SixXS::Sunset 2017-06-06

latest aiccu sources
[pl] Shadow Hawkins on Monday, 01 October 2012 01:21:47
Where are they? On page https://www.sixxs.net/tools/aiccu/ source download has date 20070115. https://www.sixxs.net/tools/aiccu/changelog says the latest version is 20120905. https://github.com/SixXS/aiccu is empty. I'm asking because Fedora and openSUSE rpm packages contains some patches. It would be great if they could be added to upstream.
latest aiccu sources
[ch] Jeroen Massar SixXS Staff on Monday, 01 October 2012 09:04:23
Where are they?
Coming. We are planning on releasing AICCU to github soon.
I'm asking because Fedora and openSUSE rpm packages contains some patches. It would be great if they could be added to upstream.
As neither of these distributions have bothered to contact us (info@sixxs.net) we are not aware of any of the changes they are making.
latest aiccu sources
[pl] Shadow Hawkins on Monday, 01 October 2012 18:49:56
As neither of these distributions have bothered to contact us (info@sixxs.net) we are not aware of any of the changes they are making.
Let's say I'm from openSUSE. I stumbled upon aiccu package, that has some patches without any word about who did them and why, so I decided to do something about it.
Coming. We are planning on releasing AICCU to github soon.
Great! When it's there I'll take care of submitting openSUSE patches.
latest aiccu sources
[ch] Jeroen Massar SixXS Staff on Monday, 01 October 2012 19:14:12
Let's say I'm from openSUSE. I stumbled upon aiccu package,
that has some patches without any word about who did them and why,
so I decided to do something about it.
If you do not know why patches are there, then what is the point of those patches?
latest aiccu sources
[pl] Shadow Hawkins on Tuesday, 02 October 2012 11:10:05
aiccu-pthread.dif and aiccu-pthread.dif fixes build issues that appeared with newer versions of gcc. They are ok. aiccu_20070115-sigaction.dif replaces 'signal' with 'sigaction'. Created probably because 'man signal' says "The behavior of signal() varies across UNIX versions, and has also varied historically across different versions of Linux. Avoid its use: use sigaction(2) instead." Looks good. aiccu_20070115.dif changes in unix-console/main.c are to avoid warnings about unchecked return value. But I have no idea what is change in common/ayiya.c about.
latest aiccu sources
[pl] Shadow Hawkins on Tuesday, 02 October 2012 11:12:00
Why there is no preview and edit option? The second link should be to aiccu-linkorder.dif
latest aiccu sources
[ch] Jeroen Massar SixXS Staff on Tuesday, 02 October 2012 11:25:47
aiccu-pthread.dif and aiccu-pthread.dif fixes build issues that appeared with newer versions of gcc.
As the build system is quite different in the new AICCU, these are not needed any more.
aiccu_20070115-sigaction.dif replaces 'signal' with 'sigaction'. Created probably because 'man signal' says "The behavior of signal() varies across UNIX versions, and has also varied historically across different versions of Linux. Avoid its use: use sigaction(2) instead." Looks good.
On Linux, which tends to use glibc, signal() just calls sigaction(). signal() is supported on almost every platform in existence, sigaction() is not. As such, I really do not see any benefit for adding this change.
aiccu_20070115.dif changes in unix-console/main.c are to avoid warnings about unchecked return value.
Which warning? AICCU is compiled with full warnings on, and no warnings are emitted.
But I have no idea what is change in common/ayiya.c about.
Then why would we even bother looking at it?
latest aiccu sources
[pl] Shadow Hawkins on Tuesday, 02 October 2012 21:25:51
aiccu-pthread.dif and aiccu-pthread.dif fixes build issues that appeared with newer versions of gcc.
As the build system is quite different in the new AICCU, these are not needed any more.
That's great!
aiccu_20070115-sigaction.dif replaces 'signal' with 'sigaction'. Created probably because 'man signal' says "The behavior of signal() varies across UNIX versions, and has also varied historically across different versions of Linux. Avoid its use: use sigaction(2) instead." Looks good.
On Linux, which tends to use glibc, signal() just calls sigaction(). signal() is supported on almost every platform in existence, sigaction() is not. As such, I really do not see any benefit for adding this change.
sigaction is part of POSIX and is supported by all UNIX-like systems that are listed on https://www.sixxs.net/tools/aiccu/. Every documentation I found (not only linux; IEEE also) says that sigaction should be used rather than signal.
aiccu_20070115.dif changes in unix-console/main.c are to avoid warnings about unchecked return value.
Which warning? AICCU is compiled with full warnings on, and no warnings are emitted.
-Wunused-result
But I have no idea what is change in common/ayiya.c about.
Then why would we even bother looking at it?
To decide if it makes any sense. If not, it should be dropped.
latest aiccu sources
[ch] Jeroen Massar SixXS Staff on Wednesday, 03 October 2012 06:21:55
sigaction is part of POSIX and is supported by all UNIX-like systems that are listed on https://www.sixxs.net/tools/aiccu/. Every documentation I found (not only linux; IEEE also) says that sigaction should be used rather than signal.
As there is no advantage though in changing it, why would we bother? It is not that signal() is going away, nor is it adding any benefit to AICCU.
aiccu_20070115.dif changes in unix-console/main.c are to avoid warnings about unchecked return value.
Which warning? AICCU is compiled with full warnings on, and no warnings are emitted.
-Wunused-result
Can't replicate.
latest aiccu sources
[fi] Shadow Hawkins on Wednesday, 03 October 2012 14:43:47
This is a build log on a very new FreeBSD 9.1 release candidate using the unmodified 20070115 sources (there are no local patches for aiccu in the FreeBSD ports). The compiler is clang version 3.1 with -Wall flag turned on. As you can see there's really nothing to fix, I wonder why the Linux version needs the patches? Aiccu build log
latest aiccu sources
[ch] Jeroen Massar SixXS Staff on Wednesday, 03 October 2012 14:50:05
The compiler is clang version 3.1 with -Wall flag turned on.
Well, there are few things that clang emits there though, harmless things though. Nevertheless, from the AICCU changelog:
* Fix compiling on Mac OS X Mountain Lion with clang
And the full svn log message was even:
Did a compile on OSX Mountain Lion with "Apple clang version 4.0 (tags/Apple/clang-421.0.57) (based on LLVM 3.1svn)" which warns about a lot of things, fixes included, thus compile tested on OSX works too again
Thus I think it is fine here. But more real work first, then I'll try to hopefully get around to finalizing this new AICCU. compilation != functioning.

Please note Posting is only allowed when you are logged in.
Static Sunset Edition of SixXS
©2001-2017 SixXS - IPv6 Deployment & Tunnel Broker