Alex, I bumped zlib this week so I have a bit of history here. All symbols from 1.2.3 versions have SUNW_1.2 label and newer are labeled with 1.3. The 64bit calls were introduced after 1.2.3 and now carry 1.3 label, have a look at mapfile in zlib directory.
Cheers, Erol Am 19.07.2013 um 12:53 schrieb Alexander Pyhalov <[email protected]>: > On 07/19/2013 14:14, Nikola M. wrote: >> On 07/19/13 12:06 PM, Alexander Pyhalov wrote: >>> On 07/19/2013 13:57, Jim Klimov wrote: >>>> On 2013-07-19 11:51, Udo Grabowski (IMK) wrote: >>>> >>>>>>> Maybe indeed a C++ incompatibility ? >>>> >>>>> Welcome to the world of binary incompatiblitly, now the last >>>>> fortress has finally been conquered..... >>>> >>>> Which moves me to think: would this example mean that legacy >>>> applications built for Solaris 10 and older, running now on >>>> obsolete deployments which might be targeted for upgrade to >>>> OI, would likely not run on newer hipster-based GCC-compiled >>>> OS releases? > >> Please don't try in Hipster, it is rolling release, hot even /dev. >> Try latest OI /dev, saying 151a7 , that you can upgrade from opensolaris >> 134, anyway. > > There is no longer opensolaris there. Migrated to FreeBSD in 2011 and rarely > regret about this ;) > > About current problem - Andrzej identified the reason - zlib update. Need > advices on fixing this. > > If we roll back zlib update, we need some ugly hack to update existing > installations. > The other idea is to identify incompatible changes and isolate them with > symbol versioning. I have no idea if it is possible and how hard it is. > -- > Best regards, > Alexander Pyhalov, > system administrator of Computer Center of Southern Federal University > > _______________________________________________ > oi-dev mailing list > [email protected] > http://openindiana.org/mailman/listinfo/oi-dev _______________________________________________ oi-dev mailing list [email protected] http://openindiana.org/mailman/listinfo/oi-dev
