- Boost - lists.boost.org

[C++ committee] auto_ptr removal in C++17 - request to comittie
by Artyom Beilis 07 Aug '17

07 Aug '17

Hello, If somebody on committee is on this list. I really want to do something to prevent auto_ptr removal. It is exceptionally useful smart pointer in C++2003 and it is needed by the code that needs to be compatible for both 2003 and 2017. It is needed by libraries that aim to be compatible with C++2003. In the real world there is a huge codebase that uses C++2003 why to break something that works - especially since it works quite well. It isn't "gets" that can't be used securely. C++ committee please stop breaking existing code even if std::unique_ptr much better it does not exist in C++2003. Artyom

3 2

Microsoft vs The Boost License
by Phil Bouchard 06 Aug '17

06 Aug '17

Hi, Just to put everybody in context: 1) Microsoft had invested into Corel so that they get rid of Corel Linux back in 2000 when I was working there: https://www.forbes.com/2000/10/03/1003corel.html 2) Microsoft once again copied my Fornux PowerCalc after I had presented it to them using some web interface: http://www.fornux.com/ With their Microsoft PowerToys: https://en.wikipedia.org/wiki/Microsoft_PowerToys#Included_applications 3) And now here I am with Microsoft trying to copy the idea behind root_ptr's node_proxy: https://github.com/philippeb8/root_ptr With their "deferred_ptr heap": https://github.com/hsutter/gcpp So for some reason I do not trust Microsoft. And I was wondering if the Boost license protects us from an idea behind a library we wrote. Otherwise this makes the Boost license not very useful. Thanks a lot for your help! -Phil

9 20

[BB++] Documentary - Take 3
by Phil Bouchard 06 Aug '17

06 Aug '17

In the following video, I am explaining how objects of a lower scope are being promoted to a higher scope (please fast forward to 10:15): https://youtu.be/OBnpFUVEVe4 I added 5 slides explaining how objects are being promoted and I've renarrated everything after 6:21. I'm keeping the beginning because I am targeting newbies on the subject as well. I hope that's a good explanation because sharing this information publicly is a point of no return for me. Sincerely, -Phil

1 0

Submodule name mismatches between master and develop
by Peter Dimov 05 Aug '17

05 Aug '17

Several submodules have different names on master and develop: - compute (libs/compute on develop) - couroutine2 (libs/coroutine2 on develop) - dll (libs/dll on develop) - qvm (libs/qvm on develop) - vmd (libs/vmd on develop) Should we normalize the names on develop? I think that after fixing .gitmodules, these submodules will appear as-if deinitialized, and would need a `git submodule update --init` to reappear under the new names. There are also libs/hana and libs/metaparse, which have consistent (although "wrong") names on both master and develop. Not sure if we need to bother with them.

6 8

LIB CSV
by Greger 05 Aug '17

05 Aug '17

During some time I have created this CSV library: <a href="http://www.multi.fi/~ghaga/libcsv/html/index.html">http://www.multi.fi/~ghaga/libcsv/html/index.html</a> (For processing of CSV tables) any use?good?bad? ty G

3 3

[locale] reporting MSVC issue
by Artyom Beilis 04 Aug '17

04 Aug '17

I noticed some failures in Boost.Locale using std backend and narrowed it down to broken Japanese_Japan.932 support of codecvt facet. I tried to report it on https://connect.microsoft.com/VisualStudio but it shown me " You are not authorized to submit the feedback for this connection. " How can I report the bug to Mircrosoft? Thank You, Artyom Below is the sample to reporoduce the issue: Works ok on MSVC 10 and below Fails on MSVC2017 - and it should work ---------------------- #include <iostream> #include <locale> #include <fstream> int main() { { // Japan in Shift JIS/cp932 char const *japan_932 = "\x93\xfa\x96\x7b"; std::ofstream f("testi.txt"); f<<japan_932; f.close(); } try { std::wfstream test; test.imbue(std::locale("Japanese_Japan.932")); test.open("testi.txt"); // Japan in Unicode std::wstring cmp = L"\u65e5\u672c"; std::wstring ref; test >> ref; if(ref != cmp) { std::cout << "FAIL" << std::endl; std::cout << "REF" << std::endl; for(size_t i=0;i<ref.size();i++) std::cout <<std::hex<< unsigned(ref[i]) << std::endl; std::cout << "EXP" << std::endl; for(size_t i=0;i<cmp.size();i++) std::cout <<std::hex<< unsigned(cmp[i]) << std::endl; } else { std::cout << "OK" << std::endl; } } catch(std::exception const &e) { std::cout << "No Locale: " << e.what() << std::endl; } }

2 1

libs/sync?
by Peter Dimov 04 Aug '17

04 Aug '17

We have a submodule libs/sync on the develop branch that isn't on master. What's its fate? Is it going to become a released library at some point, or should we remove it?

2 3

Bjarne's talk on C++20
by Phil Bouchard 04 Aug '17

04 Aug '17

I won't be able to go myself unfortunately but I'm just passing the word: http://www.technomontreal.com/en/news-center/events/bjarne-s-talk-upcoming-… Regards, -Phil

1 0

[release-tools] Doxygen 1.8.6 is rather old
by Vinnie Falco 03 Aug '17

03 Aug '17

Would it be possible to use the newest doxygen instead of the default 1.8.6 for Trusty? The latest version is 1.8.13, and Beast needs a newer version in order to generate correct output. <https://github.com/boostorg/release-tools/issues/4> Thanks

1 0

[BB++] Consensus
by Phil Bouchard 03 Aug '17

03 Aug '17

After the feedback I got back from the community, people agree they would rather improve C++ than seeing a new language taking off. I definitely agree but I would like professional opinions on how the following could be achieved: - to add instances of an object implicitly as each scope; - to add implicit references to these objects in top-level classes - to overload 'operator .' - to use 'auto' for member variables instead of 'decltype' Thanks for your help I really appreciate. Sincerely, -Phil

2 4