Matt Randolph schreef:
> Holly Bostick wrote:
>
>> Matt Randolph schreef:
>>
>>
>>> I don't think Knoppix really has an administrator. It really is
>>> an enduser only flavour of Linux. It's sort of a "fire and
>>> forget" distro. Sure, someone had to go to a lot of trouble to
>>> get it set up just right in the first place, but once that was
>>> done it can perform reliably without further administrative
>>> intervention. The enduser not only probably won't set the root
>>> password, the enduser doesn't even need to know that it is unset.
>>> Or even that a root account exists!
>>>
>>
>>
>> Interesting. But again, *someone* had to administer the system to
>> set it up so that a user could be 'pure'.
>>
>>
> It sounds like we are in agreement on this point. We both state that
> someone had (past tense) to administer the system... at some point
> in time. We also both state (or imply) that the enduser doesn't take
> up the role of administrator. Is it possible to have any sort of
> computer that hasn't felt the effects of an administrator? Of course
> not. Any device of any significant complexity can only exist by the
> labors of some knowledgeable persons. I don't think anyone is
> trying to say the opposite.
>
> But does the Knoppix user's system have an administrator NOW? I say
> it does not. It has been configured by an admin... heck, the OS was
> installed to it's filesystem by an admin... but there is no admin
> looking over the shoulder of the Knoppix user.
Right.... so here's a real-world problem, from elsewhere on this list
("authorization failure when sending email")
> Matthew Lee schreef:
>> I've tried every combination of kmail settings available, no joy.
>> I've reemerged all the software that --depclean removed, no joy.
>> I've reemerged kmail, no joy. I've reemerged ssmtp, no joy.
>> However, I think ssmtp, or something associated with it is the
>> problem. But what I haven't a clue. Is there another "simple"
>> mail transfer agent I could try. I don't need anything fancy it's
>> just a laptop connected to the lab DHCP server.
>>
Since this issue seems to revolve around programs also available to
Knoppix (and likely also being used under Knoppix), it's probably a
valid example.
So you've got a user who is unable to use a simple user function (send
email). In the proposed administratorless world, who is supposed to fix
this? The "invisible administrator" (who must exist, but is no longer
necessarily present).
In the case of Knoppix, that's the Knoppix team or the Debian team, if
we're restricting ourselves purely to the packages involved. Is the user
supposed to download and install another "fixed" Knoppix disk in order
to be able to use KMail as they did last week? Or is the user to follow
the Debian protocol and not use the newer version of these programs
(meaning they wouldn't be available to Debian stable in the first place,
which of course, they probably aren't)?
If everything is supposed to "JustWork" and does not, someone must be at
fault. Who? The user is experiencing some unidentified conflict between
programs that worked together well last week. Is there any way for those
who are 'to blame' (development, packaging, some admin along the line)
to work in such a way that these conflicts never ever filter down to the
user? I say no, because we persist in making the conflicting
applications known to the user before all such conflicts are
identified and eliminated-- partly because development requires that
these errors filter down to the user to be identified in the first
place, as developers cannot test under all possible conditions.
Basically the limit of software technology is that we make it
immediately available to everyone as if it does not require
administration, but it is (almost) never so stable and intuitive that
this is in fact the case.
The solution would seem to be to either not make the software available
until it has been sufficiently tested so that it does "JustWork" under
all possible
conditions (which the trained greed of users will not allow), or teach
the user
that sometimes they may have to do something a bit more complicated than
just click 'Send' (which means that the user cannot be a pure user anymore).
I don't see any middle ground here, but maybe I'm missing something.
Holly
>
>>
>>
>>> I don't believe this sort of user experience is limited to
>>> read-only systems like Knoppix, though. Look at
>>> Lindows/Linspire. How about those $200 Linux computers they are
>>> (or were) selling at Wal*Mart (strewth!). I expect those
>>> machines ARE intended to provide the enduser with an essentially
>>> administratorless (to coin a word) experience. Linspire (at
>>> least used to) have the user running everything as root. But do
>>> you think the enduser always knows that? I think the enduser
>>> simply knows that when they pay to install OpenOffice.org from
>>> Linspire's private apt servers, it just works; it installs
>>> without their ever having to `su` or `sudo` or anything. That
>>> Linspire user essentially is the admin, though she doesn't know
>>> it and she almost certainly doesn't behave like one.
>>>
>>
>>
>> And many now question whether Linspire can even be called a Linux
>> distribution for this and other reasons, despite the fact that it
>> runs on a Linux kernel. We're all wondering if that is then the
>> only requirement, or does it also need to follow 'the rules' to be
>> counted?
>>
>> But that's a whole 'nother discussion.
>>
>>
> I didn't mean to imply that Linspire is a proper Linux distribution.
> It certainly doesn't follow 'the rules' of a proper operating
> system. But neither does Windows for that matter (and for much the
> same reasons). Knoppix doesn't follow the traditional 'rules' in that
> it is read-only. Embedded versions of Linux don't follow 'the rules'
> in a sense because the user might never interface with the OS at all,
> merely a single application instead. Linspire IS trying to follow a
> set of rules. Specifically, the ones Windows goes by. So doesn't
> that mean that Linspire is at least as valid an OS as Windows is?
>
> No, Linspire is not proper Linux, but it is bringing the kernel and
> Linux apps into some peoples homes. It may not be bringing the
> traditions, the behaviors, or the ways of thinking that are a part of
> Linux, but those may come with time to those that seek them. But
> even if they never did, why should certain sorts of people be
> prevented from using Linux just because they aren't clever enough or
> are too busy to do it properly? Some people will never learn more
> than the basics of operating a computer. If those people are forced
> to chose between learning to use a proper OS properly versus using a
> typewriter, they'll start dusting off the old Selectric.
>
> I have heard rumors that some futurists are predicting the death of
> the PC in the not too distant future. Instead of PCs they predict
> people will use weird multi-function mobile phone devices with speech
> recognition interfaces. Will you want to have to log in to your
> mobile in order to answer it? Will you want to have to create a cron
> job to get it to download your email? But don't you want it to be
> Linux-based anyway?
>
>>> What I think I hear you saying is that being able to get away
>>> with this foolish behavior should not be one of our goals. I did
>>> not mean to imply that careless hardware shopping should be
>>> encouraged. Rather, I used this as an example to try to
>>> illustrate how lacking driver support slows the growth of Linux.
>>> If Linux is going to grow it's user base significantly, it's
>>> probably going to have to attract quite a few of those careless
>>> boobs too. And if Linux can't be made to work on their hardware,
>>> do you think they are going to run out and buy a new computer or
>>> will they simply rethink the decision to try Linux?
>>>
>>> Although careless hardware shopping should not be encouraged,
>>> being able to get away with it (that is, having nearly ubiquitous
>>> hardware support) should indeed be one of our goals.
>>>
>>
>>
>> OK, I understand that, but... how exactly is allowing one to 'get
>> away' with such behaviour not 'encouraging' such behaviour?
>>
>> If one has always been able to 'get away' with any behaviour, why
>> would one think that any other behaviour is possible?
>>
>>
> In my town it is the custom to drive ten to fifteen miles per hour
> over the speed limit. Almost everybody does it. You're much more
> likely to see someone greatly exceeding the limit than you are to
> find someone strictly obeying it. By allowing people to get away
> with this scofflaw behavior, do you mean to imply that the police are
> actively encouraging it some how? Do you honestly believe a single
> one of those motorists doesn't know they are breaking the law? Or
> that it is even possible to drive more slowly? So which is it? When
> they put out limit signs and mentions of fines, but then don't pull
> you over for every little violation, are they encouraging or
> discouraging speeding?
>
> I think I see what you're driving at, though. What I hear you saying
> is that whenever someone gets away with a thing, they will
> invariably feel encouraged to do it again. I feel, however, that
> this "encouragement" is distinctly different from encouragement with
> the intent to encourage. The latter is the result of an action while
> the former is the result of an inaction. Encouraging is not the same
> as not discouraging. Some people are encouraged to speed by the lax
> enforcement, while others are discouraged from speeding by the signs
> and a desire to obey the law. Whether one is encouraged to be a
> lawbreaker or not varies from one person to the next. There is
> nothing invariable about it.
>
> But we're not really talking about cars here. We're talking about
> buying hardware and the availability of drivers for a particular
> operating system. Do you mean to imply that carefree hardware buying
> is such an evil that we must ensure that the behavior never goes
> unpunished? To me that would mean that hardware manufacturers would
> have to be DIScouraged from releasing Linux drivers so that every
> hardware purchase by the uninformed would be a total crap shoot.
> Allowing people to get away with careless shopping is an unavoidable
> consequence of having universal hardware support. If we labor to
> obtain the latter, we will be enabling the former poor behavior by
> default. But if Linux actually had 100% complete universal hardware
> support, failing to look for an OS compatibility sticker would no
> longer BE poor behavior. It would have become unnecessary entirely.
>
>> Ubiquitous hardware support, on the one hand, is closer than you
>> think (there's not all that much hardware that cannot, no matter
>> what you do, be made to work under Linux; it's just not that it all
>> "JustWorks"), and on the other hand is less relevant than you
>> think (I have drivers that enable my ATI card to 'work' under
>> Linux, but they suck, so whose fault is that? Not Linux's. Nor is
>> it Linux's fault if I plug in my digicam and it is mounted, but I
>> don't know how to get the dv output into Kino, or can't figure out
>> how to properly mount my perfectly-well-detected Flash card to get
>> my pictures into whatever graphics display or editing program I
>> might use). The hardware works fine. But that's no help if I can't
>> understand how to use it, or can't use it effectively.
>>
>>
> You're right. Hardware support has come a long way. Knoppix, for
> instance, is awesome. More often than not, it really does "just
> work." But I do not believe that having even basic hardware support
> is as irrelevant as you seem to suggest. Are you not better off
> having crappy support from ATI than you would be if you had none?
>
> But I'm not wishing for poorly written or incomplete drivers. When I
> say that a piece of hardware has Linux support, I mean that it
> "runs" not "crawls" (to quote Mr. Knecht, I believe).
>
> If your digicam is mounted but you cannot get the contents into Kino,
> are you really getting support from the manufacturer? Shouldn't
> support for Linux include instructions on how to get it to work with
> Linux? If some insanely talented hacker/engineer reverse engineers
> your camera's interface and adds it to the kernel, yet no Howto
> exists to show you how to use it, do you really have "support?" I
> was talking about driver support but I guess I really meant to be
> talking about complete support: driver support, software support,
> technical support, the whole package the Windows folks get. If we
> had that from hardware manufacturers, Linux's growth would snowball.
>
>> And enabling some kind of efficient communication between the
>> hardware that is being properly detected by the kernel, and the
>> programs the user uses to utilize the device is a design issue,
>> which is an administrative task. If Wine/Cedega will run Morrowind
>> using my ATI card under certain configurations, but not others (or
>> the 'default' config), then someone has to be responsible for
>> setting that up so that the user (who is also me, of course) can
>> just click an icon and run Morrowind. Hell, someone has to make
>> sure that the ATI drivers are installed in the first place-- and
>> supposedly the user is never supposed to know about any of this,
>> and there should never be an admin, so who's supposed to do it
>> then? The Tooth Fairy?
>>
>>
> As I understand it, in the case of Cedega, the someones responsible
> for setting things up for a particular game are the techs at
> Transgaming. In Wine, that someone is you, though hopefully someone
> else has posted how they did it so you won't have to reinvent the
> wheel. But Windows gamers often have to deal with administrative
> tasks too (and other Windows endusers do as well). It happens quite
> often that a Windows game has to be patched in order to work on a
> particular computer. The enduser may have to seek help from the
> developers and follow their instructions. Does this make the enduser
> the administrator? Or is the administrator the one that solved the
> problem, made the patch, and wrote the instructions?
>
> I'm not saying "there should never be an admin" ever. But for
> certain sorts of users, there need to be products that don't require
> that an admin be present in order to keep things working properly
> (enough). If a box is configured well and it can be made to be
> static, does it really need an administrator? What if that box is a
> refrigerator, a video game console, or even a non-networked PC? I
> think that a Linux appliance like one of these can and should be able
> to be used without the further help of an admin.
>
>> The fact that you may be able to "Plug and Play" does not remove
>> the necessity that administration must occur: under Windows, a
>> Wizard does it, in an enterprise situation, IT does it, under SuSE,
>> maybe YaST does it, under Gentoo, you do it (or Mark does it for
>> you :) ).
>>
>> But the fact that at some point somebody has to be responsible for
>> administration is inescapable, and I feel that saying that's wrong
>> somehow is... wrong.
>>
>>
> "[S]omeone had to go to a lot of trouble to get it set up just right
> in the first place, but once that was done it can perform reliably
> without further administrative intervention," is what I said about
> Knoppix. There was an admin, but there isn't one now. I don't know
> what I said that led you to believe I thought something different.
>
> The more that is asked of a system, the more administering must be
> done to it. A video game console with a Linux OS only has to do one
> thing. As a result, it will require essentially no administrative
> intervention. An enterprise web server has to do more things than I
> can count. As a result, it has to be closely watched and fiddled
> with to keep things running smoothly. Somewhere in between is the
> Linspire desktop. If all it has to do is write documents, send
> emails, and surf the web, then little more than security updates
> would need to be performed (by cron, even) to keep it going. But if
> you actually want to use your computer AS a computer instead of an
> appliance, then a different distro would be a better choice and
> somebody is going to need to administer things.
>
>> Because it's a limit of technology, and pretending that such limits
>> don't exist (or worse yet, attempting to conceal such limits)
>> seems very very unwise to me.
>>
>>
> I agree that the limitations of a technology should be made aware to
> its users and it should be done in such a way that they will
> comprehend those limitations. But does that mean that trying to make
> simple systems that don't require constant babysitting by flesh and
> blood administrators is automatically a bad idea simply because it
> shields the enduser from the underlying mechanics?
>
>>> I was not aware that any company was trying to encourage careless
>>> hardware shopping. If knew it to be so, I'd be as unhappy about
>>> it as you appear to be.
>>>
>>
>>
>> One word.... Winmodem (easiest possible example).
>>
>> All winmodems are (naturally) marked that they work under Windows.
>> How many of them are marked that they *only* work under Windows,
>>
> All of them. The list of hardware and software requirements on each
> package *only* indicates that it works under Windows. Granted, they
> don't say "won't work with Linux," but they don't say "won't work
> with a Cray," either. If you buy a piece of hardware and the
> manufacturer didn't say it would work with your OS, and you can't get
> it to work with your OS... then you're on your own.
>
>> because a Winmodem is an incomplete piece of hardware, where the
>> functioning of certain physical chips (which are physically no
>> longer present) are replaced by software functions available only
>> in the Windows Operating System (because the Windows Operating
>> System was specifically designed with closed-source APIs to replace
>> the functions of specific chips formerly on the modem PCB)?
>>
>> How many 'real' hardware modems (which have all the chips, and do
>> not replace any hardware functionality with OS-based functions) are
>> distinguished on their packaging from WinModems, or vice versa?
>>
>>
> None that I've seen, but that doesn't mean there aren't any. I'm not
> sure, but I think a proper external modem can be made to work with
> both PC's and Macs. If so, it might even say so on the box. I
> believe I was able to tell that my last (last as in final) modem was
> a proper one because the box said it worked in DOS too.
>
>> And do you think that the 1) creation of, and 2) lack of disclosure
>> on the packaging of, such crippled hardware was somehow not
>> 'encouraged' by the company whose product's market share benefits
>> the most from the existance of such hardware (because the hardware
>> seems to JustWork with their software)? The benefit to the
>> hardware companies, of course, is that their product becomes
>> cheaper to produce, since it requires less chips... and there's
>> little chance that the old PCB with all the chips will need to make
>> a reappearance, because the software being used to replace the
>> hardware functioning is eternal (not least because of the
>> manufacturer's new hardware design).
>>
>>
> Do you really believe that some little Taiwanese company failing to
> state "this product will not work without Windows," on a modem
> package is evidence that Microsoft is up to no good? I'm not saying
> Microsoft isn't. Of course they are. But I don't think they told
> hardware manufacturers how to word the compatibility information on
> each box in an effort to mislead those people who might want to
> switch from Windows to Linux some day (and do it at 53Kbps to boot).
> Even if Winmodems DID come with a warning about being unusable
> without Windows, do you really think that would have affected sales
> appreciably? And why aren't there Win-NICs, or Win-mice, or Win-hard
> drives? Granted, they may come with the arrival of the Microsoft
> brand of Digital Rights Management. That may even be the real reason
> they're moving towards DRM at all. But I think the fact that the
> RIAA and the MPAA have been screaming their heads off might have
> something to do with it too. ;-)
>
> - Matt
--
gentoo-user@gentoo.org mailing list
