August 2008 - Boost - lists.boost.org

[mpl] doc issues in release beta
by Eric Niebler 06 Aug '08

06 Aug '08

In the reference section of the MPL docs, there are a couple of issues. The biggest is that there is some new CSS wackiness that is making cross-links look like regular text until you hover over the links with the mouse. This is a regression from 1.35. The other issue is a broken link. In the reference section for mpl::for_each, there is a link to the docs for value_initialized that goes nowhere. This problem exists in the 1.35 docs as well. -- Eric Niebler BoostPro Computing http://www.boostpro.com

2 3

Re: [boost] out of date release results?
by Eric Niebler 06 Aug '08

06 Aug '08

Beman Dawes wrote: > Eric Niebler wrote: >> >> Beman Dawes wrote: >>> Eric Niebler wrote: >>>> Just spoke with Rene about this ... >>>> >>>> Eric Niebler wrote: >>>>> How come the "report time" on nearly every page of release >>>>> test results is dated July 15th? >>>>> >>>>> For example: >>>>> http://www.boost.org/development/tests/release/developer/summary_release.ht… >>>>> >>> >>> AFAIK, that's the wrong page to be looking at. The page I use to >>> make decisions is >>> http://beta.boost.org/development/tests/release/developer/summary.html >>> >>> >> >> Whew, thanks Beman! How do you get to that page? I go to boost.org, >> click on "Development", and then on "Release Summary". > > I don't look at www.boost.org, because I assume it applies to the > current release. > > Instead I look at beta.boost.org, because I assume it applies to the > release under development. Then why does www.boost.org have results for Boost.Accumulators (some of which are quite out of date), when Accumulators is new in 1.36. In particular, I never saw your response to this exchange from yesterday. I think this is a serious problem: Rene Rivera wrote: > Eric Niebler wrote: >> >> Oh whoa. Something really strange is going on with these pages. >> >> These results are old: >> http://www.boost.org/development/tests/release/developer/accumulators_relea… >> >> These results are new: >> http://www.boost.org/development/tests/release/developer/accumulators.html >> >> I can reach either page via different sequences of clicks. At first >> I thought that maybe one page represents results from 1.35, but >> that's not possible because accumulators wasn't in 1.35. >> >> What's going on? > > This sounds like the release view of the results is not getting > generated, on Beman's machine, and it's using some old set of > results. Hence only the developer view has the current results. Thoughts? <snip> >> >> OK, but that doesn't address the concern about test reporting. >> Currently, it takes a human (you, Rene, people on the boost-testing >> list) to manually verify that the results are being updated. > > Say we had a tool that looked at web pages to see if they had been > updated, and sent email if not. We would point it at > http://beta.boost.org/development/tests/release/developer/summary.html, > and set the required update frequency to say 24 hours. The email > address could be set to either the testing list, the site admins, > release managers, or whoever else was interested. That would not be good enough unless it checked every web page with test results on it. For instance, checking summary.html as you suggest would reveal nothing amiss, however this page on beta.boost.org is out of date: http://beta.boost.org/development/tests/release/developer/summary_release.h… -- Eric Niebler BoostPro Computing http://www.boostpro.com

1 0

[atomic count] boost on apple gcc 4.2
by David Abrahams 06 Aug '08

06 Aug '08

Well, first, I think you prolly ought to be posting to the developers' list, and second I think you need a tag in the subject like "atomic count" so that the right people will look at it :-) on Thu Jul 31 2008, "Mat Marcus" <mat-lists-AT-emarcus.org> wrote: > On Mon, Apr 28, 2008 at 3:22 PM, Peter Dimov <pdimov(a)pdimov.com> wrote: >> Daniel Oberhoff: >>> Hi, >>> >>> Has anyone successfully used boost with the "new" apple gcc 4.2? I >>> would like to since gcc42 has openmp, but I cant seem to be able to use >>> boost. I compiled and installed boost with apple gcc 4.2 by directly >>> using bjam with --toolset=darwin-4.2. When I try to use it I get an >>> error (detailed output of gcc below) which hints towards problems with >>> inline assembly. Actually, to get it compiled at all I had to set >>> -fasm-blocks. Am I missing a flag of sorts? Or is it "just" another woe >>> with apple's modified assembler? >> >> Is the new Apple GCC documented somewhere? I wasn't able to find anything >> related to a new assembler syntax or -fasm-blocks. Can you ask Apple to >> explain the issue? Stock g++ 4.2.x compiles the code, so the problem must be >> Apple-specific. > > Ok, I've got a little bit more information on the Apple gcc 4.2 > situation. Here is a snippet from an email dialog between a user and > an apple compiler engineer: > > user appears below with with prefix ">>" > apple compiler engineer appears below with with prefix ">" > >>> Thank you very much for the detailed reply. >>> Based on your information I deduce: >>> >>> 1) Using gcc 4.2 with 4.0 (c++) libraries is a supported >>> configuration by Apple >> >> Right. >> >>> I then conclude that the C++ ABI is fully compatible between 4.0 and >>> 4.2 >> >> Right. > > [snip, then user hypothesizes as to why there are problems with boost > under gcc 4.2/libstdc++ 4.0 on current Apple systems] > >>> This is the case because boost assumes that a 4.2 compiler implies >>> 4.2 libraries in flies such as boost/detail/atomic_count_gcc.hpp: >>> >>> #if (__GNUC__ > 4) || ((__GNUC__ == 4) && (__GNUC_MINOR__ >= 2)) >>> # include <ext/atomicity.h> >>> #else >>> # include <bits/atomicity.h> >>> #endif >> >> Yes, this is a bug or faulty assumption in boost. > > [snip] > >> The 4.2 library issue is a bug in Boost, not a problem with >> our compiler. I'd suggest raising the issue on one of the boost >> mailing lists and see what they say. > > Comments? > Mat > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users > -- Dave Abrahams BoostPro Computing http://www.boostpro.com

1 0

[sandbox][move] Regular requirement
by Michael Marcin 06 Aug '08

06 Aug '08

I'm using the current move library in the sanbox by daniel james (which I believe is the 4th move emulation library I've used now). It seems to be very good so far and makes writing code that handles both move and copy a breeze. I hope this gets finalized and ready for review soon because a consistently used move emulation library would be a great aid to writing efficient libraries that work well together. However the documentation says that objects have to model the Regular concept in order to model the Movable concept. This seems unfortunate as Regular puts a lot of requirements on types that don't seem necessary to gain moving. It requires Assignable which requires a copy constructor and it requires EqualityComparable which seems unnecessary. As a matter of fact I'm currently using the library with types I derive from boost::noncopyable on msvc 2008 and it is working wonderfully. Can the Movable concept be relaxed to require only what is truly necessary? Thanks, Michael Marcin

3 3

fixed pt binary lib
by Neal Becker 06 Aug '08

06 Aug '08

I have the beginning of a fixed-pt binary lib. Fixed-pt is popular in DSP applications. It is integer arithmetic, but keeping track of the binary pt. It is built on top of constrained_value. It is small, so I include it here:

8 22

Re: [boost] out of date release results?
by Eric Niebler 06 Aug '08

06 Aug '08

Beman Dawes wrote: > Eric Niebler wrote: >> Just spoke with Rene about this ... >> >> Eric Niebler wrote: >>> How come the "report time" on nearly every page of release test >>> results is dated July 15th? >>> >>> For example: >>> http://www.boost.org/development/tests/release/developer/summary_release.ht… >>> > > AFAIK, that's the wrong page to be looking at. The page I use to make > decisions is > http://beta.boost.org/development/tests/release/developer/summary.html Whew, thanks Beman! How do you get to that page? I go to boost.org, click on "Development", and then on "Release Summary". That takes me to: http://www.boost.org/development/tests/release/developer/summary.html That page *was* dated July 15, but Rene fixed something and now it seems to be up to date. Which is the correct site for results? Why do we have two? And why was boost.org directing me to stale results on www.boost.org instead of the fresh results on beta.boost.org? Rene? > The date on that is Sun, 3 Aug 2008 10:16:33 +0000. That's the last run > before I shut the machine down for maintenance. It is back up now and > should produce an updated report in an hour or two. > >> We should *all* be checking the release test results, especially the >> release manager. > > I look at the results several times a day during critical times, such as > when when preparing the beta. Glad to hear you're on top of this! >> How did this happen? Here are some hard questions. > > No clue. I have no idea how the results on the www.boost.org web site > get updated from the beta.boost.org web site. IMO, the test results should live in *one* place, and all the links should point to it. >> On what basis did we release the beta? > > The beta was released on the basis of a release report that was only a > few minutes old. > >> (That is, do we have release criteria? Written down?) >> Do those criteria have anything to do with the test results? > > Of course. The results speak for them; most of the remaining failures > are minor nits or even false positives. > >> Is anybody in charge of the test infrastructure? > > Rene works on the scripts, but a lot of the responsibility is > distributed or ill-defined. Hmm... >> Does that person look at the test results? >> What can we do to make sure this doesn't happen again? > > Take a look at ticket #2150. http://svn.boost.org/trac/boost/ticket/2150 > > That's one attempt at automatic tools to look for things that have gone > wrong and report them quickly. It will work on files, from the > repository or generated by the doc process. > > A similar tool that looked at web sites would be a nice QA addition. It > would check for the presence of specified files and verify their date > was recent, for a definition of recent specific to each file. Maybe > check file size, too, or even some content. OK, but that doesn't address the concern about test reporting. Currently, it takes a human (you, Rene, people on the boost-testing list) to manually verify that the results are being updated. >> Clearly, we need to delay 1.36 until we can get some fresh test results. > > The test results are currently being updated several (up to eight) times > each day. > >> And probably reopen the release branch for bug fixes. > > I posted a message several days ago indicating that the release branch > was open to bug fixes that are stable on trunk. OK. But I wonder if there are other people like me who were looking at stale results. :-P -- Eric Niebler BoostPro Computing http://www.boostpro.com

3 3

Re: [boost] A library for checking overflow in the built-in int types?
by Alp Mestan 06 Aug '08

06 Aug '08

It would be interesting yes. But what's up about the constrained_value library ? -- Alp Mestan --- http://blog.mestan.fr/ --- http://alp.developpez.com/ --- In charge of the Qt, Algorithms and Artificial Intelligence sections on Developpez

2 1

Re: [boost] FSM Review Announcement
by Andrey Semashev 05 Aug '08

05 Aug '08

Jeremy Pack wrote: > Andrey, > > I can host the documentation if you'd like - is it just reference.html > and state_machine.html? > > http://boost-extension.redshoelace.com/docs/boost/fsm/doc/reference.html > http://boost-extension.redshoelace.com/docs/boost/fsm/doc/state_machine.html > > Hosting it straight from the sandbox gives it high latency. Thank you very much. Yes, these are the only two documents in the documentation. There was also index.html one level higher, but it simply redirected to state_machine.html.

1 0

windows.h
by Brian E. Coggins 05 Aug '08

05 Aug '08

I just upgraded to Boost 1.35, and I now have a major problem with windows.h, thanks to the Boost.Threads library now invoking Boost.DateTime which on Windows #includes <windows.h>. Besides the min and max macros, there are other problems, like the fact that one of the platform headers does this: #define small char A third party library that I am using has a class called small. Oops. There are very, very many of these things. I'm afraid I am going to have to go back to 1.34 indefinitely. Google pulls up a lot of messages from other users having these troubles. I don't know very much about the Boost.DateTime architecture - I have used the threading library for years in its older version, but never the date/time library - but to the DateTime developers I ask: Is there any possible way that this could be reworked in the future to avoid bringing windows.h into the user's world? The macros and namespace pollution are a very serious problem. Defining NOMINMAX helps a little, but on things like the "small" macro I'm afraid I'm stuck. I see from mailing list posts that DateTime is not the only Boost library to do this, although certainly most of the libraries do not. It surprises me that Boost has such a nice policy on its own headers, yet allows implementations to do things like this. I know sometimes it is hard to avoid, but for windows.h - which is notorious for these problems - it would seem like it would be worth some contortions in the implementation to avoid this. I would be more than happy to compile and link some extra CPP files to keep the HPP files clean. To the Boost community at large I ask: What is the thinking on this? Thanks for your consideration. Brian

4 5

Re: [boost] out of date release results?
by Eric Niebler 05 Aug '08

05 Aug '08

Just spoke with Rene about this ... Eric Niebler wrote: > How come the "report time" on nearly every page of release test results > is dated July 15th? > > For example: > http://www.boost.org/development/tests/release/developer/summary_release.ht… We should *all* be checking the release test results, especially the release manager. How did this happen? Here are some hard questions. On what basis did we release the beta? (That is, do we have release criteria? Written down?) Do those criteria have anything to do with the test results? Is anybody in charge of the test infrastructure? Does that person look at the test results? What can we do to make sure this doesn't happen again? Clearly, we need to delay 1.36 until we can get some fresh test results. And probably reopen the release branch for bug fixes. -- Eric Niebler BoostPro Computing http://www.boostpro.com

1 0