I have just built the latest snapshot using mingw and
the tests work fine. However, I do the install, cd and
run ruby -v and it just sits there doing nothing. Obviously
an install problem.
…
The Win32 version of ruby is also installed (d:\apps\ruby\bin)
however with renamed executables. There are no conflicting
env variables that I can see. The Win32 version still works
fine.
Any thoughts?
Could it be that the .so dll’s from the other installation are being loaded
instead of the mingw ones?
Try to rename the directory where the Win32 version is located so the search
paths become invalid.
Same behavior, unfortunately.
It’s not a build problem. It all seems to work as expected
on my Solaris box …
This is quite strange. I am using MSYS package 1.0.8 which includes
bash 2.0.4. The problem with ruby hanging occurs with bash, not
with cmd.exe. cmd.exe works as expected with the mingw build of
ruby.
I guess that it is a bash problem. Has anyone experienced
anything similar?
I have no windows near to check, but I remember to have read in one of the
three small rtf files that ship with msys that all the msys unix tools
should only help developers with compilation, and must not be relied upon
as a runtime environment. So I don’t understand what you mean with
The problem with ruby hanging occurs with bash, not
with cmd.exe.
[…]
I guess that it is a bash problem.
… the msys unix tools
should only help developers with compilation, and must not be relied upon
as a runtime environment.
Earnie says: “It offers only a user land environment for MinGW development.”
Given that bash worked like bash, I assumed…
So I don’t understand what you mean with
The problem with ruby hanging occurs with bash, not
with cmd.exe.
[…]
I guess that it is a bash problem.
I can compile, test and install ruby 1.7.3 without a problem under MinGW.
When I run it, say “ruby -v”, it hangs, never to return. The 1.7.3 [mingw]
binary works fine under zsh and cmd.
Oddly, the 1.7.2 [mswin32] version works fine under bash.
