March 2005 - Boost-users - lists.boost.org

Serialization library
by Przemyslaw Sliwa 22 Mar '05

22 Mar '05

All, I would like to ask you a question if someone of you has ever used the boost serialization library. I managed to compile with VC++ 7.1 the sample code from the example website http://www.boost.org/libs/serialization/example/demo.cpp. It went well, but it is giving some strange run time errors. I managed to compile just in the debug mode and it does not work with the Release one. I am getting some strange link errors. There are several libraries in the serialization module and basically there is nothing about it which one should be used. Can you help me with it? The examples use just the text archive. The serialization module is huge and I know there are several other modes, e.g. XML, etc. The problem is lack of any kind of documentation. Does anyone of you use other archive types than the text ones? Thanks for help in advance -- This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.

2 1

Re: map with minimal memory footprint
by Gottlob Frege 22 Mar '05

22 Mar '05

> 6. Re: map with minimal memory footprint (David Abrahams) > Date: Wed, 16 Mar 2005 09:45:34 -0500 > From: David Abrahams <dave(a)boost-consulting.com> > Subject: [Boost-users] Re: map with minimal memory footprint > To: boost-users(a)lists.boost.org > Message-ID: <uoedjd4z5.fsf(a)boost-consulting.com> > Content-Type: text/plain; charset=us-ascii > > Toon Knapen <toon.knapen(a)fft.be> writes: > > > Hi all, > > > > I'm using a rather big std::map which gives good performance but the > > memory footprint is a bit too big. So I figured to implement an std::map > > as an std::vector<std::pair<T1,T2>> but thus with the interface of an > > std::map on top of it. > > Can't be done. It has the wrong iterator invalidation properties. > You may be able to implement a useful associative container, but it > will never be std::map. > > -- > Dave Abrahams > Boost Consulting > www.boost-consulting.com > I believe you, but can you explain what you mean by "iterator invalidation properties". I wrote a vector_map once, and it was as "map-ish" as needed at the time, so I suspect it was just a "useful associative container", but I'm just curious as to what I was missing...

1 0

Re: [shared_ptr] get_deleter is for what?
by Mark Holloway 22 Mar '05

22 Mar '05

Ben Hutchings Wrote: >> I realise my requirements may be rather specific, but it seems >> shared_ptr has picked up some typeid baggage, which I can't afford. > Whyever not? Sorry - I wasn't playing the image size limitation game - it's just I'd rather class names were not revealed to the average image browser. Peter Dimov Wrote: >This problem should be fixed in the latest CVS version of boost::shared_ptr. >If you can try it and report back, it would be much appreciated. Will do. Thanks for the info Peter. Mark

1 0

[shared_ptr] get_deleter is for what?
by Mark Holloway 22 Mar '05

22 Mar '05

I realise my requirements may be rather specific, but it seems shared_ptr has picked up some typeid baggage, which I can't afford. I'm compiling using MSVC7.1 and for each object for which I use a shared_ptr, I gain an entry in my binary. For example, class Monty generates the following string in my exe: ?AU?$checked_deleter@VMonty@@@boost@@ This seems to be sucked in by this function in class sp_counted_base_impl: virtual void * get_deleter(std::type_info const & ti) { return ti == typeid(D)? &del: 0; } It isn't being used, and I can't actually see anything in boost using this, so maybe the compiler isn't doing a great job at stripping unused stuff? Anyone shed some light on this for me? Thanks Mark

3 2

boost parallel build, will it get better?
by Daniel Goller 22 Mar '05

22 Mar '05

Hi, i was wondering if there is a possibility boost (bjam?) will be supporting parallel builds better in the future, i was testing building with -j5 (-j5 is about the sweet spot at which distcc creates enough load on the helping XP3200 for this mac mini i work on) and boost was no longer able to successfully build, something it did before readding support for the user selected number of jobs. Basically im wondering what to tell users who start to notice we do not support their -jN settings ("bear with us it'll work in future" vs "we do not know if it will ever work") after having received notice that supporting make options might not be taken well by bjam (and being able to confirm this with distcc), while at the same time wishing i could take advantage of distcc myself. Is this something of concern to boost, or will it most likely remain at the state as it is in the 1.32.0 release? Thanks in advance for any information you can provide. Daniel

5 5

runtime polymorphism & passing and returning elements
by Kosta 22 Mar '05

22 Mar '05

The problem is not boost specific, but probably there is a boost specific solution to it... The usual way of passing and returning elements to/from a function/method is to use a templated function/method, e.g.: template< typename ITER_IN, typename ITER_OUT > void doSomething( ITER_IN inBegin, ITER_IN inEnd, ITER_OUT outIter, int param1, int param2, ... ) { while(inBegin != inEnd) { ... *outIter++ = ...; } } But what to do in case of (pure) virtual methods which can't be templated? One (neither pretty nor efficient) solution would be something like this: struct A { template<typename ITER_IN, typename ITER_OUT> void doSomething( ITER_IN inBegin, ITER_IN inEnd, ITER_OUT outIter, int param1, int param2, ... ) { const std::vector<int> out( doSomething_impl( std::vector<int>(inBegin, inEnd), param1, param2 ) ); std::copy(out.begin(), out.end(), outIter); } virtual std::vector<int> doSomething_impl( std::vector<int>& const in, int param1, int param2, ... ) = 0; }; // struct A Is there a better (boost-ish) solution?

1 0

BGL: adding vertex index property to listS listS adjacency list with with bundled attributes
by Sean Kelly 21 Mar '05

21 Mar '05

> Hi All, > > I'm trying to figure out how to add a vertex index property to the > following listS, listS adjacency list so I can use write_graphviz > > typedef boost::adjacency_list<boost::listS, > boost::listS, > boost::bidirectionalS, > > boost::shared_ptr<Sysgen::Visitable> > XTreeGraph; > > thanks for any help. > > - Sean > > Sean Kelly Ph.D, Sr. Software Engineer -- Xilinx Inc.

2 1

trim BUG??
by Groleo Marius 21 Mar '05

21 Mar '05

hello int main(){ string str1="\t \nhello world! "; trim_if(str1, is_any_of("\t") ); // str1 == "hello world!" trim_if(str1, is_any_of("\n") ); // str1 == "hello world!" cout << str1; } I get the following output: #> hello world! #> Isn't it supposed to be: #> hello world! #> ??? -- Regards, Groleo! .''`. # touch universe : :' : # chmod +rwx universe `. `'` # ./universe `- Debian - when you have better things to do than fix a system

2 1

Locking mechanisms other than scoped_lock
by Leon Mergen 21 Mar '05

21 Mar '05

Hello, I'm currently trying to achieve the following: A series of threads handle jobs (threads A). Of each of these jobs, in the end, a statistics object is placed in a container. This container will later be read-from by another thread (thread B), dedicated to continuously reading statistics data from other threads, and parse those statistics. Now, since it is very well possible a thread handles at least 1,000 jobs before statistics data is actually read from, I want it to hold a mutex lock to the data at all time. The class also has an internal volatile boolean, which flags whether another thread (thread B) wants to read from the container. If so, the writing thread (threads A) releases the lock, after which thread B will pick up the lock and write the data and resets the boolean flag. After releasing the lock, threads A will immediately try to re-acquire the lock. However, as far as I can see, for this approach, a scoped_lock will not be the most fantastic solution ? Or am I missing something really crucial about the scoped_lock being able to hold a lock outside a certain scope ? :) Thanks in advance for any suggestions about a locking type... :) Regards, Leon Mergen http://www.solatis.com/

6 7

Errors building CVS root sources with VC8
by Keith MacDonald 21 Mar '05

21 Mar '05

I have just done a checkout boost from cvs.sourceforge.net and tried to build with VC8 (cl version 14.0.40607.16) and Microsoft Platform SDK for Windows XP SP2, using the following commands: tools\build\jam_src\build.bat tools\build\jam_src\bin.ntx86\bjam "-sTOOLS=vc-8_0" This resulted in lots of these warnings: option 'Og' has been deprecated option 'Gs' has been deprecated ignoring unknown option '/MLd' ignoring unknown option '/ML' I read that Jam V2 should fix those, but I cannot see how to build it. More serious was the following compilation error: lexical_cast.hpp(150): error C2679: binary '>>' : no operator found which takes a right-hand operand of type 'std::basic_string<_Elem,_Traits,_Ax>' That file hasn't changed since last September, so am I wasting my time trying to use VC8? Thanks, Keith MacDonald

4 5