January 2015 - Boost - lists.preview.boost.org

URGENT: Regression testing repo migration..
by Rene Rivera 19 Jan '15

19 Jan '15

I'm going to remove the outdated regression testing code in the boostorg/boost super-repo the weekend of January 24th. Everyone MUST stop using the old scripts immediately as they will disappear in approximately 7 days. *** All testers please migrate to using this run.py NOW <https://raw.githubusercontent.com/boostorg/regression/develop/src/run.py>. *** All report generation please migrate to using the comprehensive build_results_all.sh script NOW: < https://raw.githubusercontent.com/boostorg/regression/develop/xsl_reports/b… >. -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

3 2

[iostreams][gzip][zlib] zlib vs gzip, and linking against external libraries
by Aaron Levy 18 Jan '15

18 Jan '15

What is the difference in functionality between gzip and zlib compression? I built Boost iostreams library on Linux after exporting ZLIB_SOURCE and BZIP2_SOURCE environment variables. However when I build a sample program using gzip / bzip2 with IOStreams, I do not need to link to libzlib or libbz2. Is the zlib / bz2 code built into libboost_iostreams in this case? For reference, this is the command line I used to build the iostreams library. ./b2 --with-iostreams --build-dir=../build threading=multi variant=release link=shared runtime-link=shared -- Aaron Levy aaron.levy(a)yandex.com

3 2

Building the entire Boost documentation as Windows HTML help
by Philip Bennefall 18 Jan '15

18 Jan '15

Hi all, Is it possible to build a Windows HTML help file from the Boost documentation? It'd be a nice way of navigating it, and it's a single file so is trivial to move around. Has anyone attempted to do this? Kind regards, Philip Bennefall

1 0

Re: [boost] (correction) Clang 3.4 failing to compile boost 1.57 headers
by Fletcher, John P 18 Jan '15

18 Jan '15

Sorry folks, my value for Clang and gcc 5.0.0 was wrong - it does report 50000 I had a fault in my configuration and it was dropping back to the default 4.6.3 John ________________________________________ From: Boost [boost-bounces(a)lists.boost.org] on behalf of Peter Dimov [lists(a)pdimov.com] Sent: 18 January 2015 18:54 To: boost(a)lists.boost.org Subject: Re: [boost] Clang 3.4 failing to compile boost 1.57 headers Fletcher, John P wrote: > Results are as follows for clang 3.5 > > gcc version gcc reports clang reports > 4.6.3 40603 40700 > 4.8.2 40802 40800 > 4.9.0 40900 40900 > 4.9.2 40902 40900 > 5.0.0 50000 40700 <- actually 50000 > > I can see that there are some serious problems getting the correct > identification with Clang. Looks like unless it recognizes 4.8 or 4.9, it reports 40700 by default, which would make things a bit complicated. _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost

1 0

[uuid] Platform-specific UUID generator (was Re: Issue 9407: please merge fix)
by Klaim - Joël Lamotte 18 Jan '15

18 Jan '15

On Fri, Jan 16, 2015 at 10:49 AM, Antony Polukhin <antoshkka(a)gmail.com> wrote: > Looks like Boost.UUID library is not maintained. This is unfortunate and reminded me this exchange: https://twitter.com/niv_anteru/status/487496938150195200 A platform-specific random generator would apparently be far more efficient (I say apparently because I didn't know there were platform-specific uuid generation until I read this tweet). I reminded this with the fix discussion and made the ticket marked as feature request: https://svn.boost.org/trac/boost/ticket/10939 I'm taking note to try a patch but don't hold your breath. Also, would there be any benefit in having another typedef in C++11 mode which would use the std mersene twister instead of the boost one?

3 4

[filesystem] path.cpp fails to compile under clang/Cygwin
by Peter Dimov 18 Jan '15

18 Jan '15

Fails like this: clang-linux.compile.c++.without-pth ..\..\..\bin.v2\libs\filesystem\build\clang-linux-3.4.2\release\link-static\path.obj ..\..\..\libs\filesystem\src\path.cpp:36:11: fatal error: 'windows_file_codecvt.hpp' file not found # include "windows_file_codecvt.hpp" ^ 1 error generated. "clang++" -c -x c++ -Wno-variadic-macros -Wno-c99-extensions -O3 -Wno-inline -Wall -DBOOST_ALL_NO_LIB=1 -DBOOST_FILESYSTEM_STATIC_LINK=1 -DBOOST_SYSTEM_STATIC_LINK=1 -DNDEBUG -I"..\..\..\include" -o "..\..\..\bin.v2\libs\filesystem\build\clang-linux-3.4.2\release\link-static\path.obj" "..\..\..\libs\filesystem\src\path.cpp" Looks like a clang bug; the file is certainly there, in the same directory as path.cpp.

4 12

How do I copy all source files?
by Robert Dailey 18 Jan '15

18 Jan '15

I do adhoc builds of boost using the boost modular git repo. All libs build just fine, but I can't for the life of me copy the source files. I have tried doing it through FileZilla (SFTP) but it skips a ton of files due to the symlinks. How can I get the complete "boost" header source tree?

2 5

[Phoenix] [MSVC] 8 9 and 10 COMDAT error
by Fletcher, John P 17 Jan '15

17 Jan '15

I am attempting to sort out the bugs and test failures in Boost Phoenix. I have one test (bind_rv_sp_test) which fails at the link stage with the following error on MSVC 8 9 and 10 but not later bind_rv_sp_test.obj : fatal error LNK1179: invalid or corrupt file: duplicate COMDAT '?value@?$result_@P8Y@@AE?AV?$shared_ptr@UX@@@boost@@XZ@?$is_mem_fun_pointer_select@$0A@@detail@boost@@2_NB' I have been researching this error and attempting to find a workaround for Phoenix. Has anyone else working on Boost encountered this error and have any suggestions about it. It does appear to have been around for some years and there are a lot of suggestions but no clear resolution I have found. How important are these compilers, and would it be O.K. to mark this test as expected to fail? I don't have these compilers so for any new tests I am relying on the develop tests to check out experiments. Thank you John

2 2

Re: [boost] [compute]/ Futures Some remarks
by Antonio Colino 16 Jan '15

16 Jan '15

> > On 01/14/2015 10:46 PM, Antonio Colino wrote: > > I have a flawless implementation of this concept I am very proud of. Is any one willing to critique it? > > On 01/14/2015 10:46 PM,?Thomas Heller?wrote: > > Wanna share it? Which concept are you refering to? > > I was hoping to share this lib, if i could generate any interest in a copyable hpc version of packaged_task > On Jan 2015 20:36:40,?Thomas Heller?wrote: > Absolutely! Note that there are already a few out there doing exactly what > you described (HPX and Habanero++ come to my mind here). >On Date: Thu, 15 Jan 2015 21:58:37, Antonio Colino wrote: >What would be the best way to arrange this? Thomas if you are sill interested I am still curious by which method you would prefer to review the code, as it is not yet made public and I would like to have a second opinion before doing so.

1 0

[function] unmerged changes
by Peter Dimov 16 Jan '15

16 Jan '15

Boost.Function's develop branch appears to have accumulated a few changes that aren't in master, to which I've added one more (I neglected to check first). Looking at the test results, nothing there seems out of the ordinary, so these should probably be merged, perhaps without my final commit (which is itself a very low risk, as it merely changes detail:: to boost::detail:: in a few places).

2 1