February 2007 - Boost-users - lists.preview.boost.org

[Boost-Users][serialization] Floating point precision
by jsanga＠cox.net 28 Feb '07

28 Feb '07

Is there a way to specify the precision of doubles when writing to xml_oarchive? Refer to the output snippet below. <Control class_id="3" tracking_level="0" version="1"> <Type>0</Type> <LowMZ>5.0499999999999998</LowMZ> <HiMZ>5.0999999999999996</HiMZ> I would like 5.05 and 5.1 for Low and High. Stepping through the code with the debugger I ended up in basic_text_oprimitive::save( const double t). void save(const double t) { if(os.fail()) boost::throw_exception(archive_exception(archive_exception::stream_error)); os << std::setprecision(std::numeric_limits<double>::digits10 + 2); os << t; } where it seems the precision is fixed. The only solution I can seem to think of is to split the serialization, and on the output convert my double to string (with the correct precsion) and then serialize the string. Is there a different or better way?

2 1

[Boost.MultiIndex] Problem using dataelement from derived class as key
by Manuel Jung 28 Feb '07

28 Feb '07

Hi, Im having a problem. I derive class B from class A. class a has the member A::id. I want to use this one as key for my MultiIndexContainer. But i get the following compiling errors: test.cpp:41: Fehler: Templateargument »&A::id« konnte nicht in »unsigned int B::*« umgewandelt werden test.cpp:42: Fehler: Templateargument 2 ist ungültig test.cpp:43: Fehler: Templateargument 1 ist ungültig test.cpp:44: Fehler: Templateargument 2 ist ungültig Sorry it on german. A testfile for reproducing my problem is attached. Can someone help? thanks.. Greetings Manuel Jung

2 3

[boost][circular_buffer] problem using copy contructor
by Andreas Rose 28 Feb '07

28 Feb '07

Hello, I'm new to circular_buffer. After reading the documentation i expect the following should work: const int c = 5; boost::circular_buffer<int> cb1(c); assert(cb1.capacity() == c); boost::circular_buffer<int> cb2(cb1); assert(cb2.capacity() == c); //failed, capacity() returns 0 But it does'nt work. After looking into the code of circular_buffer, i think the copy c'tor should be changed into something like this: circular_buffer(const circular_buffer<T, Alloc>& cb) : m_size(cb.size()), m_alloc(cb.get_allocator()) { initialize(cb.capacity()); BOOST_TRY { cb_details::uninitialized_copy(cb.begin(), cb.end(), m_buff, m_alloc); } BOOST_CATCH(...) { deallocate(m_buff, cb.capacity()); BOOST_RETHROW } BOOST_CATCH_END } Or do i missunderstand the documentation? I'm using vs2003 and the latest version of circular_buffer from boost-sandbox in ordinary debug mode. Thanks. Andreas Rose

1 0

Problem using regex library in Windows/cygwin
by Diego Molla 28 Feb '07

28 Feb '07

Hi, I have a problem using the Boost regex library. I have isolated the problem to the attached piece of code. For some reason the output returned by the function that uses regex is garbage. I have tested the same program on a Linux version of boost and it works fine. I wonder if the problem is on the Windows installation. Could someone using Windows or Cygwin test this and tell me if it works? I've just re-installed the cygwin version (1.33.1-3) and I still get the same error. Here is a typical interaction. It should extract the XML attribute and value but instead it returns garbage: [diego@valencia ~...af/src]$ g++ -o regexError.exe regexError.cpp -lboost_regex-gcc-mt [diego@valencia ~...af/src]$ regexError Write a fragment of XML: this <test arg="a1">is a test</test> yeah Write the XML tag to search: test Found element is a test Searching arg="a1" Found attribute 1g with value a1 Write a fragment of XML: this <test a2="more">is another test</test> Write the XML tag to search: test Found element is another test Searching a2="more" Found attribute or with value morY Write a fragment of XML: end [diego@valencia ~...af/src]$ -- Do you want to know what is going on in Language Technology in Australia and New Zealand? Join ALTA -- http://www.alta.asn.au/ This message is intended for the addressee named and may contain confidential information. If you are not the intended recipient, please delete it and notify the sender. Views expressed in this message are those of the individual sender, and are not necessarily the views of Macquarie University. --------------------------------------------------------------------- Diego MOLLA ALIOD diego(a)ics.mq.edu.au Department of Computing http://www.ics.mq.edu.au/~diego Macquarie University

2 1

[graph] patch to automatically build libbgl-viz.so and libbgl-viz.a
by David Baird 28 Feb '07

28 Feb '07

Hi, I'm interested in helping make the BGL easier to use right out-of-the-box. But, I'm not an expert at Boost (or Jamfiles) yet, so please guide me if this is wrong. I tried to patch the Jamfile so that it automatically builds the Graphgiz parser libraries (libbgl-viz.so and libbgl-viz.a). I only tested Jamfile (not Jamfile.v2) on some Gentoo boxes with gcc-4.1.1 and gcc-3.3.5 and it worked nicely. Some other future improvements I would like to do are to improve the Graphviz parser (to be able to parse more varieties of .dot files) and to review the documentation and check it for correctness (I already submitted some patches about half a year ago for a couple mistakes that I found). Please let me know what you think. Without further ado, here's the patch: Index: boost_1_33_1/libs/graph/build/Jamfile =================================================================== --- boost_1_33_1.orig/libs/graph/build/Jamfile +++ boost_1_33_1/libs/graph/build/Jamfile @@ -33,6 +33,21 @@ GRAPHVIZ_DEFS(graphviz_digraph_parser) = } local gGENERATOR_FUNCTION(LIB) = bgl-lib-file ; + dll bgl-viz + : ../src/$(SOURCES).cpp + : <sysinclude>$(BOOST_ROOT) <define>YY_NO_UNISTD_H + <runtime-link>dynamic + <define>BOOST_ALL_DYN_LINK=1 + # Darwin compiler ICEs if we turn optimization on + <darwin><*><debug-symbols>off + <darwin><*><optimization>off + # Intel compiler ICEs if we turn optimization on + <intel-win32><*><optimization>off + <vc-7_1><*><cxxflags>-GR- + : debug <inlining>on #inlining prevents linker name clashes due to truncation! + <debug-symbols>off # Debugging symbols become HUGE + ; + lib bgl-viz : ../src/$(SOURCES).cpp : <sysinclude>$(BOOST_ROOT) <define>YY_NO_UNISTD_H @@ -47,4 +62,7 @@ GRAPHVIZ_DEFS(graphviz_digraph_parser) = ; + install bgl-viz lib + : <dll>bgl-viz <lib>bgl-viz + ; } Index: boost_1_33_1/libs/graph/build/Jamfile.v2 =================================================================== --- boost_1_33_1.orig/libs/graph/build/Jamfile.v2 +++ boost_1_33_1/libs/graph/build/Jamfile.v2 @@ -24,3 +24,5 @@ obj graphviz_digraph_lex : graphviz_lex. <implicit-dependency>graphviz_digraph_parser ; obj graphviz_digraph_parser : graphviz_parser.yy : <bison.prefix>bgl_dir_ <define>GRAPHVIZ_DIRECTED=1 ; + +boost-install bgl-viz ;

1 1

[program_options] unsigned variables silently cast
by Torrey Pines 28 Feb '07

28 Feb '07

Hi With the current implementation, unsigned variables accept negative values, that get silently cast. E.g. on my platform, if "--length" expects size_t, --length -32 assigns it a value 4294967264 without warnings or errors. This behavior surprised me, and I'm using a workaround (parsing an int, and checking its range manually). However, is this an intended feature of the library? Thanks

2 1

[program_options] disable partial matches
by Torrey Pines 28 Feb '07

28 Feb '07

Hi With the current implementation, the user does not have to supply the full option names, e.g. "--he" is as good as "--help", unless there is an ambiguity. Is there any way to disable this, i.e. allow only full matches? Thanks

2 1

[program_options] checking the presence of required arguments
by Torrey Pines 28 Feb '07

28 Feb '07

Hi My program has a bunch of required arguments. I grouped them into one options_description: options_description required("Required options"); required.add_options() /* .... */ ; // ... options_description desc("Allowed options"); desc.add(required).add(other); I'd like to check that all required arguments are in fact supplied without having to do if(vm.count("arg1") == 0) { cerr << "usage error: arg1 is required\n"; return 1; } for each of them. Any suggestions? Thanks

2 1

boost::any template argument deduction
by Digvijoy Chatterjee 28 Feb '07

28 Feb '07

Hi, I have a function template like : template <class A> A what_type ( const boost::any &operand ){ A a; a = boost::any_cast<a>(operand); return a; } The idea is to provide the client code an easy way extract the type hidden by the boost::any value, Now there is no way to make the compiler deduce the type to return from an expression like below: std::string s = "hello"; boost::any a = s; some_op( what_type(a) ); //calls some_op(string &); Instead now the client has to write code using type() which returns std::typeinfo & if ( a.type().name() == "Ss" ) some_op(what_type<string> a); else if ( a.type().name() == "d" ) some_op(what_type<double> a); Or use functions to test type and cast: if ( is_string (a) ) some_op( what_type<string>(a)); else ( is_int(a) ) some_op ( what_type<int>(a)); ….. Is there any elegant way to handle this problem ?? --Digz

1 0

[MPL] Problem with vector
by Daniel 28 Feb '07

28 Feb '07

Hi, I have few Adapter classes. Each of them provide typedef operations which specifies which operations that Adapter supports (vector_c, with 0 or 1 at appropriate positions). I want to assert that two Adapters support different operations. However, my code doesn't compile. Any ideas? Daniel template<unsigned int N1, unsigned int N2> struct Adapter { typedef vector_c<unsigned int, N1, N2> operations; }; template<class TAdapter1, class TAdapter2> struct Check { typedef typename transform< typename TAdapter1::operations, typename TAdapter2::operations, plus<_1, _2> >::type sum_of_operations; typedef typename max_element<sum_of_operations>::type max_of_operations; BOOST_MPL_ASSERT((less_equal<max_of_operations, int_<1> >)); }; int main() { Check< Adapter<0, 1>, Adapter<1, 0> >(); return 0; }

2 1

[signals] slot<Signature>, slot::track() and bind syntactic sugar
by Frank Mori Hess 27 Feb '07

27 Feb '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I've added template parameters to the slot class in thread_safe_signals for specifying its signature, as slot<Signature> or slot1<R, T1>, etc. So now slots can be bound to each other. I've also dropped the automatic discovery of tracked objects, and added slot::track() functions. There is now some syntactic sugar for bind, if a slot constructor is called with more than one argument, then the arguments will automatically be passed through bind(). I tried using bind<R>(), since the slot knows what its result type should be, but gcc gave me ambiguous overload errors when bind<R> was given a function pointer. There is a little test program in the sandbox now too, at boost-sandbox/libs/thread_safe_signals/test/track_test.cpp. I encountered one problem with the bind syntactic sugar, described in the test program's comments. I found the error message unhelpful, therefore it must be a compiler bug :) - -- Frank -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFF5KRY5vihyNWuA4URAo2mAJ45BAJvKKl8rHPXGsgn7a2fjmDLVQCeLb0R ln07v8CCYxbwE7topMmhZQ4= =4kwy -----END PGP SIGNATURE-----

1 0

[multi index] warning when used in copy constructor: should be explicitly initialized
by Christoph 27 Feb '07

27 Feb '07

Using gcc with option -W I get the following warning when using a Boost.MultiIndex container inside a copy constructor: /usr/local/include/boost-1_33_1/boost/multi_index_container.hpp: In copy constructor 'boost::multi_index::multi_index_container<Value, IndexSpecifierList, Allocator>::multi_index_container(const boost::multi_index::multi_index_container<Value, IndexSpecifierList, Allocator>&) [with Value = ListenMonitor::ListenZeile, IndexSpecifierList = boost::multi_index::indexed_by<boost::multi_index::ordered_unique<boost::multi_index::identity<ListenMonitor::ListenZeile>, mpl_::na, mpl_::na>, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na>, Allocator = std::allocator<ListenMonitor::ListenZeile>]': liste.hpp:139: instantiated from here /usr/local/include/boost-1_33_1/boost/multi_index_container.hpp:204: warning: base class 'struct boost::multi_index::detail::header_holder<boost::multi_index::detail::ordered_index_node<boost::multi_index::detail::index_node_base<ListenMonitor::ListenZeile> >, boost::multi_index::multi_index_container<ListenMonitor::ListenZeile, boost::multi_index::indexed_by<boost::multi_index::ordered_unique<boost::multi_index::identity<ListenMonitor::ListenZeile>, mpl_::na, mpl_::na>, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na, mpl_::na>, std::allocator<ListenMonitor::ListenZeile> > >' should be explicitly initialized in the copy constructor The 'ListenMonitori::Listenzeile' stuff is from me. But the issue does not depend on that. I would assume that copying a multi index is not forbidden and using Is there a way to get rid of that warning? I am trying hard to produce warning free code, and in this case I am not sure if the warning is just ugly or if it is a real problem in the code. [gcc --version: gcc (GCC) 4.1.2 20061115 (prerelease) (SUSE Linux) ] I have attached a trivial example that demonstrates the issue. I have modified the example/basic.cpp: I have added a trivial class XXX that contains employee_set. When copying objects of XXX the above warning is issued by gcc. Best regards Christoph

3 9

Possible bugs in date_time package
by Gregory J. Sharp 27 Feb '07

27 Feb '07

I'm trying to use the date_time/ptime module to parse strings. Platform: MacOS X 10.4.7 on Intel Compiler: gcc 4.0.1 I have three potential bugs, so far. Perhaps I'm unreasonable in hoping that the following strings should be acceptable to time_from_string. 1. time_from_string( "2006-12-06" ) != time_from_string( "2006-12-06 00:00:00" ) 2. time_from_string( "2006/12/06" ) throws an exception (probably related to 1) even though time_from_string( "2006,12,06" ) doesn't throw. 3. time_from_string( "2006-12-06 " ) throws an exception because it passes a null string to str_from_delimited_time_duration(), which tries to read the first character of the string, even though there isn't one. I think 1 & 2 are probably caused by parse_delimited_time() doing a split on ' ' and getting nothing to pass to parse_delimited_time_duration(). -- Gregory J. Sharp email: gregor(a)mail.lepp.cornell.edu Wilson Synchrotron Laboratory url: http://www.lepp.cornell.edu/ ~gregor Cornell University ph: +1 607 255 4882 Ithaca, NY 14853 fax: +1 607 255 8062

2 4

multi_index uses serialization?
by Javier Gonzalez 27 Feb '07

27 Feb '07

Hi, I have a simple question. Isn't there a way to create multi_indices without using the serialization library? I don't intend to stream them! thanks, Javier Gonzalez

5 6

State of the Endian library
by Mateusz Loskot 27 Feb '07

27 Feb '07

Hi, What's the state of the Endian library announced [1] by Beman Dawes? Is there any plan to include it to the Boost collection? May be it's already there, but I can not find it? [1] http://thread.gmane.org/gmane.comp.lib.boost.devel/143279 Cheers -- Mateusz Loskot http://mateusz.loskot.net

1 0

shared_ptr and RTTI
by Florin Trofin 27 Feb '07

27 Feb '07

Hello, A few years back, boost::shared_ptr did not require RTTI. I recently discovered that it does now. I understand that there are some benefits that RTTI brings to the new implementation, but I would have expected that there is some compile flag that can be turned on/off. Florin.

1 0

[BGL] How to use an array in bundled property?
by Shufei Fan 27 Feb '07

27 Feb '07

I defined an double array inside bundled vertex property, and have trouble finding ways of accessing it. Neither can I find any where in the document an example on it. Your input will be much appreciated! my graph is defined as follows: struct Node{ // this is hard to access double mode[3]; }; struct Edge{ //no problem access them double nPixels; unsigned short nTh;}; typedef boost::adjacency_list< boost::vecS, boost::vecS, boost::bidirectionalS, Node, Edge> Graph; //Graph object colGraph Graph colGraph; I have no problem with accessing the 'nPixels' in Edge property, boost::property_map<Graph, double Edge::*>::type ew = get(&Edge::nPixels, *colGraph);//edge weight But I am in a great deal of trouble finding ways of accessing 'mode' in Node property. //compiler error here boost::property_map<Graph, double Node::*>::type nm = get(&Node::mode, *colGraph);//node modeerror message is : //---------------------------------------------------- c:\...\xxx.cpp(236) : error C2440: 'initializing' : cannot convert from 'boost::property_map<Graph,Property>::type' to 'boost::bundle_property_map<Graph,Descriptor,Bundle,T>' with [ Graph=boost::adjacency_list<boost::vecS,boost::vecS,boost::bidirectionalS,Node,Edge>, Property=double (Node::* )[3] ] and [ Graph=Graph, Descriptor=boost::property_map<Graph,double Node::* >::descriptor, Bundle=Node, T=double ] No constructor could take the source type, or constructor overload resolution was ambiguous//---------------------------------------------------- neither the following works(I added '[3]' in the parameter list): boost::property_map<Graph, double Node::*[3]>::type nm = get(&Node::mode, *colGraph);//node mode This time errors are raised in the 'property.hpp' file, they are: //---------------------------------------------------- c:\Boost\include\boost-1_33_1\boost\pending\property.hpp(40) : error C2825: 'PropertyTag::kind': cannot form a qualified name c:\Boost\include\boost-1_33_1\boost\graph\properties.hpp(198) : see reference to class template instantiation 'boost::property_kind<PropertyTag>' being compiled with [ PropertyTag=double Node::* [3] ] c:\shufei\VCProjects\edison\segm\GA.cpp(236) : see reference to class template instantiation 'boost::property_map<Graph,Property>' being compiled with [ Graph=Graph, Property=double Node::* [3] ]c:\Boost\include\boost-1_33_1\boost\pending\property.hpp(40) : error C2039: 'kind' : is not a member of 'operator``global namespace'''c:\Boost\include\boost-1_33_1\boost\pending\property.hpp(40) : error C2146: syntax error : missing ';' before identifier 'type'c:\Boost\include\boost-1_33_1\boost\pending\property.hpp(40) : error C2955: 'boost::type' : use of class template requires template argument list c:\Boost\include\boost-1_33_1\boost\type.hpp(14) : see declaration of 'boost::type'c:\Boost\include\boost-1_33_1\boost\pending\property.hpp(40) : fatal error C1903: unable to recover from previous error(s); stopping compilation//----------------------------------------------------thanks, _________________________________________________________________ Connect to the next generation of MSN Messenger http://imagine-msn.com/messenger/launch80/default.aspx?locale=en-us&source=…

1 0

faq How to compile graph?
by fcvcnet 27 Feb '07

27 Feb '07

Hi all, I know BGL just a few days. I need yours help. Now I only want to use the graph in the boost1.33.1 , how can I compile the only one component graph in vc6 or vs2005. So much thanks to your detail reply. Thanks again.

4 4

[lambda] beginner's woes
by Imre Palik 27 Feb '07

27 Feb '07

Hi, I try to use boost::lambda in the following function: string build_headers(shared_ptr<connection> conn, vector<wstring> headers) { if (headers.empty()) return ""; return accumulate(++headers.begin(), headers.end(), string(1, '\'') + conn->quote(headers[0]) + '\'', (_1 + ", '" + bind(&connection::quote, var(conn), _2) + '\'')); } and wisual studio gives me the following error message error C2664: 'boost::lambda::bind' : cannot convert parameter 1 from 'overloaded-function' to 'SOCKET' 1> Context does not allow for disambiguation of overloaded function Could anybody help me decypher this message? What is this 'SOCKET' and how does it come into the picture? Thanks ImRe

1 0

[signals] slot with no SlotFunction template parameter
by Frank Mori Hess 26 Feb '07

26 Feb '07

I'm implementing a more feature-full slot class with some of the things that have been discussed recently. In particular, it is templated only on signature. It uses a templated constructor to store the slot function in a pimpl object that is templated by both signature and slot function type. So now I'm about to rip out the SlotFunction parameter everywhere from thread_safe_signals, as there no longer seems to be any need for it. I feel like I might be overlooking something though. Is there any reason to keep it? -- Frank

2 2

[signals] thread-safe post-disconnect() cleanup
by Frank Mori Hess 26 Feb '07

26 Feb '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 We could implement a thread-safe post-disconnect callback by using the existing tracking functionality. That is, the connection body could hold a "cleanup" shared_ptr with a custom deleter. The custom deleter would run the user's cleanup callback. A weak_ptr copy of the cleanup shared_ptr would be kept just like it was any other tracked object. When disconnect() is called, the cleanup shared_ptr would be reset. If the signal is being invoked concurrently when disconnect() is called, the cleanup code won't get run until the last invocation is done with the slot and destroys its copy of the cleanup shared_ptr. BTW, I've added support for a two-parameter version of signals::track() to thread_safe_signals/track.hpp. It seems to work okay, the only thing I don't like about it is it implies there a coupling required between the value and the tracked shared_ptr, when really there isn't. - -- Frank -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.1 (GNU/Linux) iD8DBQFF3iTv5vihyNWuA4URAkzWAKDHL0MjcFGmAvKbWoBYmhNN5RiCMACgypCS SLJPrYcUaR0u4YTTS35lEZQ= =dCrK -----END PGP SIGNATURE-----

5 58

Re: [Boost-users] [boost::any] - reference vs pointer
by Benjamin Winkler 26 Feb '07

26 Feb '07

References can not be copied. Thus, you can't make boost::any contain a reference (which is what mean when you write boost::any<double&>, I assume). In your example, msgargs.push_back(dref) calls the any-constructor any(const T&) with T being double, and this copies the value, not the reference itself. When you call msgargs.push_back(dptr), it calls the any(const T&) with T being double*. The value is copied again, only this time, the value is a pointer. Benjamin Winkler >Hi all, > >I've been trying out boost::any in a messaging system to package up function arguments. A message >call is basically > >Message::Call(std::vector<boost::any>& args) >{ >try >...a few any_cast<..> >...fire the actual function... >catch >...etc >} > >It's all good, except for one niggling problem. Say I have a double-precision value that I need >changed inside the message, so it needs to be passed as a double* or a double& in an argument. The >problem is, if I create a double& inside a boost::any, it doesn't point to the right memory location >once I'm inside the message, and it looks like it's copying the original dereferenced value and >giving me a reference to the copy. If I create a double* inside a boost::any, I get a new pointer >inside the message like the reference, but because the pointer target is correct everything works. >The code looks like this: > >... >... >std::vector<boost::any> args; >double d = 1.5; >double* dptr = &d; >msgargs.push_back(dptr);//works fine... >double& dref = d; >msgargs.push_back(dref);//seems to copy d instead of a reference to d... >... >msg->Call(args); >... > >I can live with this, I just have to pass "large" objects by pointer so they don't get copied. BUT: > Can anyone tell me why references are creating a copy of the original and not a copy of the >reference, or how I could avoid the copying? Or am I doing something dumb? > >Damien >_______________________________________________ >Boost-users mailing list >Boost-users(a)lists.boost.org >http://lists.boost.org/mailman/listinfo.cgi/boost-users >

3 2

[FileSystem] Getting modified date and time from a file?
by Lawrence Spector 26 Feb '07

26 Feb '07

Does boost provide a portable way to do this? I didn't see an obvious way; I was hoping others might have ideas. Thanks in advance, Lawrence

5 5

[program_options] Documentation of configuration file format?
by Tarjei Knapstad 26 Feb '07

26 Feb '07

I can't seem to find any documentation on what kind of file formats are accepted when calling boost::program_options::parse_config_file(...) Am is missing it or is it missing? :) -- Tarjei

4 7

BoostCon'07 Registration Now Open
by info＠boostcon.com 26 Feb '07

26 Feb '07

The following message is a courtesy copy of an article that has been posted to comp.lang.c++.moderated,comp.std.c++,comp.lang.c++ as well. I'm pleased to announce that the official website for the inaugural Boost conference is up and registration is open: http://www.boostcon.com Due to the very high quality of the submissions we've received, our program committee has been able to assemble a lineup of speakers and topics the likes of which I haven't seen anywhere (http://www.boostcon.com/program) Combine that with the relaxed atmosphere of the conference venue (http://www.boostcon.com/location) and I think we can all look forward to a truly unique and memorable week of working and learning together. I invite you to visit http://www.boostcon.com/registration to register. -- Dave Abrahams Boost Consulting www.boost-consulting.com

1 0