- Boost - lists.preview.boost.org

xtime_get unsafe on Windows for IA64
by Ben Hutchings 26 Aug '04

26 Aug '04

The implementation of the boost::xtime_get() function in libs/thread/src/xtime.cpp includes the following code for Windows: #if defined(BOOST_HAS_FTIME) FILETIME ft; ... const boost::uint64_t TIMESPEC_TO_FILETIME_OFFSET = ((boost::uint64_t)27111902UL << 32) + (boost::uint64_t)3577643008UL; xtp->sec = (int)((*(__int64*)&ft - TIMESPEC_TO_FILETIME_OFFSET) / 10000000); xtp->nsec = (int)((*(… [View More]

2 1

[Serialization] question
by Russell Hind 26 Aug '04

26 Aug '04

I have a std::list<ublas::vector<double> > and also a list of iterators in to this list (which basically contain a subset of the initial list e.g. std::list<std::list<ublas::vector<double> >::iterator> Is there a way with serialization library to serialise both lists so that when both lists are re-loaded, the second list contains valid iterators again? Thanks Russell

1 0

[program_options + test] cw 8.3 fails all tests
by Vladimir Prus 26 Aug '04

26 Aug '04

As shown in http://tinyurl.com/3smdm all program_options tests fail on cw 8.3, because boost.test refers to the _errno variable, but that reference is somehow unresolved. Could anybody suggest any solution? TIA, Volodyaya

4 3

Re: [boost] [1.32 Plan] Stalled forever?
by Robert Ramey 26 Aug '04

26 Aug '04

Aleksey Gurtovoy wrote: >JOAQUIN LOPEZ MU?Z writes: >> I don't want to sound harsh or anything, and I truly appreciate the >> effort that's being put in this, but a month has slipped by since the >> first schedule for releasing, and AFAIK we don't even have an agreed >> upon branching date for the near future. >The branching date is pending MPL check-in. The latter has been postponed >several times as the new issues surfaced (in particular, due to better &… [View More]gt;tests), but it's crucial to ensure that the library is in the predictable >state before the release (remember that it goes on the CD with the TMP >book), so please bear with us. The branch should occur now. MPL update can go into the next release. Or you can see it a slightly different way. Branch and release 1.32 now. Rebranch and release 1.32 when MPL update has been added and any related issues resolved. > There are external constraints for this release's timeframe, > but within those, the release is criteria-driven. One of the > criteria is regression-free codebase, This is criteria for release - not for branching > and another is the new > MPL version. There is no point of releasing "on time" if the > release is useless. Nothing in the current development tree - i.e. the 1.32 release candidate depends on the MPL update that has yet to be integrated. There is no reason to hold up branch/release for new features. > > Shouldn't we settle now on a realistic branching date > and proceed with the release? Or, should I shut my mouth > and let the release manager handle this as he deems > convenient? > > Again, don't get me wrong. I am not willing to push > anybody, but the process looks, well, erratic. > Please don't forget that we are all volunteers here. The best > way to speed up things is to contribute to them. MPL issues > notwithstanding, we still have the regression field to be > cleared -- > http://www.meta-comm.com/engineering/boost-> regression/developer/summary.html. > Turning it green while MPL is in works will make a huge difference. The only way an MPL update can be expected to not create new issues is it makes only small changes. If that's the case its not worth holding things up for. If the changes are not small, it likely will create a slew of new issues. >> If you are willing to take over managing that one >> until the MPL check-in, be my guest. I wish we could all do as well as Joaquin has been doing in clearing issues related to his library. He has been doing an excellent job. When this topic was discussed at the beginning of the release process, it was indicated that subsequent releases would already be scheduled and would be soon after 1.32 . This was to assure developers that if they missed this release, it wouldn't be a big issue as the next one would be soon anyway. This is still a good idea and should be implemented. So Branch now Release branch available only for bug fixes and test clearing Main branch new MPL and what ever else. Robert Ramey [View Less]

2 1

[filesystem] Status of issues
by Beman Dawes 26 Aug '04

26 Aug '04

For reasons unrelated to Boost, I've been sidelined from most work on filesystem issues since late March and am just now starting to catch up. It will take awhile just to read backed up email. Thanks to John Maddock, Dave Abrahams, and Vladimir Prus for resolving several issues when I wasn't available. As far as the 1.32.0 release goes, I'll try to give priority to bug fixes. But there probably won't be a lot of change compared to the current CVS state, since the library is pretty stable … [View More]

5 4

RE: [boost] Re: Java style GUI in C++
by Reece Dunn 26 Aug '04

26 Aug '04

Ralf W. Grosse-Kunstleve wrote: >--- Reece Dunn <msclrhd(a)hotmail.com> wrote: > > Thus, in Java Swing-style C++, you can do something like: > > > > boost::gui::frame * main = new boost::gui::frame( "My GUI App" ); > > frame -> set_pane( new boost::gui::button( "PUSH" )); > > frame -> set_size( 500, 400 ); > > frame -> show(); > > > > boost::gui::message_loop msg; > > return msg.execute(); > >How … [View More]

1 0

RE: [boost] Re: Java style GUI in C++
by Reece Dunn 26 Aug '04

26 Aug '04

Miro Jurisic wrote: >I also believe this community is capable of producing such a framework, but >I >think that it's telling that the title of this thread is "Java style GUI in >C++"; a Java style GUI typically implies making user experience sacrifices >on >every platform, and therefore I believe that a "Java style GUI in C++" is >likely >to be driven by misguided goals. What I meant by "Java style GUI" is not in terms of L&F, but in terms of programming the … [View More]

7 8

more fixes in como-vc71
by Bronek Kozicki 26 Aug '04

26 Aug '04

I have received two more fixes from Comeau for libcomo 31, run on como 4.3.3, backend MSVC71: - better support for "long long" type - "int swprintf(wchar_t*, size_t, const wchar_t*, ...)" is available As usually, all fixes are available to everyone upon request. I believe these fixes will be incorporated into future version of libcomo, but this might require some more work (eg. make swprintf available on other backends etc.). B.

1 0

boost::add_reference breaks type deduction
by Judd Tracy 25 Aug '04

25 Aug '04

I am trying to use boost:add_reference in a template function to protect against the reference of reference problems, but it is breaking type deduction and I was wondering if any one know of a solution to this. template<typename Type> read(std::istream& stream, Type& data); This works properly of course. ----------------------------------------- template<typename Type> read (std::istream& stream, typename boost::add_reference<Type>::type data); This one does … [View More]

2 1

[mpl] A simple question about using lists/vectors
by Jody Hagins 25 Aug '04

25 Aug '04

I apologize in advance for this seemingly simple question, but I am having a hard time finding a simple solution in the online docs (I am in line for the MPL book, though ;-) Lets say I have some code like this... template <class A, class B> class Bar { public: Bar() { foo.template do_something<A>(); foo.template do_something<B>(); } private: Foo foo; }; Bar<Type1, Type2> bar; but I want to create a Bar with an arbitrary number of classes... Bar… [View More]

2 1