- Boost - lists.preview.boost.org

Re: [boost] [compute]/ Futures Some remarks
by Antonio Colino 15 Jan '15

15 Jan '15

> > On 01/14/2015 10:46 PM, Antonio Colino wrote: > > I have a flawless implementation of this concept I am very proud of. Is any one willing to critique it? > > On 01/14/2015 10:46 PM,?Thomas Heller?wrote: > > Wanna share it? Which concept are you refering to? > > I was hoping to share this lib, if i could generate any interest in a copyable hpc version of packaged_task > On Jan 2015 20:36:40,?Thomas Heller?wrote: > Absolutely! Note that there are already a few out there doing exactly what > you described (HPX and Habanero++ come to my mind here). What would be the best way to arrange this?

1 0

Marking explicit failures -- need help
by Andrzej Krzemienski 15 Jan '15

15 Jan '15

Hi All, I would like to make some regression tests as expected to fail in explicit-failures-markup.xml, but I have the following problem. Not sure if this is specific to Optional tests: http://www.boost.org/development/tests/develop/developer/optional.html I have two "columns" with the same toolset label "gcc-4.4.7" from tester "Sandia". On one the test passes, on the other: it fails. This is because on one C++0x is enabled. But I cannot tell it from the toolset label alone, and therefore I cannot make the correct entry in failures-markup.xml: either I mark too much or too little. Does anyone know away around it? Can you mark failures based on something else than toolset name? Regards, Andrzej

1 1

Re: [boost] [compute]/ Futures Some remarks
by Antonio Colino 15 Jan '15

15 Jan '15

On 01/14/2015 10:46 PM, Antonio Colino wrote: > I have a flawless implementation of this concept I am very proud of. Is any one willing to critique it? On 01/14/2015 10:46 PM,?Thomas Heller?wrote: > Wanna share it? Which concept are you refering to? I was hoping to share this lib, if i could generate any interest in a copyable hpc version of packaged_task

2 1

Re: [boost] [compute]/ Futures Some remarks
by Antonio Colino 15 Jan '15

15 Jan '15

On 01/14/2015 10:46 PM, Antonio Colino wrote: > I have a flawless implementation of this concept I am very proud of. Is any one willing to critique it? On 01/14/2015 10:46 PM, Thomas Heller wrote: > Wanna share it? Which concept are you refering to? I was referring to both type traits on lambdas and an implementation of a tasking library that is allocator free, that you can reset function arguments reuse the task, reuse threads, reset arguments on lambdas, copy tasks while running and still always get accurate information. Thread safe and lock free. Basically a scheduler, task implementation that dare i say rivals boost bind and std::packaged_task. As it can bind any set of arguments to a function , reuse the same function to create new states . It is the core of a bigger lib I have been working on that sends these tasks across machine bounds ( hpc ) and onto the gpu.

1 0

Tests that enable <warnings-as-errors>on
by Peter Dimov 15 Jan '15

15 Jan '15

Some test suites, for example Pool and Units, use '<warnings-as-errors>on' in their project requirements, to test for a warning-free compilation. Pool, f.ex. has project : requrements <toolset>gcc:<warnings-as-errors>on <toolset>msvc:<warnings-as-errors>on ; (among others.) This, however, is propagated to dependencies, so that in [ run test_threading.cpp : : : <threading>multi <library>/boost/thread//boost_thread ] Boost.Thread is also built with '<warnings-as-errors>on', which fails. (In addition, some other tests also fail because of warnings.) So I tried to use '<warnings-as-errors>off' only for some tests, but I was unable to come up with a syntax that worked. Putting -<toolset>msvc:<warnings-as-errors>on in the test requirements didn't (no effect). <toolset>msvc:<warnings-as-errors>off didn't (conflict). -<warnings-as-errors>on didn't (no effect). -<warnings-as-errors>on <warnings-as-errors>off didn't (no effect). Is there a way to override the project requirements from the run rule?

2 1

[thread] fails to build on Cygwin with weird errors
by Peter Dimov 15 Jan '15

15 Jan '15

Vicente, I'm trying to get the pool tests up and running (they fail across the board, at least on develop), and one of them requires Thread, which fails to build under g++/Cygwin. The errors are of the kind ..\..\..\libs\thread\src\win32\thread.cpp: In member function 'bool boost::thread::start_thread_noexcept()': ..\..\..\libs\thread\src\win32\thread.cpp:337:127: error: 'struct boost::detail::thread_data_base' has no member named 'id' uintptr_t const new_thread=_beginthreadex(0,0,&thread_start_function,thread_info.get(),CREATE_SUSPENDED,&thread_info->id); which leads me to think that, on Cygwin, you're using the POSIX header with the Win32 source file.

2 2

Re: [boost] [compute]/ Futures Some remarks
by Antonio Colino 15 Jan '15

15 Jan '15

I have a flawless implementation of this concept I am very proud of. Is any one willing to critique it?

2 1

Re: [boost] boost interval arithmetic
by Marc Glisse 14 Jan '15

14 Jan '15

(please keep the conversation on the list) On Wed, 14 Jan 2015, Павел Кудан wrote: > Dear Marc, > > Thank you for your answer, I also tried to contact Mr.Gubenko, have no information about other maintainers. Please, excuse my bad English, I will try to explain in more detail, please, ask if something will be uncertain. I thought the maintainers would be the 3 authors. > Boost interval library was planned as extension of mathematical functions on float/double/double long or other singe numeric values to operations on intervals. The problem is that intervals have some specific. For example, if float value can be either > either < or = some other float value (statement ! (a>b) == (a <= b) is always true), for intervals it is not correct. Intervals can also overlap. That means that it is not possible for pair of overlapping intervals to say certainly which of them are greater than other. It is the intermediate situation, which is impossible with original single float values. > > And boost interval class has comparison operators and user can expect that they will behave in similar way with usual float comparison operators. > > For that reason, the behaviour of that comparison operators was made changeable and depends on some kind of policy, defined by user. Default behaviour is to throw exception when such intermediate situation (not possible for single float values) takes place. > > That was very wise solution, especially useful in very probable case, when some programmer will try to adapt some existing code with float calculation just changing type of value to boost interval. He will get exception in than exceptional situation - when usual float comparison operator is not applicable any more. > > That is OK. > > > > But as well boost interval class defines mathematical functions and operators. And user can expect that they also will behave similarly to usual float comparison operators. But great disappointment and multiple possible errors which is hard to locate is waiting for him here. Because it is not so. Arithmetic operators behave in completely different manner. > > Let's look on of most basic operator =/ and see what it is doing. > > But before let's see what is division of intervals and what is it's result. > > Boost interval class has two function - division_part1() and division_part2(). It is correct as when you divide one interval with another, the result will be single interval or a pair of intervals. The last situation takes place when in statement A / B, interval B cross a zero. That is not normal for intervals to have a pair of intervals as a result of division, but it is exceptional for single float numeric division. > > Now let's get back to /= operator and watch what he will do in such case. > > Let A = [1,2] and B = [-1, 1] > > and let's see what we will get after A /= B and what we should get. > > The true result of division in that case will be a union of pair of intervals [-infinity, -1] and [2, -infinity] Actually, [-inf,-1][1,+inf] > Interval /= operator, naturally, must return one interval. It is, obviously, impossible in this case - to return one of two. Again intermediate situation, not applicable to single float calculations. Again, throwing exception must be a result to let user know that he does not know what he are really doing. The same behaviour, as comparison operators. But it is not so. > > Instead of such natural behaviour, /= operator of boost interval class are doing impossible thing. It silently decides for itself which one of two parts of whole result to return and kill another part of whole solution somewhere under the scene. > > The result of A /= B will be [-infinity, -1]. Not correct result - as it is not complete. One of to parts is missing. Did you really observe that? If so, it is a bug. [1,2]/[-1,1] is supposed to give [-inf,inf] (and it does indeed, in my tests), which certainly contains the right value ;-) If you have a testcase that shows differently, it would be good to open a bug report on trac. > Very very bad operator behaviour. User get not correct result and even does not know that it is so. It will be hard to find when this error is located, and when it will be done, library user will not think that he should know better what he are doing and learn some basics of interval arithmetic - no, he will accuse boost interval as part and boost as whole for bugs in libraries implementation. U nfairly , of course, but who cares. > > > > To correct it, at least, arithmetic operators and functions of interval class must behave similarly to comparison operators - throw exception in situations, not applicable for single float calculations. > > That is what I meant when said: "The difference in behaviour of comparison operators and basic arithmetic operators looks as probable misconception." > > > > I know, it is hard to understand, especially because bad English, but I hope it is important and hope it will help. -- Marc Glisse

2 2

Re: [boost] Futures (was: Re: [compute] Some remarks)
by Antonio Colino 14 Jan '15

14 Jan '15

Concerning std::futures and promises, my thought is wrapping an immutable state in a task, nither future or promise, but with elements of both. Creating a mutex from an asm lock cmpxchg wrapped in c ( available most platforms ) to a shared_ptr gives you a copyable mutex. Not that it matters, but if the state manager ensures a swap every time the arguments are updated, and the state has a ref count destructor, a value storage, function/lambda/member func/functor storage, the effect is an immutable state. once the args are set nothing can change but the execute function or function brackets, a flag in the state ensures that the func will only execute once ( all flags set by single asm instruction xchg, wrapped in c ). this task , or promise if you like, is now copyable, and doesn't require an allocator as the internal state is set by setting the arguments and the state ptr and can be created then. Every time the task/ promise gets copied the shell only exists to point to the state. Having a generic function that wraps any func call in a std::function<void()> using a lambda , works nicely for a basic way to fill a queue generically. In the end you only have two exposed objects. A task and a scheduler. No mutexes, futures, promises, or any locking mechanisms of any kind. The task implementation I'm speaking of also accepts lambdas, and can reset there arguments. The thought is creating the shell (task , promise, whatever) once. Then setting the arguments many times. Each time it would internally create a new state, the user still has to take care of the memory, less the ref count get to zero, but a simple vector does nicely. Since it is all wrapped in the copyable task/promise, it is transparent to the user. So... task(any function in with or without args), task.set_args, scheduler.queTask, and go. The scheduler is just a thread grouping dictated by the user or by the hardware concurrency as a default. The result can be retrieved directly through the task or it can create a future as well or fill an existing one with the original future ref in the state. Don't really need it but... if it must be future/promise. Biggest problem i had was dealing with the lambda and functor case intuitively and cross platform. Functions and member functions were a snap.

1 0

PDF versions of Quickbook generated documentation - Do you like the serif font?
by Paul A. Bristow 14 Jan '15

14 Jan '15

I've noted that for documentation using the Quickbook toolchain, we use a serif font (Times Roman?) for the PDF, but a sans serif for the equivalent HTML version. Is there a reason for this? Should they not be the same font? (So that they look the same?) Which do we prefer? Should we change either or both? Paul --- Paul A. Bristow Prizet Farmhouse Kendal UK LA8 8AB +44 (0) 1539 561830

2 2