dev builds
themes manual
device status forums
mailing lists
IRC bugs
dev guide

Rockbox mail archive

Subject: Re: Behaviour of the "natural" sort: consensus reached?

Re: Behaviour of the "natural" sort: consensus reached?

From: Mike Holden <>
Date: Wed, 22 Apr 2009 19:17:12 +0100 (BST)

Al Le wrote:
> On 22.04.2009 14:46, Mike Holden wrote:
>> We specifically reported on how Windows treats 00file.txt and
>> 000file.txt,
>> and that it always treats more leading zeroes as "less", so 000file.txt
>> always sorts before 00file.txt, and decided that we should probably
>> follow
>> that convention, as other file browsers do the same on Linux (as does a
>> ksh shell on the Solaris box in front of me, for yet another data
>> point).
> But that would be in line with the agreement that "000file" and "00file"
> are equal with respect to natural sort. In that case "usual" sort
> (strcmp or strcasecmp?) is used as the arbiter which would then treat
> "000file" as "less."
> I'm not stating anything about the consensus, I just want to say that
> "000file < 00file" does not contradict to what kugel said (or wrote).

I don't plan to go over all this again, look in the archives!

You're correct in as much as you say, but you aren't answering the
original statement on this current thread. The original statement on this
current thread was that multiple leading zeroes would be reduced to a
single zero for sorting purposes, which I don't believe was the consensus
(such as it was).

Seriously though, this was done to death a few weeks ago. If anyone want
to pick up where that left off then fine, but I wouldn't advocate going
back to the beginning again! I'm not sure the thread ever reached a
definitive list of requirements, but I do believe it got pretty close.
Mike Holden - the place to shop for all manner of hand crafted
items, including Jewellery, Greetings Cards and Gifts
Received on 2009-04-22

Page template was last modified "Tue Sep 7 00:00:02 2021" The Rockbox Crew -- Privacy Policy