If boost 1.61.0 has not been released, and its not too late, there are some warnings when building clang on OSX: https://svn.boost.org/trac/boost/ticket/12204 On Fri, May 13, 2016 at 12:00 PM, Michael Witten wrote:

On Fri, May 13, 2016 at 12:07 PM, Rene Rivera wrote:

...

Release 1.61.0 of the Boost C++ Libraries is now available.

These open-source libraries work well with the C++ Standard Library, and are usable across a broad spectrum of applications. The Boost license encourages both commercial and non-commercial use.

This release contains four new libraries and numerous enhancements and bug fixes for existing libraries.

New Libraries

* Compute: Parallel/GPU-computing library, from Kyle Lutz. * DLL: Library for comfortable work with DLL and DSO, from Renato Tegon Forti and Antony Polukhin * Hana: A modern C++ metaprogramming library, from Louis Dionne. * Metaparse: A library for generating compile time parsers parsing embedded DSL code as part of the C++ compilation process, from Abel Sinkovics.

For details, including download links, see http://www.boost.org/users/news/version_1.61.0

You can also download directly from SourceForge: http://sourceforge.net/projects/boost/files/boost/1.61.0/

Tom Kent has made pre-built windows binaries, and they are available at: https://sourceforge.net/projects/boost/files/boost-binaries/1.61.0/

To install this release on your system, see http://www.boost.org/doc/libs/release/more/getting_started/index.html

Thanks,

--The Boost release team

Vladimir Prus Rene Rivera Marshall Clow Eric Niebler Daniel James Beman Dawes

No, it's not released yet.

It's not released until there are associated with the files immediately obvious digital signatures that can be verified with a well known personal public key (like that of Vladimir Prus).

Please, quit fscking around.

_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost

-- Follow me on Github: https://github.com/vinniefalco

...

...

Please, quit fscking around.

Thanks for your reminder! The signed hashes file will be available later, most likely next Tuesday. Sadly, if you require this file before using 1.61.0, you would have to wait. While you're at it, please don't forget the tag in the git repositories (if that's automatic somehow, that's great, but it's not there yet in

On 13-05-16 18:54, Vladimir Prus wrote: the libs I check). Thanks! Seth

On 14 May 2016 at 9:57, Tom Kent wrote:

Ideally we could cryptographically sign the tags going into git as well, but I have even less experience doing/checking that.

You might as well sign every commit just as easily. Historically that introduced a lot of bloat (2Kb-4Kb per commit of signing due to the RSA key size), but in the AFIO v2 github repo I've been experimenting with the very new Ed25519 elliptic curves for signing. These add only dozens of bytes to each commit, and github correctly understands them as you can see by the purdy green "Verified" tag per commit at: https://github.com/ned14/boost.afio/commits/master Your big problem is going to be configuring the tooling for git signing on Windows - you need exactly the right combination of recent or beta versions of git, tortoisegit and gpg4win. I'd personally speaking recommend Boost leave off commit signing for at least another release or two until the tooling required for ed25519 curves comes out of beta, especially on Windows. I'd strongly recommend *against* using RSA keys for commit signing. The bloat introduced isn't worth the gain. The remaining question is whether signing tags is worth it. I personally don't think so, it's creates a false sense of security unless all committers are signing their commits. Better to SHA the zip archive as Rene already does. Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/

On Fri, May 13, 2016 at 4:54 PM, Vladimir Prus wrote:

Hi Michael,

On Fri, May 13, 2016, 19:01 Michael Witten wrote:

...

On Fri, May 13, 2016 at 12:07 PM, Rene Rivera wrote:

...

Release 1.61.0 of the Boost C++ Libraries is now available.

No, it's not released yet.

It's not released until there are associated with the files immediately obvious digital signatures that can be verified with a well known personal public key (like that of Vladimir Prus).

Please, quit fscking around.

Thanks for your reminder! The signed hashes file will be available later, most likely next Tuesday. Sadly, if you require this file before using 1.61.0, you would have to wait.

That said, the definition a Boost release is made by release managers, it's not a universal law of physics. Presently, that definition does not include any digital signatures. The signatures I included with a couple of release candidates were an experiment to see how many people appear to care (the answer was one, Tom), and how complicated it is (the answer is that GPG is quite a mess, especially on Windows or if you do not want your master key on a random cloud server). No decison is made yet.

You are welcome to push on this and make concrete suggestions, but the impact will be directly proportional to how polite your messages are worded.

So, because only Tom explicitly responded, it must be the case that only Tom cared; it must be the case that only Tom will ever care. Actually, Tom's emails are quite useful, because his own digital signature compounds the confirmation of your digital signature. I have made many concrete suggestions and I have been very polite; it doesn't seem to do anything. Your personal inability to find a convenient way to produce and present digital signatures is irrelevant; they are something that *must* be done in this day and age, and I cannot fathom why there is so much incredulity about that fact. Perhaps, Tom would be willing to take on the task of constructing the files and/or composing the digital signatures, at least until some agreeable release recipe can be established for other maintainers to follow mindlessly.

On Fri, May 13, 2016 at 12:40 PM, Michael Witten wrote:

On Fri, May 13, 2016 at 4:54 PM, Vladimir Prus wrote:

...

Hi Michael,

On Fri, May 13, 2016, 19:01 Michael Witten wrote:

...

On Fri, May 13, 2016 at 12:07 PM, Rene Rivera wrote:

...

Release 1.61.0 of the Boost C++ Libraries is now available.

No, it's not released yet.

It's not released until there are associated with the files immediately obvious digital signatures that can be verified with a well known personal public key (like that of Vladimir Prus).

Please, quit fscking around.

Thanks for your reminder! The signed hashes file will be available later, most likely next Tuesday. Sadly, if you require this file before using 1.61.0, you would have to wait.

That said, the definition a Boost release is made by release managers, it's not a universal law of physics. Presently, that definition does not include any digital signatures. The signatures I included with a couple of release candidates were an experiment to see how many people appear to care (the answer was one, Tom), and how complicated it is (the answer is that GPG is quite a mess, especially on Windows or if you do not want your master key on a random cloud server). No decison is made yet.

You are welcome to push on this and make concrete suggestions, but the impact will be directly proportional to how polite your messages are worded.

So, because only Tom explicitly responded, it must be the case that only Tom cared; it must be the case that only Tom will ever care.

Actually, Tom's emails are quite useful, because his own digital signature compounds the confirmation of your digital signature.

I have made many concrete suggestions and I have been very polite; it doesn't seem to do anything.

Your personal inability to find a convenient way to produce and present digital signatures is irrelevant; they are something that *must* be done in this day and age, and I cannot fathom why there is so much incredulity about that fact.

Perhaps, Tom would be willing to take on the task of constructing the files and/or composing the digital signatures, at least until some agreeable release recipe can be established for other maintainers to follow mindlessly.

For the record, the files containing the windows binaries have been hashed and signed. If you really need that, you could download one of them and extract the source from it as the entire source distribution is within them. Tom

On 15 May 2016 at 05:27, Boris Schäling wrote:

The Asio link http://www.boost.org/doc/libs/1_61_0/libs/asio/ is broken (on http://www.boost.org/doc/libs/1_61_0/). I get ERR_TOO_MANY_REDIRECTS with Chrome.

Fixed on the website, but the offline copy won't work until it's fixed in the build system - I had to use the 1.60.0 build system to generate working documentation.

On Sun, May 15, 2016 at 7:16 AM, Daniel James wrote:

On 15 May 2016 at 13:13, Daniel James wrote:

...

On 15 May 2016 at 05:27, Boris Schäling wrote:

...

The Asio link http://www.boost.org/doc/libs/1_61_0/libs/asio/ is

broken (on

...

http://www.boost.org/doc/libs/1_61_0/). I get ERR_TOO_MANY_REDIRECTS with Chrome.

Fixed on the website, but the offline copy won't work until it's fixed in the build system - I had to use the 1.60.0 build system to generate working documentation.

Of course, the offline copy will never work. But at least the redirect will go to working documentation on the website once the build is fixed.

What went wrong? -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail

On 17 May 2016 at 04:18, Rene Rivera wrote:

...

What went wrong?

Hmm.. To me it looks like ASIO has an incorrect redirect HTML file < https://github.com/boostorg/asio/blob/develop/index.html> (and it's always been that way). It also looks like the docs are not generated properly (missing images). Which could be from a variety of reasons.

Just looked into it, in the past the documentation was always built under doc/html, but because its location is now based on the target name it was being built in the wrong place. This is my fix: https://github.com/danieljames/asio/commit/6acd8f0a8d89c867cff7305cc9c4d5cce... Which tricks the toolset by setting a name property with the target directory. That's not ideal, it would be nice to have a proper way to specify the target build directory, but I'll go ahead with that if there isn't.

On Mon, May 16, 2016 at 8:08 AM, Karsten Ahnert < karsten.ahnert@googlemail.com> wrote:

...

The Asio link http://www.boost.org/doc/libs/1_61_0/libs/asio/ is broken (on http://www.boost.org/doc/libs/1_61_0/). I get ERR_TOO_MANY_REDIRECTS with Chrome.

The odeint link is also broken. I get

"404 Not Found

File "/home/www/ live.boost.org/common/code/../../doc/archives/boost_1_61_0/libs/numeric/odeint/doc/html/index.html " not found.

Unable to find file."

This looks like the docs where not included in the release :-( Likeliest cause is that they didn't get built. Maybe because of the subdir location. Really wish authors had checked this when I asked for them to test the new release archive build results months ago :-( -- -- Rene Rivera -- Grafik - Don't Assume Anything -- Robot Dreams - http://robot-dreams.net -- rrivera/acm.org (msn) - grafikrobot/aim,yahoo,skype,efnet,gmail