November 2014 - Boost - lists.boost.org

Re: [boost] [review][sort] Formal review period for Sort library begins today, November 10, and ends Wednesday, November 19
by Niall Douglas 17 Nov '14

17 Nov '14

On 10 Nov 2014 at 0:39, Edward Diener wrote: > - What is your evaluation of the design? It's okay. I would prefer a design where there is a sorted_view which if std::copy gives an actually sorted copy. Obviously one would then specialise that scenario with a more directly optimised specialisation, and you'd also have the API presented here which looks like std::sort(). I'd also prefer to see fallback implementations for less than random iterators, right down to forward only iterators. > - What is your evaluation of the implementation? There seems to be an assumption that user supplied overloads and specialisations e.g. swap() don't throw exceptions. I think that unwise. BOOST_NOEXCEPT can test if something can throw. Explicit guarantees of what happens when something does throw needs to be in the documentation, per API. There seems to be a general lack of explicit C++ 11 support as well. In particular, I think all Boost libraries should use inline namespaces to implement ABI version management. Implementation isn't capable of dual build as standalone use without any dependency on Boost. I'd suggest my forthcoming Boost.BindLib library. > - What is your evaluation of the documentation? Lacks exception guarantees, one per API. Lacks scaling graphs near front so people can quickly evaluate usefulness compared to std::sort. I believe the author will fix this though. > - What is your evaluation of the potential usefulness of the > library? > - Did you try to use the library? With what compiler? Did you > have any problems? No. > - How much effort did you put into your evaluation? A glance? A > quick reading? In-depth study? Glance. > - Are you knowledgeable about the problem domain? Author of a very popular open source indexing algorithm library. > And finally, every review should attempt to answer this question: > > - Do you think the library should be accepted as a Boost library? Yes with these conditions: 1. Either change its name to Boost.Spreadsort, or do a much better job at a general sort library e.g. sorted_view, more than one sort algorithm not in the STL, more than random access iterator support. 2. Exception safety is not obvious, and explicit exception guarantees need to be made and stated per API in the documentation. 3. Explicit ABI version management in C++ 11 using inline namespaces. Forthcoming Boost.BindLib can do this for you, or it can be done by hand. It would also be nice if this library were also capable of standalone build without Boost, but this is not a requirement. Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/

2 4

Re: [boost] How do I authenticate myself?
by Felix Uhl 14 Nov '14

14 Nov '14

Andrey Semashev wrote: > Note that you can have anonymous read-only access if you change the URL a little: > git://github.com/boostorg/boost.git > That should probably be mentioned in the docs. Thank you very much, that was exactly what I needed. Should've probably thought about that myself. I agree, that should be mentioned. Especially under the point "try out modular boost". --- Felix Uhl ________________________________ Von: Andrey Semashev<mailto:andrey.semashev@gmail.com> Gesendet: ‎14.‎11.‎2014 14:17 An: boost(a)lists.boost.org<mailto:boost@lists.boost.org> Betreff: Re: [boost] How do I authenticate myself? On Friday 14 November 2014 10:55:36 Felix Uhl wrote: > I want to clone the current develop branch > > to verify that a bugfix to a trac item I opened > > does actually fix the bug. > > > It seems like this information should be easy > > to find on the boost webpage or wiki, but neither > > of both explain how to authenticate myself > > when using “git clone --recursive git@ > > github.com:boostorg/boost.git modular-boost >clone.log”. > > The output of this is “Permission denied (publickey)”, > > which seems to mean to me that I need a > > ssh key the server actually knows about or that I have > > to upload my public key somehow. > > > Have I just missed something on the wiki? As Boost is hosted on GitHub, all the options it offers also work with Boost. See here: https://help.github.com/articles/which-remote-url-should-i-use/ Note that you can have anonymous read-only access if you change the URL a little: git://github.com/boostorg/boost.git That should probably be mentioned in the docs. _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost

1 0

Re: [boost] How do I authenticate myself?
by Felix Uhl 14 Nov '14

14 Nov '14

Adam Wulkiewicz wrote: >FYI, instead of preparing a patch and attaching it to a bug report you >could consider working directly on the development version of a library >and propose a pull request with the fix on GitHub. The fix is not mine, I have opened a ticket, the author has written a fix and I just wanted to clone to check if the fix has fixed the issue. Thanks for the link, though, I'll make sure to check it out. --- Felix Uhl ________________________________ Von: Adam Wulkiewicz<mailto:adam.wulkiewicz@gmail.com> Gesendet: ‎14.‎11.‎2014 14:48 An: boost(a)lists.boost.org<mailto:boost@lists.boost.org> Betreff: Re: [boost] How do I authenticate myself? Andrey Semashev wrote: > On Friday 14 November 2014 10:55:36 Felix Uhl wrote: >> I want to clone the current develop branch >> >> to verify that a bugfix to a trac item I opened >> >> does actually fix the bug. >> >> >> It seems like this information should be easy >> >> to find on the boost webpage or wiki, but neither >> >> of both explain how to authenticate myself >> >> when using “git clone --recursive git@ >> >> github.com:boostorg/boost.git modular-boost >clone.log”. >> >> The output of this is “Permission denied (publickey)”, >> >> which seems to mean to me that I need a >> >> ssh key the server actually knows about or that I have >> >> to upload my public key somehow. >> >> >> Have I just missed something on the wiki? > As Boost is hosted on GitHub, all the options it offers also work with Boost. > See here: > > https://help.github.com/articles/which-remote-url-should-i-use/ > > Note that you can have anonymous read-only access if you change the URL a > little: > > git://github.com/boostorg/boost.git > > That should probably be mentioned in the docs. FYI, instead of preparing a patch and attaching it to a bug report you could consider working directly on the development version of a library and propose a pull request with the fix on GitHub. Here is a tutorial showing how this could be done: https://github.com/boostorg/geometry/wiki/Contribution-Tutorial It was made for Geometry but the procedure looks the same for any other library. Regards, Adam _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost

1 0

How do I authenticate myself?
by Felix Uhl 14 Nov '14

14 Nov '14

I want to clone the current develop branch to verify that a bugfix to a trac item I opened does actually fix the bug. It seems like this information should be easy to find on the boost webpage or wiki, but neither of both explain how to authenticate myself when using “git clone --recursive git@ github.com:boostorg/boost.git modular-boost >clone.log”. The output of this is “Permission denied (publickey)”, which seems to mean to me that I need a ssh key the server actually knows about or that I have to upload my public key somehow. Have I just missed something on the wiki? --- Felix Uhl

3 2

[sort] Review of the Sort library by Steven Ross
by Edward Diener 14 Nov '14

14 Nov '14

As review manager I am forwarding this review sent to me: - What is your evaluation of the design? - What is your evaluation of the implementation? Design an implementation look good to me. Spreadsort has a clean interface and the code is readable. I tried the algorithm out a few years ago in a project at work (our product in the EDA industry) for sorting 32-bit {x,y} coordinate data cast to a 64-bit integer, and for sorting 32-bit values within scanlines. The speedup was about 1.8x that of std::sort() for ~100M values. This translated into a 10-20% overall runtime reduction in that tool, with only a few lines of code changes. - What is your evaluation of the documentation? I haven't read through the documentation in detail, but I have read through the code and the comments so I feel I understand the basics of how the sorting algorithm works. - What is your evaluation of the potential usefulness of the library? I would say this library is very useful for specific tasks involving large datasets that need to be sorted, especially in cases where the data is already partially sorted and contains many duplicate values. For example, it could be useful for EDA applications, computational geometry, some types of databases, etc. It may be a fairly narrow user base but I think it could add significant value. - Did you try to use the library? With what compiler? Did you have any problems? I built and tested spreadsort on both linux/gcc and Windows/MSVC++. I was never successful in running the tests in Cygwin. It seemed to work as expected when I tried using it inside an EDA tool and caused no problems. - How much effort did you put into your evaluation? A glance? A quick reading? In-depth study? I spent some hours reviewing the algorithm a few years ago, and I provided Steve with feedback to improve the code. I was mostly interested in the integer sorting aspect at that time, but did experiment with floating-point numbers as well. I read through the code again recently to familiarize myself with it. - Are you knowledgeable about the problem domain? I would say that I'm knowledgeable of applications and uses of sorting large datasets, and algorithms in general. I don't have any specific experience in developing or testing sorting algorithms. - Do you think the library should be accepted as a Boost library? Yes, I believe Steve has put in enough work to make this a well designed, tested, and documented library. There are real numbers to show the performance improvements over other sorting algorithms. If this was part of Boost, I would probably use it in our product at my company. Frank Gennari

1 0

which submodule are boost headers in?
by Robert Ramey 13 Nov '14

13 Nov '14

This is making me feel really dumb. I want to remove a header in the /boost/ directory. This has been superseded by a header of the same name in the serialization library. boost/strong_typedef.hpp I expected to find a sub-module in modular boost which I can just select, get permission to make changes and upload my change - just remove this file. But I don't find a sub-module which seems to contain this header. So I need some help. Actually the simplest would be for someone who understands this and has appropriate privileges to just do the dirty work for me and remove the file. Failing that, someone can explain the procedure for finding the file, and getting the requisite privileges. Robert Ramey -- View this message in context: http://boost.2283326.n4.nabble.com/which-submodule-are-boost-headers-in-tp4… Sent from the Boost - Dev mailing list archive at Nabble.com.

2 2

[site] uBLAS documentation page link is broken
by Renato Florentino Garcia 13 Nov '14

13 Nov '14

The uBLAS documentation link[1] on Boost libs documentation page[2] is broken. The previous released version[3] is yet working nevertheless. Regards, Renato Garcia [1]: http://www.boost.org/doc/libs/release/libs/numeric/ublas/doc/index.htm [2]: http://www.boost.org/doc/libs/ [3]: http://www.boost.org/doc/libs/1_56_0/libs/numeric/ublas/doc/index.htm

2 1

[smart_ptr] explicit instantiation of boost::shared_ptr
by Petr Machata 12 Nov '14

12 Nov '14

Hi there, I got a bug report that the following code doesn't compile: #include <boost/tr1/memory.hpp> template class boost::shared_ptr<int>; int main(int, char**) {} The errors this gives under GCC 4.9 are as follows: In file included from /usr/include/boost/shared_ptr.hpp:17:0, from /usr/include/boost/tr1/memory.hpp:56, from foo3.cc:1: /usr/include/boost/smart_ptr/shared_ptr.hpp: In instantiation of 'typename boost::detail::sp_array_access<T>::type boost::shared_ptr<T>::operator[](std::ptrdiff_t) const [with T = int; typename boost::detail::sp_array_access<T>::type = void; std::ptrdiff_t = long int]': foo3.cc:2:23: required from here /usr/include/boost/smart_ptr/shared_ptr.hpp:663:22: error: return-statement with a value, in function returning 'void' [-fpermissive] return px[ i ]; This in particular is with boost 1.54. The problem is however present on master as well, where it also gives more errors relating to boost/core/demangle.hpp that I didn't look into. I went ahead to do some investigation in case the issue ends up being trivial. We can't use enable_if to disable a particular member function. Not unless it's a template. Making it a template would lead to a code like this: template <class U = T, class V = boost::disable_if <boost::is_void <boost::detail::sp_array_access< U >::type>, U>::type> boost::detail::sp_array_access< T >::type operator[] (etc. etc.); ... which is not C++98, and I don't have ideas how to C++98-ify it while also having it do what we want without breaking API. (The indirection through U may not be necessary, I'm not in fact sure, but that doesn't change anything.) One way that I can see is extracting element_type * px to a base class (say, boost::detail::sp_operator_at) that either does or does not define operator[] (which would then use the px). But that implies some amount of code motion, and not knowing what your position would be, I didn't want to blindly go ahead and start moving stuff around. So, would a patch to the effect described above be possibly acceptable? Thanks, Petr.

2 5

[Fusion] Lots of test failures on develop
by Louis Dionne 12 Nov '14

12 Nov '14

I'm getting a lot of test failures on the develop branch of Fusion on OS X. There are too many to paste the output here. Note that master is just fine. Are the unit tests being run automatically each time a pull request is merged? Louis

5 8

Boost.Build: new tutorial
by Vladimir Prus 12 Nov '14

12 Nov '14

Back in 2009, Boris Schäling wrote a Boost.Build tutorial that many deemed to be better than the tutorial in the official documentation. It took quite some time, but it's now part of Boost.Build, here: http://www.boost.org/build/tutorial.html I'd appreciate any comments. Note that previously, I planned to integrate this tutorial with the boostbook documentation. It's becoming clear that there's just no time for that, and the tutorial is added as HTML page. I haven't decided yet what to do with old tutorial in boostbook sources. -- Vladimir Prus CodeSourcery / Mentor Embedded http://vladimirprus.com

9 14