<p dir="ltr"><br>
On Aug 16, 2015 12:50 PM, "Reindl Harald" <<a href="mailto:h.reindl@thelounge.net">h.reindl@thelounge.net</a>> wrote:<br>
><br>
><br>
><br>
> Am 16.08.2015 um 18:14 schrieb Roberto Ragusa:<br>
>><br>
>> On 08/16/2015 10:55 AM, Reindl Harald wrote:<br>
>>><br>
>>><br>
>>> no the architecture was created by Intel<br>
>>><br>
>>> AMD added the 64bit capabilities in a compatible way other than Intel itself tried with Itanium which was not able to run i686 instructions and later Intel was forced to license the AMD extensions<br>
>><br>
>><br>
>> Well, saying that the architecture was created by Intel is an evident<br>
>> "rewrite the history" exercise.<br>
><br>
><br>
> no, the main architecture is still Intel<br>
> AMD extended it and Intel adopted the extenisons<br>
><br>
> but *the point* is that calling it AMD64 is *plain wrong* because AMD/Inetl *and others* are using *the same* architecture and using a AMD64 suffix would imply "that binary is for AMD CPU's only" which is wrong<br>
><br>
><br>
>> You immediately contradict in the second sentence, where you describe the IA64 fiasco,<br>
>> and the adoption of AMD64 by Intel.<br>
>> Or maybe you think that "licensing the AMD extensions" is equivalent to "the architecture<br>
>> was created by Intel"?<br>
>><br>
>> Let's recap how it really went:<br>
>><br>
>> - Intel designed a new incompatible arch (IA64), it was useless at emulating the i386<br>
>> and was a substantial market failure<br>
>> - AMD designed the AMD64, as an extension if IA32<br>
>> - Linux was running on AMD64 immediately at day 0, as AMD had been giving around simulators<br>
>> for chips not created yet<br>
>> - Microsoft, who had already ported Windows to IA64, created an AMD64 version too<br>
>> - Intel tried to avoid the humiliating acceptance that their rivals did a better job<br>
>> than them, by going to extend the ia32 in a different way<br>
>> - Microsoft told Intel "I already did a port for you, you do not dare asking me another"<br>
>> - Intel released the "EM64T" architecture<br>
>> - Linus wrote a furious email saying that he had spent time studying the EM64T manuals<br>
>> only to finally realize that it could all have been replaced with just the sentence<br>
>> "it's AMD64" (differences are only found in little details)<br>
>><br>
>> Nowadays some use "AMD64" and some "x86_64", with Intel preferring the second for obvious reasons.<br>
><br>
><br>
> i know the history well, long enough in the business<br>
><br>
><br>
>> Having said that, the cost of change has got probably too high, so we will keep<br>
>> the current mix of AMD64 (used by BSD, Windows, Solaris, Java, Debian)<br>
>> and x86_64 (used by Linux, Fedora, SuSE, gcc)<br>
><br>
><br>
> it's not a point of "costs"<br>
> it's simply plain wrong<br>
><br>
> and just because Debian is here wrong as well as calling the httpd package "apache" don't make it right</p>
<p dir="ltr">OK you can stop this thread now. We aren't going to change in any case and it doesn't matter which of you is correct here.</p>
<p dir="ltr">If you really must argue further please do it off list.</p>
<p dir="ltr">josh</p>