- Boost - lists.boost.org

[1.32 release] Intel 8.0/Mingw regression results
by Aleksey Gurtovoy 22 Jul '04

22 Jul '04

Stefan Slapeta's Intel 8.0 and Rob Lievaart's Mingw/gcc 3.2.3 results are now available in the Boost-wide summary reports: http://www.boost.org/regression-logs/developer/summary.html -- Aleksey Gurtovoy MetaCommunications Engineering

4 4

RE: [boost] Re: Regression results
by Rob Lievaart 22 Jul '04

22 Jul '04

Hi Alekey, I am currently running your script, for tests with mingw (gcc 3.2.3) So you should see these results appearing as well, sometime today (machine is not that fast). I'll watch the output of the upload and if it fails I'll try the new script. After a quick glance at the bjam.log (still growing) I noticed that python was not setup correctly, so those tests are not run. I'll try to correct that for the next run. And in boost tests there is a problem with wcscmp not being declared in std, that causes a lot of tests to fail, but I thought I read on the list that a fix for that was submitted ?!? Well, lets see if the upload works. Kind regards, Rob Lievaart. -----Original Message----- From: Aleksey Gurtovoy [mailto:agurtovoy@meta-comm.com] Sent: 21 July, 2004 1:35 PM To: boost(a)lists.boost.org Subject: Re: [boost] Re: Regression results for Intel 8 online Stefan Slapeta writes: > The regressions are finished and should appear in the metacomm result > tables after the next update there. Stefan, It seems that the logs didn't get uploaded. Could you please run the attached 'regression.py' with the 'upload' command, e.g. python regression.py upload --runner=<your runner id> ?

2 1

Como --windows experiment failed
by Jeff Garland 22 Jul '04

22 Jul '04

Based on the regression report it looks like --windows isn't accepted: The core of it is Command-line error #571: invalid option: --windows 1 catastrophic error detected in this compilation. Compilation terminated. Details at: http://www.boost.org/regression-logs/output/metacomm-bin-boost-libs-date_ti… or http://tinyurl.com/45z6v Thanks guys -- you will need to back out the change when you get a chance. Looks like it's time to mark the dll case as failing for Comeau. Jeff

2 1

[Range] partial_workaround failure on Linux
by Jeff Garland 22 Jul '04

22 Jul '04

One of the few remaining failures on Mandrake is a range library failure that's been there for the better part of a week. Looks like it might be easy to solve... Jeff URL: http://boost.sourceforge.net/regression-logs/cs-LinuxMandrake-links.html#ra… or http://tinyurl.com/6xwh9 Small tease: range - workaround_test - gcc Compiler output: In file included from /boost_regression1/boost/boost/range/detail/common.hpp:23, from /boost_regression1/boost/boost/range/detail/iterator.hpp:14, from /boost_regression1/boost/boost/range/iterator.hpp:21, from ../libs/range/test/partial_workaround.cpp:13: /boost_regression1/boost/boost/mpl/if.hpp:141: error: namespace `algo_' undeclared /boost_regression1/boost/boost/mpl/if.hpp:143: error: `void_' was not declared in this scope /boost_regression1/boost/boost/mpl/if.hpp:143: error: `void_' was not declared in this scope ... set -e "g++" -c -Wall -ftemplate-depth-100 -g -O0 -fno-inline -I"../bin/boost/libs/range/test" -I "/home/jeff/devTools/boost_regression1/boost" -o "../bin/boost/libs/range/test/workaround_test.test/gcc/debug/partial_workaround.o" "../libs/range/test/partial_workaround.cpp" "/usr/bin/objcopy" --set-section-flags .debug_str=contents,debug "../bin/boost/libs/range/test/workaround_test.test/gcc/debug/partial_workaround.o"

2 1

Proposal for String Algorithms Library [CODE]
by zaufi＠sendmail.ru 22 Jul '04

22 Jul '04

---------- Forwarded Message ---------- Subject: Proposal for String Algorithms Library Date: Thursday 22 July 2004 09:43 From: zaufi(a)sendmail.ru To: boost(a)lists.boost.org hi boosters I would like to propose one more algorithm to SAL -- squeeze. It removes leading/trailing spaces and duplicate spaces between words. So sample string " This is \n\n a test \n" become "This is\na test". I use this function with XML/HTML related code to 'normalize' content... Hope this will help to smbd else ;) Have fun! ------------------------------------------------------- Oops. Forget to paste code... ;) <squeeze.cc> std::string squeeze(const std::string& str) { std::string out; // bool is_start_of_line = true; char last_space = 0; for (std::string::const_iterator i = str.begin(); i != str.end(); ++i) { if (std::isspace(*i)) last_space = int(!is_start_of_line) * (int(*i == '\n') * '\n' + int((*i != '\n') & (last_space != '\n')) * ' ' + int(last_space == '\n') * last_space); else { if (last_space) { out.push_back(last_space); last_space = 0; } out.push_back(*i); is_start_of_line = false; } } return out; } </squeeze.cc>

1 0

typeof: lvalue preserving
by Vertleyb 22 Jul '04

22 Jul '04

One more addition to the typeof library is a macro, that, unlike BOOST_TYPEOF(), which removes top-level consts and references, attempts to be closer to the decltype, by preserving lvalue-ness of its argument: BOOST_TYPEOF_PRESERVE_LVALUE(expr) I do admit that the name is quite clumzy :( It is directly based on the rules of binding a reference. The following table outlines which types are returned: Expression lvalue? typeof_preserve_lvalue(expression) ---------- ------- ------ int n; lvalue int& const int m; lvalue const int& int& rn = n; lvalue int& const int& rm; lvalue const int& int foo1(); int const int foo2(); const int& (unfortunately, in both VC7.1 and GCC 3.3) int& foo3(); lvalue int& const int& foo4(); lvalue const int& int(21) int 21 int The BOOST_AUTO remains based on the BOOST_TYPEOF. The references, in the same way as proposed in http://www.osl.iu.edu/~jajarvi/publications/papers/decltype_n1478.pdf, should be specified explicitly: int& foo(); BOOST_AUTO(v, foo()); // auto v = foo(); BOOST_AUTO(&r, foo()); // auto& r = foo(); BOOST_AUTO(const& cr, foo()); // auto const& cr = foo(); See http://groups.yahoo.com/group/boost/files/typeof.zip (or Spirit repository, TYPEOF_DEV branch). As always any comments are welcome. Thanks, Arkadiy

1 0

Tribool missing title in docs
by Jeff Garland 22 Jul '04

22 Jul '04

Doug -- Maybe you've already noticed, but the Tribool title isn't showing up in the generated docs table of contents... http://www.boost.org/regression-logs/cs-win32_metacomm/doc/html/libraries.h… Jeff

2 1

Re:Re: [boost] Regression Tests, Boost.Serialization
by Robert Ramey 22 Jul '04

22 Jul '04

Jeff Garland wrote: >> Houston, we have a problem. >> >> I had a few GB of disk space left a few days ago. >> Now, I'm unable to run all the tests. I could >> drop one of the gcc-3.4 versions. However, I'd >> still be short by ~10GB after that. It's a fact - running the bjam serialization test is going to consume 20 GB. If the test machine doesn't have it the best is just to remove the serialization test from the status jamfile for that machine. Actually I think it would be unusual for a machine used for testing boost not to have this much free space available. So I would hope it only affects martin's tests. >It would be nice if we could drop serialization on compilers that just >aren't going to work. Having looked at the test results on all the compilers the only ones that I really wonder about are HP, and for now Comeau. All tests run with MSVC 6.0 and borland will fail without spirit 1.61 being installed. This is because spirit 1.61 is needed to build the libraries. So if the library build fails then all the tests are skipped. They are marked "fail" which is a little misleading - "skipped" might be better. >This is where we were hoping to be able to allow regression levels like >'basic' and 'torture' that would provide standard ways of controlling the >number of tests. But none of this is currently available. All in all I think the testing program is working well and in reasonable time. I need this to shake out issues with each platform. It just takes the resources that it takes. So far I've only been through two test cycles so it's premature to panic. >As for what we should do now, it would be nice to have an answer to the >question I posed this morning: is serialization going into 1.32 or are we >cutting it out in the final release: I want to keep working to finish regardless of which release it becomes a part of. >If we are cutting it out the we can just remove it from status bjam now.... But that would mean that progress would be stopped on getting this done. Robert Ramey

7 8

RE: [boost] Re: Lock unification [move]
by Batov, Vladimir 22 Jul '04

22 Jul '04

Yes, I fully agree with your argument against >> basic_lock(mutex_type&, const elapsed_time&, defer_type); That suggestion was OTOH and clearly messy and bad. Although the point I tried to make was that the interface appeared somewhat inconsistent -- I can create a deferred blocking lock but not a deferred "try" lock. However, I now clearly see that I was thinking in terms of separate classes "scoped, try, timed". That is not true anymore. After thinking about it some more (I know, I should have done it from the start :-) ), I like the proposed interface. Essentially, scoped_lock(mutex_type& m, detail::defer_lock_type); is a constructor in its own right as (unlike the other three) it does not create a specific (blocking, try, timed) lock. The more I think about it, the more I like it. > -----Original Message----- > From: boost-bounces(a)lists.boost.org [mailto:boost-bounces@lists.boost.org] > On Behalf Of Howard Hinnant > Sent: Thursday, 22 July 2004 11:19 AM > To: boost(a)lists.boost.org > Subject: [boost] Re: Lock unification [move] > > On Jul 21, 2004, at 7:16 PM, Batov, Vladimir wrote: > > > Howard, > > > > Thank you for taking on the task of shaping up the interface. That > > activity is of great importance. A few comments. > > > > explicit scoped_lock(mutex_type& m); > > scoped_lock(mutex_type& m, detail::defer_lock_type); > > scoped_lock(mutex_type& m, detail::try_lock_type); > > scoped_lock(mutex_type& m, const elapsed_time& elps_time); > > > > The problem (as I can see it) with the set above is that "deferred", > > "tried" and "timed" appear to be alternatives or mutually exclusive. > > However, "deferred" and, say, "try" functionalities are orthogonal. > > That > > is, as one might need to defer a blocking lock > > > > scoped_lock l(m, deferred); > > > > one might need a deferred try_lock as well. Therefore, as we are trying > > to provide one universal lock, it appears we have to have something > > complex like > > > > basic_lock(mutex_type&, const elapsed_time&, defer_type); > > > > that would address all the lock variability/configurability. > > I'm not seeing that the combination of time and defer always make sense: > > time defer description > --------------------------------------------------- > 0 yes interpret as don't lock > 0 no interpret as try-lock > finite yes no interpretation > finite no lock if you can get it in time > infinite yes no interpretation > infinite no interpret as lock > > Some of the combinations don't appear to have clear meanings. > > Even more importantly, the underlying mutex might not support time. > Below is a list of mutex functionality that I have so far identified in > the spec. Recursion is ignored for this discussion (it isn't relevant > to locks, and thus not relevant to mutex interface, only to mutex > implementation details): > > Mutex interface: > ---------------- > > void lock(); > void unlock(); > > bool try_lock(); > > bool timed_lock(const elapsed_time&); > > void lock_sharable(); > void unlock_sharable(); > > bool try_lock_sharable(); > > bool timed_lock_sharable(const elapsed_time&); > > void lock_upgradable(); > void unlock_upgradable(); > > bool try_lock_upgradable(); > > bool timed_lock_upgradable(const elapsed_time&); > > void unlock_and_lock_sharable(); > > void unlock_and_lock_upgradable(); > void unlock_upgradable_and_lock(); > > void try_unlock_upgradable_and_lock(); > > The spaces between signatures are relevant. They represent groupings > where I believe it is reasonable for a mutex to support a group of > signatures, but exclude other groups. For example, it is very > reasonable for a mutex to support lock and unlock but not try_lock. Or > perhaps lock, unlock and try_lock, but not timed_lock. If a lock > constructor forces the semantics of "deferred lock" to call > m_.timed_lock(elps_time), then a compile time error results if the > mutex does not support timed_lock(), even though timed_lock isn't > needed to support "deferred lock" (no mutex interface is required to > support deferred lock). So I'm very reluctant to have a lock function > that calls two mutex functions that can reasonably be expected to not > be implemented in the same mutex. > > -Howard > > _______________________________________________ > Unsubscribe & other changes: > http://lists.boost.org/mailman/listinfo.cgi/boost

1 0

RE: [boost] Re: Boost.Test problem
by Rozental, Gennadiy 22 Jul '04

22 Jul '04

> > My personal preference is fully qualified reference > everywhere. And I > > don't see how other solution is any better. > > Whatever... Just fix the thing! That's great! First you say don't do a thing, than - why you did not do it yet. I had enough headache with things that were "sure thing". I did not touch anything I am not 100% sure about. Gennadiy.

2 1