Re: Building headless Bitcoin and Bitcoin-qt on Windows
@Nitro
After what me and Ron came through, perhaps it is a good idea to expend your post to include the following import pointers:
* Make sure that your MSYS bin directory precedes any other Windows system directory, like c:\Windows\System32 (this is necessary if you want to use the Unix find and not the Windows variant).
* The WSAPOLLFD structure bug in MinGW (not really bug, it is just missing)
* UnregisterWaitEx() problems when not setting _WIN32_WINNT
(also look for a reference to the two above here: http://stackoverflow.com/questions/20957727/boostasio-unregisterwaitex-has-not-been-declared)
Also, Ron mentioned that makefile.mingw is not longer used and there are more problems (Perl is unnecessary and so is Python too I believe, although I'm not sure about Python). It would also be nice if you explain a bit about the inner workings of the building process (i.e. why this utility or command), because as it is now it is sending a wrong message to people. I don't only want to know what to do, but also why to do it.
Don't get me wrong, I really appreciate this post and I think it is very useful, but it can be made so much better.
(we could work together a bit and also update the build_msw file, should be much more accessible than this forum thread)
Jori.
After what me and Ron came through, perhaps it is a good idea to expend your post to include the following import pointers:
* Make sure that your MSYS bin directory precedes any other Windows system directory, like c:\Windows\System32 (this is necessary if you want to use the Unix find and not the Windows variant).
* The WSAPOLLFD structure bug in MinGW (not really bug, it is just missing)
* UnregisterWaitEx() problems when not setting _WIN32_WINNT
(also look for a reference to the two above here: http://stackoverflow.com/questions/20957727/boostasio-unregisterwaitex-has-not-been-declared)
Also, Ron mentioned that makefile.mingw is not longer used and there are more problems (Perl is unnecessary and so is Python too I believe, although I'm not sure about Python). It would also be nice if you explain a bit about the inner workings of the building process (i.e. why this utility or command), because as it is now it is sending a wrong message to people. I don't only want to know what to do, but also why to do it.
Don't get me wrong, I really appreciate this post and I think it is very useful, but it can be made so much better.
(we could work together a bit and also update the build_msw file, should be much more accessible than this forum thread)
Jori.
Msys bin folder should automagically get precedence in path when running msys (no need to add it manually to windows environment variables):
Code:
$ which find
/bin/find.exe
/bin/find.exe
WSAPOLLFD and UnregisterWaitEx() issues appear only when using MinGW default toolchain, this is one of the reasons I stress on using mingw-builds.
makefile.mingw was used in pre v0.9 versions, a lot of altcoins are still based on older versions, so I think it is useful not to remove 0.8.6 notes (yet).
Perl and python were remnants and have been removed (you still need perl when configuring openssl, but msys-perl gets installed with msys-base).
I agree I could have been a bit less cryptic, but this can be improved
Yep you are right, my db did not build correctly. I had the wrong version of gcc
Anyways, this thread has been very helpful for me, thanks! I'll be sending a btc tip your way.
Anyways, this thread has been very helpful for me, thanks! I'll be sending a btc tip your way.
Thanks for the tip