > On Tue, Feb 14, 2017 at 11:53 AM Fæ <fae...@gmail.com> wrote: > > Usecases are appearing, thanks to whomever is intervening, though in a > narrow column so hard to read. > > Now I can read it, I see that it is out of date. As a test sample, I > JethroBT (WMF) was granted m:admin rights in June, these expired by > August 2016 and were eventually removed by a volunteer steward in > October 2016. Though I JethroBT is an admin on meta right now, this > was via a separate use case dated "42676", which I presume is > November. Could the spreadsheet be properly reviewed and updated > please, including reformatting the date field so it's easy to > understand? > > Pine - yes this process of "WMF Advanced Permissions" includes admin > rights for any WMF website and so by-passes the community procedures. > > Fae > > > Hi Fae, > > As I’ve mentioned on previous occasions when you’ve brought up this > spreadsheet on the mailing list, it occasionally breaks. That was the case > here. If you send me a quick note if you see the issues, we can fix it, as > we did today with the use case query (including make sure that it’s > multiple columns again.) Pointing that out so it can be quickly fixed is > much better done via a private poke that we'll see quickly rather than a > public mailing list post that we may not see until after hours or until > somebody lets us know about it. Obviously if we ignore your emails or > refuse to fix it, then the math changes, and a post to this list makes more > sense. I do not, however, think breakage (or overlooking notes about > breakage) has been a frequent problem over the past couple years (though we > have certainly had a couple breakages). > > The public sheet is up to date to the internal version of the data (which > is done automatically). However, the automated data collection is better at > “adding new” than “removing old.” A member of the team does annual audits > of the data to ensure that defunct entries are removed and that everything > else matches reality. The time for the next one is coming up. > > James > > *James Alexander* > Manager, Trust & Safety > Wikimedia Foundation > > PS: I also fixed the weird date thing you were seeing on some of them... > not sure what caused that (was just a format display thing). > _______________________________________________ > Wikimedia-l mailing list, guidelines at: > https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and > https://meta.wikimedia.org/wiki/Wikimedia-l > New messages to: Wikimedia-l@lists.wikimedia.org > Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, > <mailto:wikimedia-l-requ...@lists.wikimedia.org?subject=unsubscribe> -- John Vandenberg _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l New messages to: Wikimedia-l@lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:wikimedia-l-requ...@lists.wikimedia.org?subject=unsubscribe>
