- Boost - lists.boost.org

FreeBSD regression testing
by Bruce Simpson 22 Jul '09

22 Jul '09

Hi all, Thanks to whomever cleaned up the old runs on this page: http://www.boost.org/development/tests/release/developer/summary.html I just wanted to ask: * what's the procedure here, are old runs cleaned up automatically? * Is there a preferred naming convention for the runs? We're aware of some issues with Boost.Build on 64-bit FreeBSD platforms at the moment. I'll be starting another run shortly, but will be performing it on i386, so as to avoid the FAILs we're seeing. I just merged a fix to FreeBSD's system gcc to deal with some issues in global 'using' declarations, that should bring FreeBSD-STABLE/i386 closer to the other platforms in terms of Boost stability. I would like to get a slice of FreeBSD cluster resources to run the Boost regression tests periodically, so I'll have a chat with the cluster admins about it. thanks, BMS

2 1

[threadpool] renmaed to task
by k-oli＠gmx.de 22 Jul '09

22 Jul '09

Hello, boost.threadpool is renamed to boost.task - the review schdule list should be updated soon. Oliver

1 0

[1.40.0] Release plans
by Beman Dawes 22 Jul '09

22 Jul '09

I've been missing in action for most of the month, working on C++ committee stuff. Sigh. Rather than set a new release schedule for 1.40.0 right now, I'm planning to spend the next week or so looking at the state of the release branch, and pestering folks as needed. As that process proceeds I'll get a better sense of what a reasonable schedule would be to finish the release. In the meantime, the release branch is still open for merges as long as they are stable on trunk tests. --Beman

1 0

[introspection] Status
by Richard Hazlewood 22 Jul '09

22 Jul '09

I see the vault contains some introspection helpers (for checking for member function existence). AFAICT these utilities are not currently in boost nor scheduled for review. Are there any plans to bring them forward? (There appears to be an overlap with some of the introspection utils in mpl - though only for member variables there.)

5 6

FW: [uuid] changes to fix warnings from the linker
by christian.leutloff＠web.de 22 Jul '09

22 Jul '09

Hi, thank you for the boost.uuid lib! Could you please add some more "inline"s to some of the methods to avoid warnings from the VS2008 compiler? I'm using the uuid implementation from uuid_alt_v13_pod.zip downloaded from http://www.boostpro.com/vault/index.php. This is one of the warnings: warning LNK4006: "bool __cdecl boost::uuids::operator<(class boost::uuids::uuid const &,class boost::uuids::uuid const &)" (??Muuids@boost@@YA_NABVuuid@01@0@Z) already defined in xxx.obj; second definition Here is the patch to fix the warnings: Index: boost/uuid/uuid.hpp =================================================================== --- boost/uuid/uuid.hpp (revision 65) +++ boost/uuid/uuid.hpp (working copy) @@ -152,28 +152,28 @@ x.swap(y); } -bool operator==(uuid const& lhs, uuid const& rhs) /* throw() */ +inline bool operator==(uuid const& lhs, uuid const& rhs) /* throw() */ { return (lhs.data_ == rhs.data_); } -bool operator!=(uuid const& lhs, uuid const& rhs) /* throw() */ +inline bool operator!=(uuid const& lhs, uuid const& rhs) /* throw() */ { return !(lhs == rhs); } -bool operator<(uuid const& lhs, uuid const& rhs) /* throw() */ +inline bool operator<(uuid const& lhs, uuid const& rhs) /* throw() */ { return (lhs.data_ < rhs.data_); } -bool operator>(uuid const& lhs, uuid const& rhs) /* throw() */ +inline bool operator>(uuid const& lhs, uuid const& rhs) /* throw() */ { return rhs < lhs; } -bool operator<=(uuid const& lhs, uuid const& rhs) /* throw() */ +inline bool operator<=(uuid const& lhs, uuid const& rhs) /* throw() */ { return !(lhs > rhs); } -bool operator>=(uuid const& lhs, uuid const& rhs) /* throw() */ +inline bool operator>=(uuid const& lhs, uuid const& rhs) /* throw() */ { return !(lhs < rhs); } @@ -191,7 +191,7 @@ } // As named in RFC 4122, 4.1.7 -uuid nil() { +inline uuid nil() { uuid u = uuid(); return u; } Index: boost/uuid/uuid_io.hpp =================================================================== --- boost/uuid/uuid_io.hpp (revision 65) +++ boost/uuid/uuid_io.hpp (working copy) @@ -22,7 +22,7 @@ namespace detail { -int get_showbraces_index() +static int get_showbraces_index() { static int index = std::ios_base::xalloc(); return index; Index: boost/uuid/uuid_namespaces.hpp =================================================================== --- boost/uuid/uuid_namespaces.hpp (revision 65) +++ boost/uuid/uuid_namespaces.hpp (working copy) @@ -42,7 +42,7 @@ namespace namespaces { /* Name string is a fully-qualified domain name */ -uuid dns() { +inline uuid dns() { uuid u = { /* 6ba7b810-9dad-11d1-80b4-00c04fd430c8 */ 0x6b, 0xa7, 0xb8, 0x10, 0x9d, 0xad, @@ -53,7 +53,7 @@ } /* Name string is a URL */ -uuid url() { +inline uuid url() { uuid u = { /* 6ba7b811-9dad-11d1-80b4-00c04fd430c8 */ 0x6b, 0xa7, 0xb8, 0x11, 0x9d, 0xad, @@ -64,7 +64,7 @@ } /* Name string is an ISO OID */ -uuid oid() { +inline uuid oid() { uuid u = { /* 6ba7b812-9dad-11d1-80b4-00c04fd430c8 */ 0x6b, 0xa7, 0xb8, 0x12, 0x9d, 0xad, @@ -75,7 +75,7 @@ } /* Name string is an X.500 DN (in DER or a text output format) */ -uuid x500() { +inline uuid x500() { uuid u = { /* 6ba7b814-9dad-11d1-80b4-00c04fd430c8 */ 0x6b, 0xa7, 0xb8, 0x14, 0x9d, 0xad, Bye Christian -- Christian Leutloff, Aachen, Germany ______________________________________________________ GRATIS für alle WEB.DE-Nutzer: Die maxdome Movie-FLAT! Jetzt freischalten unter http://movieflat.web.de

1 0

[1.40] release schedule?
by joaquin＠tid.es 22 Jul '09

22 Jul '09

Hi, The Google Calendar schedule for Boost 1.40 (http://www.boost.org/development/index.html ) says we're one day past the beta delivery date. How does this schedule still apply? Is it OK to make doc changes to the release branch? Thank you, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

2 1

[Date_Time] -proposal to overload tz_database::load_from_file method
by Slawomir Dors 21 Jul '09

21 Jul '09

Hi ! I'm using the Date_Time library on Windows. The "tz_database" has only one method to populate it with the data from "date_time_zonespec.csv" file. It is "load_from_file" method which takes a filename in std::string. I wonder if it's possible to overload that method so it could be more generic (and much more useful). It could take for example a reference to std::istream object. The idea is to avoid to save the data in a .csv file on a disk. I could then put the data in the resources of an .exe or .dll file, create some stream object derived from std::istream and load into the "tz_database" object. The second limitation of the existing version of "load_from_file" is that I can't use unicode string to provide the filename. What the Developers think about such small addition ? Cheers, Slawomir Dors

1 0

_WIN32_WINNT in thread\src\win32\thread.cpp and filesystem\src\operations.cpp
by James Mansion 21 Jul '09

21 Jul '09

Why does win32\thread.cpp unconditionally define _WIN32_WINNT and WINVER? That's a bit rude, especially when I go to lengths to set them properly in a file I pass as BOOST_USER_CONFIG. :-( This also applies in filesystem\src\operations.cpp: in that case it does check that _WIN32_WINNT isn't defined already, but there has been no inclusion of config.h so BOOST_USER+CONFIG has not been processed yet. I would have though that in both cases the correct response is to remove these explicit settings and just: #include <boost/config.h> James (Using 1.39.0 on Vista, BTW - and setting _WIN32_WINNT by detection in my waf script)

2 1

[smart_ptr] shared_ptr template type
by Frank Mori Hess 21 Jul '09

21 Jul '09

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I wonder if it might be better/possible to have a shared_ptr<T> that didn't assume its corresponding "raw" pointer type was T* but rather T. For example, what is now a shared_ptr<int> would be specified as a shared_ptr<int*>. The motivation would be to allow support for other classes as the "raw" pointer type. For example, I might want to have a shared_ptr<monitor<T*> > where the monitor<T*> has an operator->() which does automatic locking/unlocking of a mutex before/after forwarding the operator->() to the plain T pointer. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux) iEYEARECAAYFAkpWA5YACgkQ5vihyNWuA4UQOwCeI7sjr4/pByXIqRtactVfi8fk MUgAoLmstL+7V5NW+yYNXyDPlVog7VAI =JE48 -----END PGP SIGNATURE-----

8 29

[TypeTraits] Type traits for member pointers?
by Celtic Minstrel 20 Jul '09

20 Jul '09

I recently noticed that the Boost Type Traits library does not include any traits for member pointers apart from identification. For example, there are no "add member pointer" or "remove member pointer" traits, nor a trait to extract the containing type from a member pointer (ie extract Q from T Q::*). And there is no trait for decomposing member function pointers. It would appear that these traits could be added fairly easily; I was able to create them within a few minutes, and verify that they worked (with g++). Is there sufficient reason to add them to the library? For me, completeness seems like reason enough (unless it would be a lot of work). On the other hand, member pointers are, as far as I know, rarely used. So, um... thoughts? Is it a good idea? Should I post my code? --SPCD "Celtic Minstrel"

4 7