August 2004 - Boost - lists.boost.org

RE: [boost] Re: [Boost.Test] New testing procedure?
by Rozental, Gennadiy 31 Aug '04

31 Aug '04

> FYI, if there are no files in that directory of the > repository, cvs will generally not create the directory on > update/checkout. Not sure what's happening here, but you > might just have to add a dummy file to ensure it gets created. This directory have some files already. > That said, Boost tests should be able to be run from a > read-only CVS tree, so if you're creating files you probably > ought to do it elsewhere (like /tmp/...). Could you show how to do this portably? Gennadiy.

2 1

toolset proliferation
by Robert Ramey 31 Aug '04

31 Aug '04

I would like to see a little more order/organization in the creation of new toolsets. For example in the current CVS we have for Codewarrior Cw Cw-9.2 Cwpro-9.2 Cw-8.3 Cwpro-8.3 This is only an example. The same problem is occurring in gcc and mingw It creates a couple of problems a) we're not all testing with the same Jamfile. This leads to a lot of confusion. b) If I have a jam file which needs special treatment for a particular compiler, I have to included a line for each toolset - and that can change in the future creating another obscure bug to track down. c) The documentation on the "getting started" page is way out of sync with the toolsets in existences. A somewhat related problem - or maybe the cause of the above is the fact that the Jamfile doesn't have access to the compiler version. I'm not sure what to do about this. But, here's a suggestion for consideration to start the discussion. Right now we pass a variable to bjam with the toolset name. e.g bjam -sTOOLS=cwpro-9.2 How about the idea Bjam -sTOOLS=cw -sCOMPILER_VERSION=9.2 -sSTD_LIBRARY=stlport The "base" toolset would fill in variables COMPILER_VERSION and STL_LIBRARY_VERSION with defaults - probably the currently shipping version. Or a "base" toolset could require one or the other of the above switches. Also the "base" toolset could include a "config_check" target what would double check that compiler version invoked is the same as the one that is specified by the switch. This would catch the case where one has one compiler in his path and tries to build for another. This would have the following advantages: a) a smaller number of toolsets b) a greater likelihood that we're all testing the same thing c) documentation more likely to be insync with the actual toolsets available The recently re-organized stlport.jam "add-in" with all stlport stuff factored out so it can be "added-in" to any library is a good idea. Unfortunately it has a bug but that's fixable. This would allow us to specify usage of STLPort (or any other library) without creating a new toolset. I guess we sometimes use: bjam "-sTOOLS=cwpro-9.2 msvc" to handle multiple compilers in one shot. If its important to keep this feature then we could use Bjam "-sTOOLS=cw msvc" "-sCOMPILER_VERSION=cw-9.2 msvc-6.0" -sSTD_LIBRARY=cw-stlport or something similar. Just a small suggestion to keep the pot boiling. Robert Ramey

2 1

[multi_index] Question about 'modify'
by Vladimir Prus 31 Aug '04

31 Aug '04

I have a question about the 'modify' method of ordered indexes in the multi_index library. I have an iterator obtained from an ordered index and would like to modify the pointed element. I'm pretty sure that the modification won't change the order, won't create collisions and so on. The documentation suggests to use either 'replace', or more efficient 'modify' method. Why can't I do the following: typedef variables_t::nth_index<1>::type index_t; index_t& index = variable_declarations.get<1>(); index_t::iterator existing = index.find(vd); const_cast<Variable_declaration&>(*existing) .make_external(); ? TIA, Volodya

2 3

user defined facets in date_time library
by Markus Schöpflin 31 Aug '04

31 Aug '04

While trying to make date_time tests compilable on tru64cxx65, I have come accross an issue with user defined facets (namely date_names_put and all_date_names_put) that leaves me scratching my head. Is a user defined facet required to be default constructible? Or does it need a constructor of the form "foo(size_t refs = 0) : locale::facet(refs) {}"? [lib.facets.examples] doesn't say so but most examples I could find either have a default constructor or a constructor taking a reference count. Markus

2 3

copyright question (bjam examples)
by John Maddock 31 Aug '04

31 Aug '04

Some of the examples in bjam v2 are actually taken verbatim from the QT libraries examples, and currently have no copyright or license declaration on them at all. Vladimir asks: "Will /* This file is part of Qt and is available under the Q Public License, available from http://doc.trolltech.com/3.3/license.html */ be enough to make inspect tool happy. If we're really going to be extra-picky, I don't think I can add any "Copyright" string to the file I don't own." Is adding such a string acceptable? Clearly we need to do something, but what? Thanks, John.

4 4

Something odd with the meta-comm tests?
by David Abrahams 31 Aug '04

31 Aug '04

It looks like some of them are being run against the mplbook branch of boost/mpl. How else to explain the lack of mpl/aux_/has_xxx.hpp reported in http://tinyurl.com/524ur? -- Dave Abrahams Boost Consulting http://www.boost-consulting.com

2 1

RE: [boost] Re: Review Wizard Status Report
by Glen Knowles 31 Aug '04

31 Aug '04

> From: David B. Held > David Abrahams wrote: > > > [...] > > What is the point of these IDs? It seems like it adds busywork for > > reviewers and no especially useful information. Now we're talking > > about making them more complicated... > > Maybe they have to do with a nagging feeling that Boost > Libraries don't all have a canonical name, and even when it > seems like they do, people tend to call them whatever they > feel like. Perhaps we should either go with a numerical > index (which the DBA in me prefers), or insist that all > libraries have a canonical name and that everyone use that > name to refer to it (perhaps to make it easier to filter > messages containing the name). There is also the case of a library having multiple reviews, the serialization library comes to mind. Glen Knowles Centor Software Corporation

2 1

Is boost::fsm suitable for embedded systems?
by Thomas Stevens 31 Aug '04

31 Aug '04

Hi, forgive me for my ignorance but I'm a boost newbie... The rationale document for boost::fsm states that the library should meet a number of requirements, numbered from 1 to 9. Especially, the sixth requirement says "support the development of arbitrarily large and complex state machines.". Ok, this may be good - I guess it is in fact good. But I'm a bit anxious that this implies that the library is therefore not suitable for enviroments where no unnecessary overhead is acceptable, like in many embedded applications. However, the seventh requirement indicates that my worries are unjustified since it reads: "allow the user to customize all resource management so that the library could be used for applications with hard real-time requirements". And further down the Rationale document one can read a section on the Resource usage of Memory which looks good. So, just to be sure, is boost::fsm suitable for the development of embedded applications where constraints such as memory footprint is an issue? (Hopefully, someone can answer this based on the experience of using the library or by knowing the code - not by reading the documentation (like me ;-)). Thanks, Thomas S.

4 7

VC8.0 and std::list
by Pavol Droba 31 Aug '04

31 Aug '04

Hi, I have encountered a problem with vc8.0 std library. I was able to fix it, but I'm not sure what is correct according to standard. The problem can be narrowed to one simple case. In vc8.0 method iterator std::list<>::erase(iterator First, iterator Last) returns a 'singular' iterator when the list is empty after erase operation. Because I'm dependant on the returned iterator to make insertion at this point later on, I had to make an explicit workaroud. I would expect it to return .end() iterator in this case. Is this behavior correct according to standard or not? If not, is it reasonable to submit bug report? Thanks, Pavol

3 3

Workflow, process flow, mapping of an activity diagram to code.
by Batov, Vladimir 31 Aug '04

31 Aug '04

In my spare time I've been fooling around with an implementation of an infrastructure that would allow workflow to be configured at run-time. A well-defined process flow is clearly easier to configure/implement at compile-time using the conventional means (coding). Say, bool my_process_flow() { if (call_a() == success/true) { return call_b(); } else { return call_c(); } } However, to be able to quickly adjust to changing customer requirements and/or procedures we might decide to provide a set of building blocks (like call_a, call_b, call_c) that could be assembled into a certain process flow at run-time. Maybe even give that ability to the customer themselves through, say, GUI interface. So, the infrastructure boils down to an Activity class that has an operation associated with it and can have two sub-activities to be called when the operation succeeds or fails. For example, // Create an activity with call_a() as an operation. Activity a(call_a); // Register call_b to be called when call_a succeeds. a.success() = call_b; // Regiter call_c to be called when call_a fails. a.failure() = call_b; call_a, call_b, call_c can be Activity instances themselves. So, we are building hierarchical tree-like structure to represent a certain process flow. We can easily support concurrent activities like a.success() = Concurrent(call_b)(call_c); Loops can be supported in a similar fashion (via pre-defined functiors). All mostly based on boost::function (a very handy gadget I must say. Kudos to the authors). Now after such a long and boring introduction to my actual question. :-) Is there any interest in having/developing such an infrastructure under the boost umbrella? Any input is welcomed even if you think it is a crappy idea. Best, Vladimir.

2 1