Stupid end-user tricks: darcs for /etc and /boot

[Alexander Skwar]

Daniel Carrera wrote:
> Alexander Skwar wrote:
>>> Alex, please calm down.
>> 
>> I'm getting somewhat upset about you, if you must know... Why are
>> you picking on me? Why are you defending this FUD?
> 
> I can see that you are upset. That's why I thought you had taken the 
> Gentoo comment personally.

No, I haven't taken the Gentoo comment personally. I'm getting annoyed
by you defending what he wrote. Maybe it really wasn't "FUD" - how else
would you want to call it? Would "nonsense" be okay for you?

> I defending the OP's comments because 
> I am concerned that using the FUD label too freely leads to very bad 
> consequences

Okay, I agree with that, but why should invalid criticism be defended?
That's another point that annoys me about what you've written in this
thread.

>>>  From the information available, this seems like a true statement.
>> 
>> Well, it's not. It's the user who might be hosing the systgem.
> 
> Well, I think we can agree that this is a place where we have a 
> difference of opinion. I think that fdisk and dd are dangerous commands 
> and you think that they aren't because they must be ran by the user. It 
> seems that by your definition no command can ever be dangerous, which 
> tells me that it's not a reasonable definition.

Wrong. This *is* a very reasonable definition. There are tools which
can be dangerous, if not used correctly or if used by people who
don't know what they are doing.

>>> That etc-update may indeed hose the system.
>> 
>> No, it may not - not easily at least, and that's what he's implying
>> by what he wrote.
> 
> If the comparison with fdisk and dd is appropriate

It 's not really.

> (which, from your 
> emails, seems to be the case) then I would agree in calling it a 
> dangerous program that must be used with caution.

Okay, but as dd and fdisk aren't dangerous, if used by people
who know what they are doing...

>>>>> FUD is an attempt to spread uncertainty over a product, to make 
>>>>> people not use it.
>>>>
>>>> Yep. Just what he did. What he wrote can very well be understood
>>>> that way.
>>>
>>> I don't get that impression.
>> 
>> I do. And it's made even stronger by his new reply.
> 
> I thought that his new reply was informative.

Yes. It was informative in reassuring that there really was no problem
with etc-update, but rather that he maybe didn't quite agree with "the
Gentoo way" - ie. that you've got the freedom to configure everything
yourself and that there are very few "wizards" offered by the distribution.

That's of course very fine - but it has nothing at all to do with etc-update
and blaming etc-update for that is very wrong.

> I read it to say that he 
> is not naive about this program and elaborates on how it can hose his 
> system.

That's not how I read it. I read it so, that he tells us, that in
reality he doesn't have a problem with etc-update.

>> But as you might have seen by now in his new reply, it's not the fault
>> of etc-update, but his very own.
> 
> That's like saying that it's not dd's fault if you it hoses your system. 

Of course its not dd's fault. It would be dd's fault, if it were
to overwrite /dev/hdb when "dd if=/dev/null of=/dev/hda" was
invoked. Or if it were to also change ("fix") the partition table,
if "dd if=/dev/null of=/dev/hda2" were invoked.

> That might be strictly true, but it doesn't make dd not dangerous.

As dd isn't dangerous, I don't see what your getting at.

>> Anyway - how *could* etc-update be made better?
> 
> I think you are missing the point. Whether etc-update can be improved or 
> not was never a point of discussion.

Yes, it was. It really makes no sense to complain that e.g. a tree
has so many leaves. It's just the way it is.

> It is possible for a command to be 
> dangerous and not be actually badly designed.

Agreed. But a command can be made safe to use or dangerous to use.
It depends on how the command is designed - and also on the "target
audience".

> I can't say if this is 
> true for etc-update, but consider another example: you must have a way 
> to partition your disk. You can't get around that. And whichever command 
> is used for this will have the ability to hose your system, it'll be 
> dangerous.

Not really. It depends on how this program is designed. It can be
designed so that it's dangerous if you're not careful or it can
be made so, that the user is "guided" and can't do much wrong.

> It might well be that the OP feels that etc-update is unnecessarily 
> dangerous. If I tried to read minds I might guess that he probably does. 

I'd agree with your "mind reading". But I strongly disagree with
saying that etc-update is dangerous, as it just is not. Also, how
should etc-update be changed?

> But this is not strictly relevant to his assertions.

Yes, it is.

>>> A statement like "Dapper is stable" or "etc-update is dangerous" are 
>>> not gospel truth.
>> 
>> And "etc-update is dangerous" is plain wrong and FUD.
> 
> It seems to me like it isn't. The OP has given examples of how it has 
> caused significant damage to his system,

Once again: Not etc-update has caused damage to his system. He did this
by himself. Saying that etc-update caused damage to the system means,
that etc-update did things which it wasn't supposed to do or which
weren't obvious. That's not true, no matter how often you repeat his
FUD. It was wrong, it is wrong and it will be wrong.

> and your counter argument is 
> simply that only a knowledgeable user should use it,

No, it's not. I'm saying, that there isn't much that can be done
about it. And I'm saying, that it's not the fault of etc-update.

> so you haven't 
> actually refuted the OP's point.

Yes, I have. And he has as well, by his 2nd reply.

> This leads me to think that etc-update 
> really is quite dangerous

Think what you want, but you're wrong. Not etc-update is dangerous, but the
way a user uses his tools can lead to bad consequences.

> and you are redefining dangerous in a way that 
> almost no conceivable software could be called dangerous.

Wrong. Buggy software can very well be dangerous. Software, which is
doing "surprising" things can very well be called dangerous.

>>> There is a degree of opinion, and perceptions will vary based on 
>>> personal experience. If Dapper hoses your system you will think it's 
>>> unstable even if most people in this forum think it's rock solid. That 
>>> wouldn't make your statement FUD.
>> 
>> It might - depending on how such a complaint is formulated.
> 
> Phrasing a complaint in a way you don't like doesn't make it FUD.

No, but a complaint can be phrased "neutrally" or in a "FUD-dy way".

>>> It looks to me that the OP had valid criticism and not FUD.
>> 
>> As you might have seen in his 2nd reply, he didn't have valid criticism
>> reg. etc-update and thus just FUD.
> 
> I've read his second reply. It seems like valid criticism to me.

To me, it doesn't seem like valid criticism about etc-update.

>>>>> unless you're going to accuse him of lying, which is quite an 
>>>>> accusation.
>>>>
>>>> I'm not - how did you get that idea? Is it, because you tend to do
>>>> such things? Well, live with yet another fact: I'm not like you.
>>>
>>> Alex, an ad hominem is not going to improve your position.
>> 
>> But yours?
> 
> I have not attacked you.

You haven't? Well, yet another point where we seem to disagree.

>> If you wish to put it that way. I'd rather say, that the OP complained
>> about the wrong thing ("barked at the wrong tree"?). It's not etc-update,
>> it's himself who got tired and then hosed his system.
> 
> This seems like a reasonable basis for saying that a command is 
> dangerous and should be used with caution.

With such a definition, close to every command is dangerous. I
disagree that "echo" is dangerous, but you agree with that. That's
fine, but we've just got another disagreement.

> And from the OPs point of 
> view it seems like a reasonable basis for changing distribution.

Fine. There are enough distributions out there.

>>> That's the point I'm trying to make. So I am led to think that 
>>> etc-update is indeed a tool that one should be careful with.
>> 
>> Well - not really. It's not that you've got to be extra careful
>> with etc-update. No more than you've got to be careful with
>> eg. "echo" or "perl" or ....
> 
> My understanding is that etc-update is more comparable to fdisk and dd 
> than it is to echo and perl.

But etc-update isn't really comparable to fdsik and dd. etc-update is
basically just a combination of find, diff, mv and rm.

>>> Even if the OP is actually wrong, it seems that calling it FUD is 
>>> really quite extreme.
>> 
>> Maybe it was extreme - what he wrote was somewhat extreme as well.
> 
> Provided that it was (I didn't think it was),

I disagree - it really was. Just read, *how* he phrased his OP and also
his reply.

> two wrongs don't make a right.

Disagreement.

>>> You could say the same statement about fdisk, Windows, viruses and any 
>>> software product.
>> 
>> Yes.
> 
> So, by your definition no software product can be deemed dangerous.

No, that's not my definition.

>> Again (it's getting boring...): It's not etc-update that warrants
>> caution.
> 
> If it can wreck your system if you use it wrong, it warrants caution.

That's too broad and because of this, I disagree.

>> And, basically, I don't see how the working of etc-update
>> could be changed very much.
> 
> dd can't be changed much

Yes, it could. It could warn the user, if "dangerous" things are to be
done or just refuse to do them.

> But in any event... at least we've isolated the point of disagreement. 
> You feel that commands like dd, fdisk and any software product cannot, 
> by definition, be called dangerous,

Wrong, that's not my feeling.

> because they have to be run by the 
> user. I think that a command that, when run, can cause major damage to 
> the system (like fdisk and dd), is dangerous should be used with caution.

"echo" is dangerous?

No, I don't think so.

>>> The statement "as it was wrong, it was FUD" is  fundamentally illogical.
>> 
>> No, it's not.
>> 
>>> There are many things that are wrong and are not FUD.
>> 
>> That's true. It's a question about how something is worded.
> 
> You are contradicting yourself.

Am I?

I agree that not every wrong thing is FUD. It's just a matter of how
a complaint is phrased.

> First you disagreed and then agreed with 
> essentially the same statement.

Where?

>>>>> I it crashed on me, that's solid enough to complain.
>>>>
>>>> No, not necessarily. Maybe your hardware is somehow broken?
>>>
>>> If my hardware is broken that doesn't make my statement FUD.
>> 
>> Yes, it might. If it's the fault of your hardware, it makes no sense
>> to complain that the distribution is unstable. It's the hardware which
>> makes the distribution unstable.
> 
> No, being wrong is not FUD.

True. But a complaint can be phrased so, that it is FUD.

> If I blame Ubuntu for a hardware problem 
> then I would be merely wrong, but not FUDing.

It depends on how you phrase your complaint.

>>> It's important to understand the distinction between "wrong" and "FUD".
>> 
>> It's mainly a matter of how something is written.
> 
> I'd be interested to see how you can say that the sky is green in a way 
> that makes it FUD.

I'm no good in such things, but how about "Don't dare to look in the
sky! It makes you blind because of all the 'green-nes'!".

Alexander Skwar
-- 
Must destroy mankind!  (His watch alarm goes off)  Ooh, lunchtime!
		-- Homer Simpson
		   Homer Goes To College