>>> "AdG" == Adriaan de Groot writes: Hi Adriaan, > Hi Uwe, > On Monday, 10 August 2020 18:02:34 CEST Allan Sandfeld Jensen wrote: >> On Montag, 10. August 2020 09:12:12 CEST Martin Steigerwald wrote: >> > Uwe Brauer - 10.08.20, 09:02:15 CEST: >> > > The issue concerns an announcement by the gmail developers to drop the >> > > option to connect to their IMAP service via name and password, known >> > > as less-secure-apps. The main email reader of GNU Emacs, called gnus >> > > relies on this method. Someone on the GNU emacs developer list >> > > suggested that kmail solved the problem and is able to connect to the >> > > GMAIL imap server using their OAuth2 method. >> >> I would recommend anyone concerned with privacy rights to not use Gmail. >> OAuth2 is just a way for Google to force people to use Google Chrome for >> logging into google services. Thanks for your answer. > There's an intransparent and arbitrary vetting process in place, where Google > approves an application's access to Google services. > I blogged about the issue, on behalf of Dan Vratil who was doing the actual > work of dealing with Google, > https://euroquis.nl/kde/2020/05/12/google.html > https://euroquis.nl/kde/2020/01/13/google.html > I'm sure Dan could describe the process in more detail, but it (slightly > cynically) comes down to this: > - you will need a Google account *for the application developer*, > - this application needs some API keys, > - this application needs a privacy policy, > - the application + policy must be submitted to the arbitrary process, > - wait anywhere from 3 days to 6 months, > - possibly the application will be allowed to connect to Google services. This is really bad, deadline is in February. So there is quite a bit of work to do and then it all depends on the mercy of the «Dark Lord». I am not sure what RMS is going to say about this. Then there is a python based solution published on https://luxing.im/mutt-integration-with-gmail-using-oauth/ Not sure how relevant that could be. Regards Uwe > There is code in KDE-PIM somewhere that does the actual API-key-wrangling, > that's not really the issue. > [ade]