FYI: pp-lib configuration update
I just updated the EDG configuration of the pp-lib on both the trunk and the release branch at the request of EDG. EDG's newest front-end preprocessor goes into a Microsoft-emulation mode when the compiler proper is set to run in Microsoft-emulation mode. Because of that, the pp-lib will fail without the configuration choosing the MSVC configuration when EDG is running in Microsoft-emulation mode--hence the change. Regards, Paul Mensonides
Paul, Paul Mensonides wrote:
I just updated the EDG configuration of the pp-lib on both the trunk and the release branch at the request of EDG.
The release branch is frozen since March 1st. The idea being no changes and if there is a pressing need please run them by me first. I guess for now we have to hope this one does not break anything. Thomas -- Thomas Witt witt@acm.org
-----Original Message----- From: boost-bounces@lists.boost.org [mailto:boost-bounces@lists.boost.org] On Behalf Of Thomas Witt
I just updated the EDG configuration of the pp-lib on both
Paul Mensonides wrote: the trunk
and the release branch at the request of EDG.
The release branch is frozen since March 1st. The idea being no changes and if there is a pressing need please run them by me first.
Yes, I know. I just didn't know who I was supposed to run them by. Sorry.
I guess for now we have to hope this one does not break anything.
It shouldn't, but if it does, it is a trivial rollback. Regards, Paul Mensonides
Paul Mensonides wrote:
-----Original Message----- From: boost-bounces@lists.boost.org [mailto:boost-bounces@lists.boost.org] On Behalf Of Thomas Witt
I just updated the EDG configuration of the pp-lib on both
Paul Mensonides wrote: the trunk
and the release branch at the request of EDG.
The release branch is frozen since March 1st. The idea being no changes and if there is a pressing need please run them by me first.
Yes, I know. I just didn't know who I was supposed to run them by. Sorry.
Sorry, it's a little bit weird. If you knew about the freeze you could have posted a patch saying "I don't know who's supposed to approve this, but there's a patch".
I guess for now we have to hope this one does not break anything.
It shouldn't, but if it does, it is a trivial rollback.
Depends on the meaning of "trivial". A commit that touches a library that many other libraries depends on might trigger rebuild of a lot of tests, and that would take a lot of time. So, assuming a commit breaks something, we spend a couple of days running tests after the change, notice breakage, trivially rollback it, and wait a couple days for new results, which easily translates into 4 days, during which time results of other changes are not updated as often as desired. - Volodya
participants (3)
-
Paul Mensonides -
Thomas Witt -
Vladimir Prus