Attendees
Stephen
Clive
Tullio
Larry
Hubert
Tong
Michael Wong
Teleconference/Webex
Meeting Times:
13 February 2017: 2100-2300 UTC
IMPORTANT:
|
||||
2018 |
||||
|
|
|
|
|
Pre-mtg 55 |
01/11/18 |
|
|
|
#54 |
TBD September 2018 |
CSA Toronto, Canada with SC 22 |
|
|
#53 |
15-16/06/18 |
With WG 9 and Ada Europe |
|
|
Pre-mtg-53 |
Teleconference |
|
|
|
#52 |
TBD April 2018 |
Czech Republic with C |
|
|
Pre-mtg 52 |
TBD March 2018 |
|
||
#51 |
22-23 January 2018 |
Phoenix, AZ |
||
|
||||
2017 |
||||
pre-mtg-51 |
20/11/17 |
Teleconference (UTC 2000, 2 hr) |
|
|
post-mtg-50 |
16/10/17 |
Teleconference (UTC 2000, 2 hr) |
|
|
#50 |
17-18 August 2017 |
BSI London (with SC 22 Plenary) |
|
|
#49 |
19-20 June 2017 |
Vienna, Austria with Ada Europe(2 day) |
|
|
post-mtg-48 |
15/05/17 |
Teleconference (UTC 2000, 2 hr) |
|
|
#48 |
6-7 April 2017 |
IBM Markham, Canada (2 day) |
|
|
pre-mtg-48 |
Mar 6, 17 |
Teleconference (UTC 2100, 2 hr) |
||
Post-mtg-47 |
13/02/17 |
Teleconference (UTC 2100, 2 hr) |
Dedicated topic, Templates and Generics in Part 1, capturing C++ issues. |
Document N0689, clause 40 Templates and Generics.
At the present time, clause 40 reflects the view of templates and generics from a perspective similar to Ada. C++ has a different model of generics. This meeting is to discuss the different models of templates and generics, and determine how to update 6.40 to reflect a broader view and to capture what vulnerabilities exist from the C++ model.
Since attendees from a WG 21 background are not familiar with the document, our style of writing and constructing vulnerability descriptions, nor the model presented in clause 6.40, we defer discussion until the next WebEx session or meeting 48.
Document N0691 is the first draft of a potential C++ Part. It is largely a remapping of Part 3 C. C++ participants are requested to review the document and suggest ways to sharpen the C-specific discussions, to capture ways that C++ features can mitigate vulnerabilities, and to help capture C++ vulnerabilities that do not exist in C.
Section 3 Terms and definitions – We agree to leave the definitions in place and review them later once significant portions of the vulnerabilities have been worked on.
AI 47-12 – Clive – check that updates made to Part 1 6.4 floating point to see if captured in Part 3.
We discuss N0691 and begin to make some changes, but terminate that effort when some of the participants state that they do not think that the larger C++ standardization committee will accept the approach taken.
Suggestion – create a C++ annex that reflects the vulnerabilities in C++ without restating all of the vulnerabilities in the C subset. At the same time this annex needs to acknowledge that the vulnerabilities exist, and recommend ways to ensure that developers are not inadvertently hitting those vulnerabilities.
Team of Paul, Michael and Hubert to try a tack of pushing the C-specific guidelines to Part 3 or Part 1 and running the approach through the WG 21 leadership. Clive to continue working through the document to capture the differences between C and C++ w.r.t vulnerabilities.