<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">On 01/30/14 10:00, Renato Golin wrote:<br>
</div>
<blockquote
cite="mid:CAMSE1kd9QpVBha-rw98MKdHcMR_xFfmE=C_QBX4jPG7pLjC3zg@mail.gmail.com"
type="cite">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote">On 30 January 2014 17:53, Behan
Webster <span dir="ltr"><<a moz-do-not-send="true"
href="mailto:behanw@converseincode.com" target="_blank">behanw@converseincode.com</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div bgcolor="#FFFFFF" text="#000000">
<div class="im">
<div><span style="color:rgb(34,34,34)">Quite frankly
most of the people I talk to think that "AArch64"
was a bizarre choice as the official name; most
people I've met think "arm64" is a much more
descriptive and sane name. The kernel community
has a habit of naming things what they think they
should have been called...</span></div>
</div>
</div>
</blockquote>
<div><br>
</div>
<div>If anything, "arm64" is easier to say out loud. Any
other argument is meaningless, since they'll all end up as
strings on a path or a file that will be auto-completed
with TAB.</div>
</div>
</div>
</div>
</blockquote>
Well, also arm64 is clearly an arm (even to the uninitiated) whereas
aarch64 could be something entirely different. "arm64" satisfies the
rule of "least surprise", and is the most obvious choice IMHO. ARM
missed the boat on this one I think. <br>
<br>
<blockquote
cite="mid:CAMSE1kd9QpVBha-rw98MKdHcMR_xFfmE=C_QBX4jPG7pLjC3zg@mail.gmail.com"
type="cite">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote">
<div>The kernel community has the habit of calling whatever
they want, and that's not always "sane". Least of all, it
causes confusion.</div>
</div>
</div>
</div>
</blockquote>
True; not always what is sane. But IMHO their track record is
arguably better than most. ;)<br>
<br>
<blockquote
cite="mid:CAMSE1kd9QpVBha-rw98MKdHcMR_xFfmE=C_QBX4jPG7pLjC3zg@mail.gmail.com"
type="cite">
<div dir="ltr">
<div class="gmail_extra">
<div class="gmail_quote">
<blockquote class="gmail_quote" style="margin:0 0 0
.8ex;border-left:1px #ccc solid;padding-left:1ex">
<div bgcolor="#FFFFFF" text="#000000">
<div class="im"><span style="color:rgb(34,34,34)">compilers
tend to follow architectural names; paths in the
kernel do not have to. They chose a name which makes
more sense to the wider development community.</span></div>
</div>
</blockquote>
</div>
<br>
</div>
<div class="gmail_extra">I don't buy this argument... we're
still using x86_64 (which is a pain to type), but what do I
know... ;)</div>
</div>
</blockquote>
Agreed. That is also a stupid name, though at least it is
descriptive and obvious. But since the port was done primarily by
Intel, they chose it... I don't believe ARM did the majority of the
arm64 kernel port.<br>
<br>
Another example of this is that Debian (and all Debian based
distros) still use the arch name of "amd64" for x86_64 which Intel
doesn't like at all. AMD's amd64 chips predate the current x86_64
Intel chips (though Intel had IA64 Itaniums at the time).<br>
<br>
None of this really matters though, and is all just opinion. The
reality is that the arch name can be different between kernel and
toolchain, and is different in this case whether we like it or
not...<br>
<br>
Having said all that, regardless of whether a name is better or not,
I would prefer a consistent naming convention (even though I prefer
amd64 and arm64, it would be more consistent to use x86_64 and
aarch64 since the powers-that-be chose those names)...<br>
<br>
Both "sane" and "consistent" is preferable, but I'm okay if it is at
least one or the other. :)<br>
<br>
Behan<br>
<pre class="moz-signature" cols="72">--
Behan Webster
<a class="moz-txt-link-abbreviated" href="mailto:behanw@converseincode.com">behanw@converseincode.com</a></pre>
</body>
</html>