bug-fileutils
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: GNU rm's man page doesn't name chattr as a "See Also" program


From: Craig Carey
Subject: Re: GNU rm's man page doesn't name chattr as a "See Also" program
Date: Fri, 07 Feb 2003 16:55:46 +1300



I seem to have lost the case (argument) rather rapidly and unexpectedly.



At 03\02\05 11:36 +0000 Wednesday, Richard Dawe wrote:
>Hello.
>
>Craig Carey wrote:
>[snip]
>> Eighthly, the GNU project is defending its observable 'in action'
>> trait of taking a stubborn stand over data loss. Normally it just
>> causes it (Glitchix).
>[snip]
>
>I don't think the Linux kernel is part of the GNU project. That's one reason
>why it's called GNU/Linux and not just GNU.

I left unfinished, this text

>Which would be cold brained enough to permit the problem to remain.

We are getting fairly dense errors here (when Mr Szmidt states his views
[see below]) and so a Penguin is out since a bird of the air. Also the
animal absolutely can't be animal of the water since there is still no
admission that users are relevant. To symbolise the discussion it would likely
be an earthy animal, but not the GNU.
It moves from above and downwards: it asserts views that get worse the
more considered by persons above. So it would not even have vestigages of
wings like a Penguin.

I produce a suggestion on he animal at the bottom. It drops one spot,
and observation that GNU/Whatever is there, but not two: the use made
use of that.


>
>If a document on data recovery of ext3 filesystems exists, I would expect it
>to tell you about the chattr command.
>
>Having been a victim of minor data loss from a Linux kernel bug that caused
>corruption in FAT filesystems, I can sympathise with you. It still surprises
>me that kernels relatively late into the stable series can cause data
>corruption. I got burnt by 2.2.9 (I think).
>
>The man pages are automatically generated. They're just a summary of the
>output of "foo -h", listing the switches and directing the user to the info
>documentation. So the information about chattr would belong in the info
>documentation.
>
>I agree it would be good if the info documentation to mention reasons other
>than file permissions for why the file can't be removed. Perhaps there should
>be an OS-specific section in the "Changing file attributes" node, which could
>be linked to from the rm node.
>
>I'll write a documentation patch for coreutils (which is the merger of
>fileutils, sh-utils, textutils).
>
>Regards,
>
>--
>Richard Dawe [ http://www.phekda.freeserve.co.uk/richdawe/ ]


That document completely destroyed the plausibility of my suggestion
I could slur from a criticism from Mr Szmidt to the GNU project, and
it quite differs from Mr Szmidt's own style in that users, stats,
facts, adjustability, action, distress at data loss.


---------------------


Below I reply to 2 of Mr Szmidt's message comments.

Once again you deleted what I wrote and seemed to have no success
in finding something I wrote that was disagreed with.


At 03\02\05 15:54 +0100 Wednesday, Alfred M. Szmidt wrote:
><sniped lots of stuff that I didn't bother to read, please keep mails
>short, people don't have time to read essays>
>
>   Can you remember the page number holding the policy or the date you
>   received the policy from the FSF/GNU lot ?. I guess not.
>
>What policy first of all?  You seem not to understand that chattr is
>_not_ a GNU project.  It might be licensed under the GPL, but that
>does not make it a GNU project.
>

Don't forget to put the idea 'man page' into your paragraphs otherwise
you have not got the world divided into 2 part.

If Mr Szmidt did not answer my question. Linux users do not want
unhelpful documents and when the view wrongly upholds that but it
had not basis of reasoning, I hope for Mr Szmidt to explain the
reasoning. But a later message shows that there is a false
appearance of the thread seem to proceed as if a casual friendly
conversation.


------------------
>Sixthly, I read a page by the Mr Richard Stallman which said that
>volunteers were need to create the GNU documentation. What starts
>as hopes can later become FSF policies.
------------------

I was trying to say I was commenting adversely on GNU while sticking
close to Mr Szmidt's text, but the very potentiality of casting the
whole GNU project down a drain pipe if Mr Szmidt's influence were to
be copied, makes it too unlikely that I could have been commenting
on its leaders.


>Just like Linux being GPL'ed doesn't mean that it is a GNU project.
>To be a GNU project you have to assign copyright to the FSF. You can
>read more about it at: http://www.gnu.org/evaluation/evaluation.html
>
>   Is there a principle against withholding details from FSF mailing
>   lists?.
>
>Yes, since withholding information makes it impossible to help
>people.

You neither modeled readers thinking, by asserting their wishes using
text, nor tried to consider something so real as to be statistical.

Beyond a doubt that others readers can agree that a modelled or real
clump of GNU/Linux users want to have at some point, a concession
that they exist. Mr Dawe might improve that. You could probably
do better work at the FreeBSD project, E.g. FreeBSD's TCP stack is
used by all Linux programs running in the Linuxulator. The Linux
stack loses lone TCP packets without reporting any error which makes
automatical maximisation of speeds impossible in Linux. Discard the
Linux stack, keep the GNU code and an infrequent man page, run heavy
duty TCP code in FreeBSD that runs 35% slower in Linux, and the
developers never knew a worse world that they wanted to help unless
dealing with principles or abstractions, etc.

>
>And next time, please keep your mail a bit shorter, I surely do not
>have the time to read a long essay about totally unreleated subjects.


Isn't that a comment that you can add to any message you want ?.
There is no need to answer that.

More here from the same: and this is untrue:

>
>It is quite impossible to document chattr, since as I said it differs
>from system to system.
>

The instantaneousness with which that can be rejected is obvious.
Or is it ??.
Certainly there is no reasoning.

Are grounds for not documenting chattr (in the rm man page) in that
paragraph ?.

No. It says it restates something.
Here is the earlier statement [below], and grounds are not here too

Here is the earlier statement asking for the reasoning Mr Szmidt
might have been using:

At 03\02\04 12:27 +0100 Tuesday, Alfred M. Szmidt wrote:
>   A defect in the 'man' reference manual document of the 'rm' program,
>   is that it does not name the 'chattr' program.
>
>Why should it even name chattr?  chattr differs from file-system to
>file-system.
>

That has not got grounds in it that substitute for reasoning for
that conclusion that the 'rm' man page shan't name chattr.

                       Mold drops spots

It is seeming bleak so far.

I never found out how many GNU GPL variants there are.
The Ronnie Barker Seafood Licence is fishily spelled out by me
to be incompatible with all GNAT GMPL (modified LGPL) licences
with an exception being whenever the licence applies to GNAT.com's
own software. Copyability of copyright statements is a target.
Szmidt's own ideas are one possible way for the GNU community
to counter what seems to be rather minor. It is the faster Ada
strings package on the known earth, just as the strings handling
Gcc compiler is leveraging on garbage collecting ideas as a way
to speed it up. We have yet to see good results from that.

If I had to pick an animal for Mr Szmidt's GNU, I'd volunteer that
idea of mold. It is actually a plant and green, but dropping down
a Kingdom at least gets us into the realm of little real adjustment
when in contact with fier of the FreeBSD mascot. A cat could be
a fiery too: needs no advice when walking over a low plank.








reply via email to

[Prev in Thread] Current Thread [Next in Thread]