Please find attached the fixed R2. 
 
Whoops -- don't do that -- once a P paper is issued it's fixed (like an IETF RFC). This avoids any confusion about which version of a P paper you're referring to...
Please re-issue a D1386R3 if it's a draft (I suspect it is), or P1386R3 if it's not.
 
Roger.
 
-----Original Message-----
From: SG13 [mailto:sg13-bounces@lists.isocpp.org] On Behalf Of Timur Doumler via SG13
Sent: 18 June 2019 16:01
To: sg13@lists.isocpp.org
Cc: Timur Doumler
Subject: Re: [SG13] Thursday's Telecon

Turned out that the whole “code examples” section was completely wrong as the code examples given were still on an old version of the API. So they wouldn’t even compile. Let’s treat that as a glitch in the matrix ;)

 

Cheers,
Timur



On 18 Jun 2019, at 17:50, Timur Doumler via SG13 <sg13@lists.isocpp.org> wrote:

Update:

On 18 Jun 2019, at 17:29, Timur Doumler via SG13 <sg13@lists.isocpp.org> wrote:

7. 7.1/7.3 the while(true); is ihmo UB. Could we make it while(device.has_unprocessed_io()); or similarly check the state of it.

Also correct. Will fix.

I fixed this, however while(device.has_unprocessed_io()); would be the wrong condition if you do async processing (the normal case). Better to use while(device.is_running()). This way the loop will stop as soon as the device becomes unavailable or the user stops it.

Cheers,
Timur

--
SG13 mailing list
SG13@lists.isocpp.org
http://lists.isocpp.org/mailman/listinfo.cgi/sg13