cover "SET X-MICROSOFT-DISALLOW-COUNTER"

[pbiering]

relates to https://github.com/Kozea/Radicale/issues/1524

[da4089]

Waiting on example file from @nl06857

[nl06857]

Attached is a meeting invitation as created by the online meeting scheduling service 'datumprikker.nl' (it is available in Dutch and English). The .ics file is accepted by the iCalendar Validator as valid. Some more information:

• RFC 5545 states that unrecognised x-params must be ignored

• There is something strange in this .ics file in that it contains not just an x-param, but that it is preceded by 'set'. This is something that isn't defined in the RFC, nor in the Microsoft documentation (https://learn.microsoft.com/en-us/openspecs/exchange_server_protocols/ms-oxcical/8fcaed94-cfc7-4881-83cd-36317969711d). Thus, the .ics file is not a clean object with properties. I believe that the scheduling service is to blame. Nevertheless, the iCalendar Validator does accept it, as other calendar programs do…

Tjoelker-reunie.ics.tar.gz

[da4089]

See also this setting mentioned here

https://stackoverflow.com/questions/10426661/sending-outlook-meeting-requests-without-accept-decline-option

https://learn.microsoft.com/en-us/openspecs/exchange_server_protocols/ms-oxcical/8fcaed94-cfc7-4881-83cd-36317969711d

I have pinged the Validator maintainer to ask why it is accepted there: perhaps they will have some useful information.

Otherwise, I suspect it was originally a Microsoft bug, and people have just interoperated with it ever since.

I'm increasingly of the view that vobject should perhaps support an optional "accept known oddities" mode when parsing: I'm reluctant to make it the default, but for things like Outlook, refusing to parse them is kinda silly.