> On 12.9.2011 18:40, Charles Wilson wrote:
>> On 9/12/2011 10:29 AM, Teemu Nätkinniemi wrote:
>>> Everything seems to be working without crashing. I am trying to compile
>>> GCC 4.3.4 natively just to see if it can be done. One problem: at the
>>> moment MSYS2 is as slow as Cygwin is which in my mind makes it unusable.
>> Guys, TANSTAAFL. You can't get something for nothing: cygwin-1.7 (and
>> 1.5, before it) adds a LOT more posix compatibility functionality and
>> new features -- like unicode support -- than older versions. MSYS-1.0
>> is only SLIGHTLY faster than cygwin *of its day* -- e.g. 1.3.4 -- and
>> that is only because msys skipped a lot of the permission checking and
>> user/access emulation stuff that cygwin did.
>> Cygwin has become more processor hungry to support the new features. The
>> same will be true of msys-2: it will be only slightly faster than the
>> cygwin it is based on, and only if you short-circuit a lot of the
>> permission/access handling.
> This has been the exact reason I've started patching MSYS from the
> beginning and the ultimate goal being getting better support for
> threading and unicode. I still like the idea of MSYS2 if only to get to
> know Cygwin better and to see if there is any way to speed up MSYS2 when
> compared to Cygwin.
i would just like an msys2 faster for configure scripts than msys 1.
Configuring gettext, gnutls, etc... are too long with msys1. If msys 2 is
not faster than msys 1, i don't see the benefits of msys 2 for me.