[serialization] regressions due to recent changes in /libs/serialization/test/tes_tools.hpp
Hello, Recent changes in test_tools.hpp aimed at switching from Boost.Test to boost/detail/lightweight_test.hpp are causing lots of regressions, seemingly because the switch is not complete and the still linker misses some components from Boost.Test, see for instance http://tinyurl.com/yfohqh . As a result, the regression count has raised to 280 :( and possibly growing. I don't want to sound rude, but was it necessary to do this potentially dangerous change precisely now that we were so very close to being able to release a beta? If the change was aimed at solving some regression, it'd be good to know what the problem was so that maybe we can find some other, less drastic, workaround. Again, I don't want to offend anyone and maybe there's good reason for having commited the change. Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo
Uh oh - I meant to check this into the HEAD only. If this is not the case its a HUGE mistake on my part. Let me look into this. Robert Ramey "JOAQUIN LOPEZ MU?Z" wrote:
Hello,
Recent changes in test_tools.hpp aimed at switching from Boost.Test to boost/detail/lightweight_test.hpp are causing lots of regressions, seemingly because the switch is not complete and the still linker misses some components from Boost.Test, see for instance http://tinyurl.com/yfohqh . As a result, the regression count has raised to 280 :( and possibly growing.
I don't want to sound rude, but was it necessary to do this potentially dangerous change precisely now that we were so very close to being able to release a beta? If the change was aimed at solving some regression, it'd be good to know what the problem was so that maybe we can find some other, less drastic, workaround. Again, I don't want to offend anyone and maybe there's good reason for having commited the change.
Thank you,
Joaquín M López Muñoz Telefónica, Investigación y Desarrollo _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
Robert Ramey wrote:
Uh oh - I meant to check this into the HEAD only. If this is not the case its a HUGE mistake on my part. Let me look into this.
Robert Ramey
Browsing the online CVS tree looks to me that the changes have been checked into the HEAD and only the HEAD. This is what I intended and I would have thought this was OK. What am I missing here? Is checking into the HEAD affection the RC_1_34_0 branch? Is the RC_1_34_0 somehow dependent on the HEAD.Is it possible that the test RC_1_34_0 test results in RudbekAssociates-V2 are using the head rather than the RC_1_34_0 ? I would be happy to fix it if I knew how. Robert Ramey
Robert Ramey wrote:
Robert Ramey wrote:
Uh oh - I meant to check this into the HEAD only. If this is not the case its a HUGE mistake on my part. Let me look into this.
Robert Ramey
Browsing the online CVS tree looks to me that the changes have been checked into the HEAD and only the HEAD.
Does not seem this way, here's the part of the changelog that I see: 2006-12-16 09:05 ramey * test/: test_tools.hpp (RC_1_34_0), A.hpp, D.hpp, Jamfile: moved to lightweight testing framework 2006-12-16 09:00 ramey * src/basic_iarchive.cpp (RC_1_34_0): remove superfluous functions for helpers Looks like you're mistakenly checked in part of your changes in RC_1_34_0. - Volodya
Vladimir Prus wrote:
Robert Ramey wrote:
Robert Ramey wrote:
Uh oh - I meant to check this into the HEAD only. If this is not the case its a HUGE mistake on my part. Let me look into this.
Robert Ramey
Browsing the online CVS tree looks to me that the changes have been checked into the HEAD and only the HEAD.
Does not seem this way, here's the part of the changelog that I see:
2006-12-16 09:05 ramey
* test/: test_tools.hpp (RC_1_34_0), A.hpp, D.hpp, Jamfile: moved to lightweight testing framework
2006-12-16 09:00 ramey
* src/basic_iarchive.cpp (RC_1_34_0): remove superfluous functions for helpers
Looks like you're mistakenly checked in part of your changes in RC_1_34_0.
So it seems. I found the "Sticky Tag" set for this one file. I'll back this change out of RC_1_34_0 and put in the head. Now, if I can figure out the easy way to do this... Sorry for the inconvenience. Robert Ramey
OK - I believe I've fixed this. RC_1_34_0 should be in the state it was up until yesterday. I found two sticky tags in my files which I've now removed so I'm hoping this should be solved. Thanks for catching this. Robert Ramey Robert Ramey wrote:
Vladimir Prus wrote:
Robert Ramey wrote:
Robert Ramey wrote:
Uh oh - I meant to check this into the HEAD only. If this is not the case its a HUGE mistake on my part. Let me look into this.
Robert Ramey
Browsing the online CVS tree looks to me that the changes have been checked into the HEAD and only the HEAD.
Does not seem this way, here's the part of the changelog that I see:
2006-12-16 09:05 ramey
* test/: test_tools.hpp (RC_1_34_0), A.hpp, D.hpp, Jamfile: moved to lightweight testing framework
2006-12-16 09:00 ramey
* src/basic_iarchive.cpp (RC_1_34_0): remove superfluous functions for helpers
Looks like you're mistakenly checked in part of your changes in RC_1_34_0.
So it seems. I found the "Sticky Tag" set for this one file. I'll back this change out of RC_1_34_0 and put in the head. Now, if I can figure out the easy way to do this...
Sorry for the inconvenience.
Robert Ramey
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
"Robert Ramey" <ramey@rrsd.com> writes:
Looks like you're mistakenly checked in part of your changes in RC_1_34_0.
So it seems. I found the "Sticky Tag" set for this one file. I'll back this change out of RC_1_34_0 and put in the head. Now, if I can figure out the easy way to do this...
Sorry for the inconvenience.
FYI, it's usually a lot easier to keep two working copies, one of which covers the branch. -- Dave Abrahams Boost Consulting www.boost-consulting.com
participants (4)
-
"JOAQUIN LOPEZ MU?Z" -
David Abrahams -
Robert Ramey -
Vladimir Prus