[Vm-dev] Squeak removed from Gentoo Linux

John M McIntosh johnmci at smalltalkconsulting.com
Thu May 27 21:11:55 UTC 2010

On 2010-05-27, at 2:06 PM, Igor Stasenko wrote:

> On 28 May 2010 00:02, Andreas Raab <andreas.raab at gmx.de> wrote:
>> On 5/27/2010 1:56 PM, Josh Gargus wrote:
>>> I loves me my bit-identical computation, but statically linking to the
>>> same version of the library doesn't necessarily give it.  For example, a
>>> video decoder may check hardware capabilities, and depending on what it
>>> finds decode the video on the GPU, using SSE, or just using the non-vector
>>> CPU instructions.
>>> There are some examples where we would make the case that the library
>>> needs to be statically linked in order to provide bit-identical
>>> computation(eg: fdlibm for Croquet math operations), but for things like
>>> JPEG I think that the argument is pretty weak.
>> Agreed.
> I really don't wanna put that on flame.
> Just a single word, which could say more than rest of arguments: UUID plugin.

The UUID plugin does not contain the UUID source code, it's a library call and expect it to use the 
system library. The fact that the system library is broken on some versions of some platforms is of course 
an issue, but frankly it's an issue for all software that uses UUID on that platform.  Also the behaviour of UUID is 
not bit identical on every platform the host operating system.

Obviously this is the problem if you rely on the platform code to do the right thing, but you choice here is 
wanting to enable bit identical behaviour versus ensuring Squeak is put into various Linux distributions? 

John M. McIntosh <johnmci at smalltalkconsulting.com>   Twitter:  squeaker68882
Corporate Smalltalk Consulting Ltd.  http://www.smalltalkconsulting.com

-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 2445 bytes
Desc: not available
Url : http://lists.squeakfoundation.org/pipermail/vm-dev/attachments/20100527/7029e644/smime.bin

More information about the Vm-dev mailing list