From mboxrd@z Thu Jan 1 00:00:00 1970 From: Kuba Ober Subject: Re: Filesystem errors Date: Wed, 17 Jul 2002 08:40:42 -0400 Message-ID: <200207170840.42655.kuba@mareimbrium.org> References: <200207160146.01440.cabernet@internode.on.net> <200207170811.44229.kuba@mareimbrium.org> <20020717161632.A2614@namesys.com> Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Return-path: list-help: list-unsubscribe: list-post: In-Reply-To: <20020717161632.A2614@namesys.com> List-Id: Content-Type: text/plain; charset="iso-8859-1" To: reiserfs-list@namesys.com On =B6roda 17 lipiec 2002 08:16 am, Oleg Drokin wrote: > Hello! > > On Wed, Jul 17, 2002 at 08:11:44AM -0400, Kuba Ober wrote: > > Did I mention that spinrite was about 100kb in size and written in > > assembler? > > This is not an advantage by any means in my view. It only means it is > x86 specific, most probably requires dos/windows, and hard to debug. You're right, but the guy simply kept it in assembler since mid-80's. And as far as I know him, and I know his coding style a little bit from=20 disassembling the early spinrite II, his code is mostly more legible than=20 poorly written C++. And the program, believe it or not, does not require debugging as it is. I = think that the current version works fine w/o any changes nor patches since= =20 about 3 years or so. He does one hell of a testing job, and he always bases= =20 on previous code that worked (most probably one of the reasons why it's sti= ll=20 in assembler). He is kind of assembly-maniac, but he has some windows utilities which are = in=20 assembler and do very cool things. Cheers, Kuba Ober