What's happenning to the regression results?

Hi all, What is the reason of this sudden increase in the number of failures for a lot of libraries? Has something underlying been broken? Was some platforms status changed from "broken" to "not broken", etc. Or is it just that a lot of new failures have been introduced? Thanks, Arkadiy

Arkadiy Vertleyb wrote:
Hi all,
What is the reason of this sudden increase in the number of failures for a lot of libraries? Has something underlying been broken? Was some platforms status changed from "broken" to "not broken", etc. Or is it just that a lot of new failures have been introduced?
Good question: my failure count has gone from a healthy zero up to 20. Other people have faired a lot worse :-( Most of the failures relate to something in MPL that's completely broken the Borland results, if I could figure out what it was that changed, I would suggest reverting it. John.

Hi, John Maddock wrote:
Arkadiy Vertleyb wrote:
Hi all,
What is the reason of this sudden increase in the number of failures for a lot of libraries? Has something underlying been broken? Was some platforms status changed from "broken" to "not broken", etc. Or is it just that a lot of new failures have been introduced?
Good question: my failure count has gone from a healthy zero up to 20. Other people have faired a lot worse :-(
AFAICS it is not clear yet what exactly the problem is. What is clear is that we will have a better idea once all testers switched to build v2. What is hopefully happening any day now. Please hang tight for now. Thomas -- Thomas Witt witt@acm.org

Arkadiy Vertleyb writes:
What is the reason of this sudden increase in the number of failures for a lot of libraries? Has something underlying been broken?
Yes, I inadvertently broke MPL on Borland: http://article.gmane.org/gmane.comp.lib.boost.testing/3514 -- Aleksey Gurtovoy MetaCommunications Engineering
participants (4)
-
Aleksey Gurtovoy
-
Arkadiy Vertleyb
-
John Maddock
-
Thomas Witt