- Boost - lists.preview.boost.org

Regression testing: too many runners?
by Dmitry Moskalchuk 09 Feb '15

09 Feb '15

Guys, As you may know, I'm working on automatic regression testing of Boost on Android. I spent several weeks debugging it and now it seems to be close to the finish - i.e. we'll have soon fully automatic 24x7 regression testing with uploading all results to the Boost public FTP. As you can see, some results are already uploading on a regular basis - http://www.boost.org/development/tests/master/developer/summary.html (there are results of just few libraries at the current moment since I've limited testing by only those libraries to save time needed for debugging testing scripts; as far as I finish with my debugging, I'll enable testing of all libraries). However, I've realized that for thorough testing matrix of the variants become really big. Look, right now there are nine runners for Android, and this number is in fact very limited. They differs by target ABI (three arm variants, x86 and x86_64) and version of Android (API level 19 - Android 4.4 and 21 - Android 5.0). However, it has sense to test also on Android 4.0, 4.1, 4.2 and 4.3 since theirs market share is still large (see https://developer.android.com/about/dashboards/index.html) Add to that MIPS target ABI (not yet included to the testing) and multiple by two to have tests running with both default settings and with -std=c++11 - and you'll get really big total number of runners. I'm asking for advice from community. It's not the big problem for us to run all such tests in all variants, but I'm unsure if it will be acceptable for Boost community to see such wide table of runners. I'm afraid it will look like a flood. We also publish Android-only results on https://boost.crystax.net/master/developer/summary.html and we'll definitely display all results from all runners there. Please let me know if the same approach would work for http://www.boost.org/development/tests/master/developer/summary.html or should I limit somehow runners to be uploaded to the Boost FTP. -- Dmitry Moskalchuk

4 5

[githelp] Relative submodule paths..
by Rene Rivera 09 Feb '15

09 Feb '15

I've been trying to clean out the pull requests for the boost super-project and I ran into this one <https://github.com/boostorg/boost/pull/33>. And I don't really know if it makes sense, as I have personally not done a fork of the super-project. Is it a real problem? Should we change all the sub-project references? If yes, what are the likely impacts on others and existing release processes? -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

4 5

[polygon] voronoi diagram in a bounding box
by Dan Chen 08 Feb '15

08 Feb '15

Hi all, I am not sure whether this is the right place to ask this question. Basically, I want to compute the the voronoi diagram of a set of points in a bounding box, and output the voronoi cells. My idea is, after getting the voronoi diagram vd of the point set, for each voronoi cell in vd, compute the intersection of the cell and the bounding box. I am wondering whether this can be done in a better way with the boost libraries. Thanks,Dan

2 1

[voronoi] Issue with boost::polygon::voronoi_diagram<> as class variable
by duke 08 Feb '15

08 Feb '15

For some reason when I have a voronoi_diagram as a class variable (which it needs to be in this case), I get this error on compile: Error 49 error C2248: 'boost::polygon::voronoi_diagram<T>::operator =' : cannot access private member declared in class 'boost::polygon::voronoi_diagram<T>' I'm new to C# so it may be something obvious. - George

2 1

[MPL] Advice on how to proceed with a bugfix pull request
by Bruno Dutra 08 Feb '15

08 Feb '15

Hello everyone, I've found a fix to this bug <https://svn.boost.org/trac/boost/ticket/7769#no3>, which is based on BOOST PP library and thus should be supported across the various platforms, even though I just tested for a specific set up as I mentioned on my comment there. I've added some code snippets, from which it is straightforward to develop a patch, but I guess it would be best if I just issued a pull request myself. Since this is my first time contributing to MPL and to BOOST itself for all that matters, I'd like to first hear some advice on how to do this properly. For instance, to which branch should I issue the pull request? By the way, is this indeed the best way to proceed at all? Kind regards, *Bruno C. O. Dutra*

2 2

A proposal
by Mudit Sharma 08 Feb '15

08 Feb '15

After creating QR,LU and Cholesky for Boost..can we develop algorithms for determining which of QR,LU,Cholesky..etc. should be used in solving linear equations based on analyzing the running time and numerical stability specific to the problem. Is it feasible? Thanks!

2 1

Re: [boost] A proposal
by Ganesh Prasad 08 Feb '15

08 Feb '15

LU is already implemented in UBLAS. As of QR, SVD and Cholesky you might want to spend more time in debugging and testing for any one or two of the algorithms. https://github.com/GnsP/viennacl-dev/tree/master/viennacl/linalg you can find some useful sources here. Best wishes Ganesh

1 0

[c++11] MSVC does not define __cplusplus?
by Oliver Kowalke 07 Feb '15

07 Feb '15

Hi, I encountered that '__cplusplus < 201103L' evaluates on MSVC (14.0 ...) to true?! At least on GCC + CLANG this test works - what about MSVC? best regards, Oliver

4 4

[new libraries] [sort] New Library Check list updated
by Beman Dawes 07 Feb '15

07 Feb '15

As we started through the process of adding the sort library as a module, it became obvious the docs for how to do this were quite out-of-date. The best document for adding new libraries is https://svn.boost.org/trac/boost/wiki/ReleasePractices/ManagerCheckList#New…, and that section has now been updated. Although this document is written from the perspective of release managers, the New Library Checklist section serves as a roadmap for what is expected as a new library goes through the stages of being added to a Boost distribution. Comments and corrections welcome, --Beman

5 6

Serialisation: Is is_trivial<T> a sufficient precondition to bypass serialisation?
by Niall Douglas 06 Feb '15

06 Feb '15

A library user has raised an interesting question at: https://github.com/BoostGSoC13/boost.afio/commit/ac32ebfceb9a08b133e8c d47000d3b4807bc3e42#commitcomment-9574522 Basically, AFIO v1.3 has new pre-serialisation metaprogramming in response to Robert's comments on his Incubator about it being too hard to serialise and deserialise data with AFIO. AFIO will now consume any STL container and will implicitly auto expand to ASIO scatter gather buffers any: 1. Trivial type T 2. C array of trivial type T 3. STL container of trivial type T, including initializer_list. A free function, to_asio_buffers(T), can be specialised by external code to extend this with custom ASIO scatter gather buffers generation. Note that AFIO explicitly and intentionally expects that anyone interested in async file i/o will be doing their serialisation and endian conversion far away from AFIO code, so to_asio_buffers(T) is really for marking extra types to be treated as implicit auto expand. The user asks the question: Is is_trivial<T> a sufficient precondition for this auto expansion to be safe, or should is_standard_layout<T> also be required? The user notes that is_trivial<T> && is_standard_layout<T> == is_pod<T> which seems a little overkill to me. There is also the possibility that is_trivial<T> is too conservative. Some may argue that is_trivially_copyable<T> would be sufficient. Thoughts? Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/

5 7

[release-mgt] missing library
by Oliver Kowalke 05 Feb '15

05 Feb '15

Hello, some weeks ago we discussed/decided that the C++11-only version of boost.coroutine should be a new lib -> boost.coroutine2. Daniel has already created a git-submodul inf the boostorg-repo and I've added the sources to branches master and develop. Unfortunately I'm missing coroutine2 (link) in <boost>/lib/coroutine2, entries in <boost>/libraries.htm and <boost>/lib/maintainers.txt. File <boost>/website/blob/master/feed/history/boost_1_58_0.qbk already contains announcement of boost.coroutine2. Could you execute the missing steps to add coroutine2 to release 1.58 please? best regards, Oliver

3 3

boost interval arithmetic
by Павел Кудан 05 Feb '15

05 Feb '15

The difference in behaviour of comparison operators and basic arithmetic operators looks as probable misconception. Default and changeable behaviour of comparison operators consists in throwing exception in intermediate situation (intersecting intervals, ! (a>b) == (a <= b) is not applicable). Great decision for correct migration from float numeric code. The only and unchangeable behaviour of arithmetic operators consists in silent returning only one of possible intervals in intermediate situation (possible return is not single, multiple interval result). The source of mistakes and undesired behaviour at migration from float numeric. At least, throwing exception by arithmetic operators at intermediate situation would provide behaviour unity of comparison and arithmetic operators and functions. Best regards, Pavel Coodan

7 51

[release-mgt] Name of release manager's repository?
by Beman Dawes 05 Feb '15

05 Feb '15

We used to have a directory <boost-root>tools/release that contained scripts used by release managers. As part of the modularization effort, it was separated into its own GitHub repository, boostorg/release. It is not a submodule of the boost super-module as it is primarily of interest only to release managers. That seems to me to be a "good thing". But I'm wondering if the "release" name might be confusing, since it doesn't contain releases or release tags themselves. Would renaming it "release-mgt" or similar be clearer? --Beman

2 2

BOOST_HAS_PRAGMA_ONCE
by Stephen Kelly 05 Feb '15

05 Feb '15

Hello, As of commit 5b20f4910d8b6b48b0b6ad2df693d36edf54cf21 in the interprocess repo, the affected files seem to expect prior include of config by the the includer. Apparently this is not obvious to everyone, so it seems worth pointing out. Thanks, Steve.

4 10

Regression tests stopped on Saturday?
by Neil Groves 05 Feb '15

05 Feb '15

Hi, I know there has been a lot go good work going on updating regression testing, and I might have missed an email. I'm currently waiting for some tests to cycle, but there appear to have been no updates since Saturday. Is this expected? Thanks, Neil Groves

6 10

Boost regression testing on Android
by Dmitry Moskalchuk 04 Feb '15

04 Feb '15

Hi all, We've just published CrystaX NDK 10.1 (https://www.crystax.net/android/ndk) which is a drop-in replacement for the Google's Android NDK, and we've included prebuilt Boost 1.57.0 libraries there, to make life easier for those developers who want to use Boost on Android. The reason I write here is that in CrystaX NDK we were able to build Boost for Android without any changes in Boost sources, just because CrystaX NDK makes Android much more POSIX-conformant than original Android is, and Boost just works on top of it. If you're interested how we did it, please read description of CrystaX NDK release on https://www.crystax.net/android/ndk. Another important point is that we were able to run Boost regression tests on Android devices - again, without any changes in Boost sources, just providing own 'adbrunner' script as tests launcher. The only thing needed is to plug one or more Android devices to USB ports and/or run one or more Android emulators and then run script for Boost testing provided by CrystaX NDK. As long as those devices/emulators are available by ADB, Boost regressions tests will run on all of them. As far as I know, Boost is not yet officially supported on Android platform, mainly because of lack of regular regression testing procedure. I wonder if someone would be interested in our results and, if yes, would he/she like to participate in further Boost-on-Android improvements? -- Dmitry Moskalchuk

3 6

[mixin] Library name
by Borislav Stanimirov 04 Feb '15

04 Feb '15

To those unfamiliar with the Boost.Mixin - it is a new library submitted for review. The docs are here: http://ibob.github.io/boost.mixin/index.html The code is here: https://github.com/iboB/boost.mixin It has been suggested multiple times that perhaps the name "mixin" isn't appropriate, so I was wondering if I can find a better one. The current name, mixin, is, as the library itself, inspired by the mixins in the Ruby programming language. However, it doesn't do exactly what Ruby's mixins do (as mentioned here: http://ibob.github.io/boost.mixin/boost_mixin/introduction.html#boost_mixin… ) My current alternative ideas are: * EntityComponent - not great since it doesn't exactly follow the entity-component-system pattern, although it can be said that it allows the construction of entities from components. * DynamicMixin - to emphasize that it's not to be confused with the CRTP mixins. It still has "mixin" in it's name though. * Trait - since it resembles the traits in Self, Scala, PHP, etc. However, a trait is more a "compile time" feature (even when talking about PHP) * TypeComposition or TyCo - a very vague description * Polymorph - since it allows the composition of polymorphic types at run time, but that's pobably even more vague Other ideas are welcome. Thanks -- Borislav

3 2

A general qurey
by Mudit Sharma 04 Feb '15

04 Feb '15

What can be done to run boost on Dev-C++? What else can be used for boost?

3 2

[chrono][multiprecision][gsoc] Project proposal, looking for feedback and mentor.
by Damian Vicino 03 Feb '15

03 Feb '15

Hi, I’m collaborating with other students working in “Simulation Driven Development” (SimDD), mostly for embed systems. The main idea in SimDD is modelling complete systems (not necessarily software) and make simulations of them. Once the simulation has the expected results start replacing some sub-models (software pieces) by real application code. The pieces that are not replaced are used to provide testing facilities. The main fields where this SimDD is used are robotics and embed systems. Where you can clearly separate models in three groups: the software, the interaction (i.e. sensors) and the environment. The software is what you are developing, the environment is the things the software will interact with and the sensors are what info the software gets from the environment and how. When the software is ready, it interacts with the sensors and the simulated environment for testing. Once tested in simulated environments it can be implemented in real world. In this context, I work in time representation problems for discrete event simulation and I will like to propose some additions to the Boost library as GSOC project. — Adding a clock to chrono for virtual simulation time, so models and real code can both work using same library for time representation. — Adding multi base support for multi-precision library. — Adding the option for an accurate mode to multi-precision (meaning a failure should happen when the result requires to be rounded and, if possible, detected at compile time). — Supporting multi base multi-precision in chrono. — For sure, documentation and tests included for everything. Is anybody interested in these additions? Can you provide some feedback about them? Is anybody interested in be my mentor for GSOC in this project? In the last year, I had being working in a general-purpose Discrete Event Simulation library as a side-project that I hope to contribute to boost. The library was submit for registration to blincubator today and in the following days I will be uploading it. Best regards, Damian Vicino

2 1

[config] Visual Studio 2015 Preview CTP 5 FWIW
by Beman Dawes 03 Feb '15

03 Feb '15

VC++ in the 2015 CTP 5 is reporting version 19.00.22512, so needs to be bumped in compiler/visualc.hpp. C++14 digit separators and C++11 universal character names in literals are supposed to be implemented, but I haven't tested them. It would also be interesting if some of bugs in recently added C++11/14 features have been fixed. --Beman

1 0

Hi everyone
by PUNEET KOHLI 03 Feb '15

03 Feb '15

Hello guys. I came to know about your organization from the melange page.Your projects really got me interested. I am a very proficient in c & c++. I have been doing a lot of competitive programming for more than a year so , i can certainly code.I wish to apply to GSOC this year and would love the opportunity if it comes from your organization. Thus, i am looking for a way to contribute and work for you.

2 3

[preprocessor] integration of 'vctuple' branch breaks [accumulator] on msvc
by Jürgen Hunold 02 Feb '15

02 Feb '15

Hi! It seems that commit 01848af60d7639fd441b5d03462e848175cf1692 Merge: fd135a0 dc54fee Author: Edward Diener <...@...> Date: Fri Jan 30 23:16:08 2015 -0500 Merge branch 'vctuple' into develop completely breaks Boost.Accumulator on all msvc versions. I've some code breaking with msvc-12.0 and the msvc-10.0 tests in the regression matrix are all red. For example: http://www.boost.org/development/tests/develop/developer/output/teeks99-08c… fails with: .../accumulators/reference_accumulator.hpp(72) : error C2061: syntax error : identifier 'BOOST_PP_CAT' .../accumulators/reference_accumulator.hpp(72) : error C2947: expecting '>' to terminate template-parameter-list, found '>' .../accumulators/reference_accumulator.hpp(72) : fatal error C1903: unable to recover from previous error(s); stopping compilation Yours, Jürgen -- * Dipl.-Math. Jürgen Hunold ! * voice: ++49 4257 300 ! Fährstraße 1 * fax : ++49 4257 300 ! 31609 Balge/Sebbenhausen * jhunold(a)gmx.eu ! Germany

2 1

GSOC 2015 : Project on Concurrent Hash Tables
by Anmol Sood 02 Feb '15

02 Feb '15

Hi, I am Anmol Sood,a second year undergraduate student, pursuing Mathematics and Computing at Indian Institute of Technology,Delhi. I have been following the Boost mailing list for the last couple of months and I have been trying to understand some Boost libraries and enhancing my skills in generic programming at the same time. I saw the GSOC 2015 Boost page ( https://svn.boost.org/trac/boost/wiki/SoC2015) and the project on concurrent hash tables looks really interesting to me. I would like to know on the kind of pre-requisite skills required for the project and if I could share the progress I am making on writing code for the programming competency test here. Any tips/advice on helping me to get started on working for this project will be really helpful. Thanks.

2 1

Re: [boost] [type_traits] Rewrite and dependency free version
by Emil Dotchevski 02 Feb '15

02 Feb '15

On Mon, Feb 2, 2015 at 11:57 AM, Robert Ramey <ramey(a)rrsd.com> wrote: > this is the result of naming issues. the word array has been used for > boost::array, std::array and the built-in C/C++ array. They all ended > up in the same header so that one could find them when they were needed. > Ideally they should be separated, and boost::array's serialization should go into a boost::array header. As for serialization core dependencies, isn't it possible to define serialization function template overloads for user-defined types without including serialization library headers? E.g. aren't all function calls one makes in operator<< overloads dependent on the archive template argument? -- Emil Dotchevski Reverge Studios, Inc. http://www.revergestudios.com/reblog/index.php?n=ReCode

2 1

[sync] [thread] Faster mutex
by Antony Polukhin 02 Feb '15

02 Feb '15

Hi, One of the Mesa developers implemented a faster mutex: http://lists.freedesktop.org/archives/mesa-dev/2015-January/075260.html According to the sources of glibc, Mesa's mutex has a much smaller size (only 1 uint32_t vs 6 ints + 2 pointers), performs less actions before and after the actual syscall (no getting and setting of the mutex->__data.__owner, no type field checks) This gives an ~10% speedup. Just curious, maybe ideas from Mesa's mutex could be useful for some of the Boost libraries. -- Best regards, Antony Polukhin

4 4