January 2009 - Boost-users - lists.preview.boost.org

Using different compilers with boost.
by Dan Smithers 23 Jan '09

23 Jan '09

Hi, I need to use several compilers for my project for different target computers. I have built boost 1.37 for gcc 4.1.2 and 3.4.3 and everything seems to be nicely set up as <boost_dir>/lib/libboost_xxx-gcc34-mt-1_37.so and <boost_dir>/lib/libboost_xxx-gcc41-mt-1_37.so. Is there anyway of automatically detecting which version to link? I suppose that gcc --version and some sort of awk script could do this. thanks dan

2 1

How does one build a "release" version of Boost with the -Zi or /Zi compiler flag enabled?
by Joseph A. Marks 23 Jan '09

23 Jan '09

How does one build a "release" version of Boost with the -Zi or /Zi compiler flag enabled? Please e-mail answer directly to josephamarks(a)yahoo.com As well as to this list! Thanks Very much. Joseph A. Marks PHONE: 612-788-9658 e-mail: josephamarks(a)yahoo.com

2 1

Re: [Boost-users] 1.37: boost::enable_shared_from_this fails whil estd::tr1 succeed s
by Daniel Lidström 23 Jan '09

23 Jan '09

> Daniel Lidström: > > I have already found a workaround, but it is a little bit annoying: > > > > struct Sub : public Base, > > public tr1::enable_shared_from_this<Sub> > > { > > void f() > > { > > BasePtr basePtr = > > tr1::dynamic_pointer_cast<Base>(shared_from_this()); > > } > > }; > > > > I would like to avoid explicit type casts like this. > > You don't need the cast. Of course. Sorry I didn't see the obvious this time. /Daniel

1 0

[date time] Querying current time zone settings.
by Rodolfo Federico Gamarra 23 Jan '09

23 Jan '09

Hi all! I'm using boost::posix_time::second_clock::local_time() to retrieve my current local time, and then using operator<< to stream the resulting ptime. I'd like to add to the output my local time zone (as configured in the underlying OS). Be it just a time difference or (better) a region (like "America/New_York"). I have been looking but didn't find how to to it. The class ptime doesn't include the time zone; and all the other examples I've seen in to docs work with an explicitly used time zone in the code. For example: time_zone_ptr zone( new posix_time_zone("MST-07") ); local_date_time ldt = local_sec_clock::local_time(zone); And, for instance, in the docs about I/O it can be seen "%q ISO time zone (output only). This flag is ignored when using the time_facet with a ptime.". There's also a boost::date_time::c_local_adjustor (which warns about "This class assumes that timezone settings of the machine are correct. This can be a very dangerous assumption."), but doesn't help me much. By now, I can use (also) boost::posix_time::second_clock::universal_time() to infer my time zone. Is there any alternative to this? Thanks a lot. -- Rodolfo Federico Gamarra

1 0

how to make a vector to derivedClass::create() methods
by Victor.Whiskey.Yankee 23 Jan '09

23 Jan '09

Hi, I wish to make a std::vector to some create() methods from several different derived classes (same ABC). I could instead use the ctors if necessary. lambda::bind<> and lambda::constructor<> maybe?, but I fail getting syntax. Thank you, Vic

2 4

1.37: boost::enable_shared_from_this fails while std::tr1 succeed s
by Daniel Lidström 23 Jan '09

23 Jan '09

Hello, this sample does not compile when using the boost version: #if 1 #include <boost/enable_shared_from_this.hpp> namespace tr1 = boost; #else #include <memory> namespace tr1 = std::tr1; #endif namespace { struct Base { virtual void f() = 0; }; typedef tr1::shared_ptr<Base> BasePtr; struct Sub : public Base, public tr1::enable_shared_from_this<Base> { void f() { BasePtr basePtr = shared_from_this(); } }; } int main() { return 0; } boost\enable_shared_from_this.hpp(50) : error C2446: '==' : no conversion from 'boost::enable_shared_from_this<T> *const ' to '`anonymous-namespace'::Base *' with [ T=`anonymous-namespace'::Base ] Types pointed to are unrelated; conversion requires reinterpret_cast, C-style cast or function-style cast boost\enable_shared_from_this.hpp(48) : while compiling class template member function 'boost::shared_ptr<T> boost::enable_shared_from_this<T>::shared_from_this(void)' with [ T=`anonymous-namespace'::Base ] test.cpp(23) : see reference to class template instantiation 'boost::enable_shared_from_this<T>' being compiled with [ T=`anonymous-namespace'::Base ] The tr1 version compiles though. I am using Visual Studio 2008 with SP1 installed. I have already found a workaround, but it is a little bit annoying: struct Sub : public Base, public tr1::enable_shared_from_this<Sub> { void f() { BasePtr basePtr = tr1::dynamic_pointer_cast<Base>(shared_from_this()); } }; I would like to avoid explicit type casts like this. Why does the boost version fail? Thanks in advance! Regards, Daniel Lidström Stockholm, Sweden

2 1

[boost][interprocess] upgradable_mutex is semi-recursive...
by Lars Hagström 23 Jan '09

23 Jan '09

The subject is misleading, but hopefully eye-catching :-) I do understand that this is not an interprocess bug, but more of a result of how upgradeable locks work. But the behavior can be dangerous, as will be seen. What I've got happening is that one process locks the mutex recursively (sharable, of course). And this works (most of the time), since it is okay to lock the mutex shareably any number of times (even if it is within the same thread). But the problem occurs when a writer (in another thread) manages to get inbetween these lockings, when we get this sequence thread A locks mutex shareably thread B waits for exclusive lock thread A waits for sharable lock Since the upgradable mutex is written so that if there is an exclusive locker waiting it doesnt let any more sharable lockers in this will deadlock. Something that explains this could be useful in the interprocess documentation. Or, maybe the problem could be solved. I see two ways: 1. Detect recursive sharable locking, so that my thread A would always get an exception when trying to lock a lock that I've already got (it is a recipe for disaster) 2. Allow recursive sharable locks. If you've got the lock it's okay to take it again. Since I'm not really a fan of recursive locking (the above locking is a mistake, really) I think I would prefer solution 1 of these two. (I'm fixing my code to not be recursive now, but I thought this might be of interest to either Ion or anyone using boost interprocess.) Cheers Lars

2 2

[boost program options] missing value validation
by Andy F 23 Jan '09

23 Jan '09

If I have a parameter which needs a value, but the value is not specified in the command line, I get an unknown exception in VisualStudio 2005. Meaning, if I declare a --compression=something, but then call the program with just --compression, the exception is thrown at the po::store() method. How do I intercept this and validate? I am just using the code from the example program: po::options_description desc("Allowed options"); desc.add_options() ("help", "produce help message") ("compression", po::value<std::string>(), "set compression level") ; po::variables_map vm; po::store(po::parse_command_line(ac, av, desc), vm); po::notify(vm); if (vm.count("help")) { std::cout << desc << "\n"; return 1; } if (vm.count("compression")) { std::cout << "Compression level was set to " << vm["compression"].as<std::string>() << ".\n"; } else { std::cout << "Compression level was not set.\n"; }

2 5

[BGL] Remove vertices/edges from a subgraph
by Grzegorz Slodkowicz 23 Jan '09

23 Jan '09

What is the reason for vertex/edge removal functions in subgraphs being unimplemented? Is it simple oversight / lack of demand or would that pose some problems? Regards, Greg

2 4

Help with a boost::lambda problem and boost::array?
by jesseperla＠gmail.com 23 Jan '09

23 Jan '09

> Is there a solution? It appeared to me that the likely winner was: http://www.boost.org/doc/libs/1_37_0/doc/html/lambda/extending.html Then it seems we would use the boost array's typedef value_type to get the return type. However, I do not see operator[] overloading in the list of supporting operators here. But the following code compiles: //Testing lambda and vectors/arrays. using lambda::_1; std::vector<Real> vec(1); vec[1] = 0; boost::function<Real (std::vector<Real>)> vec_f = _1[0]; //Returns the 0'th element. Hence, it seems that this type of action is supported. Any idea where the sourcecode for the std::vector lambda operators are? might be useful to just copy/paste. Also, I wonder if there is any way to do this generically for all vector/array types supporting the Container concept? (I use a lot of ublas::vectors). And if operator() is also supported, I use a lot of ublas::matrix as well, so this would be nice to have in lambda. Jesse

3 4