Linux-Advocacy Digest #269, Volume #29 Fri, 22 Sep 00 20:13:03 EDT
Contents:
Re: filename extensions are NOT a kludge (FM)
Re: Because programmers hate users (Re: Why are Linux UIs so crappy?) (FM)
Re: Implications ([EMAIL PROTECTED])
Re: How low can they go...?
Re: so what?
Re: Because programmers hate users (Re: Why are Linux UIs so crappy?) (Richard)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (FM)
Subject: Re: filename extensions are NOT a kludge
Date: 22 Sep 2000 22:51:43 GMT
Reply-To: [EMAIL PROTECTED]
Donovan Rebbechi <[EMAIL PROTECTED]> wrote:
>The only time that this becomes a danger ( in C++ ) is when you want to use C
>from C++. And it's a problem if you want to use C from any other language.
>Any proficient programmer knows how a language is "meant" to be used, that
>is, they have a sense of what is considered "good style" in that language.
>For example, "good style" in perl would mean preferring foreach to
>for (...). In C++, it means using OO all the time.
No it isn't. C++ is quite explicitly a hybrid language
supporting multiple paradigms. There are often very good
reasons why you'd want to use C++ but use its OO features
sparingly.
>> Additionally, people
>>certainly go through a separate design phase, it just happens all
>>in my brain.
>You'll have a hard time designing anything moderately complex with that
>approach. Even a head as big as yours can only hold so much (-;
Actually it works quite well for a simple project. For more
complex projects requiring multiple programmers, writing
down becomes more important.
Dan.
------------------------------
From: [EMAIL PROTECTED] (FM)
Subject: Re: Because programmers hate users (Re: Why are Linux UIs so crappy?)
Date: 22 Sep 2000 22:44:43 GMT
Reply-To: [EMAIL PROTECTED]
Richard <[EMAIL PROTECTED]> wrote:
>> >Depending on the dialect, it's either free (Squeak) or there is support
>> >for splicing the VM into your app (VisualWorks).
>> Which makes your app bigger, right ?
>That's inevitable when you want to run apps written in one language on
>an OS that uses a completely different language.
No it isn't. It's more of the peculiarity of the Smalltalk
run-time environment.
>If you used an OS built
>on Smalltalk, you'd be paying this penalty only for C/C++ apps.
No, either the OS is extremely poorly designed that all apps
not written in Smalltalk will pay this penalty or you won't
at all. It's hard to imagine an OS not being to support
native compilation of C/C++ programs.
>I honestly don't want to get into an argument about language typing
>since I don't understand how C++ does it exactly, let alone the pros
>and cons of it.
In other words, you'd like to bash C++ and promote Smalltalk
but you don't want to get into specifics because you know
nothing about C++.
Dan.
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup,comp.software.config-mgmt
Subject: Re: Implications
Date: Fri, 22 Sep 2000 22:53:07 GMT
In article <%0Iy5.12646$[EMAIL PROTECTED]>,
"paul snow" <[EMAIL PROTECTED]> wrote:
[
> Adding software to my software library is a literal form of computer
sex.
Ah, that would explain that thoroughly fscked feeling I get when
clicking through a EULA. And people think I'm joking when I tell them
our network is a clusterfsck. It really is by your definition.
Gotta go, I'm busy installing Norton's new Condoms for Windows. Can't
wait for the lubricated version to get out of beta....
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED] ()
Crossposted-To:
comp.lang.java.advocacy,comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: How low can they go...?
Date: Fri, 22 Sep 2000 23:17:30 -0000
On Fri, 22 Sep 2000 22:43:54 GMT, The Ghost In The Machine
<[EMAIL PROTECTED]> wrote:
>In comp.os.linux.advocacy, [EMAIL PROTECTED]
><[EMAIL PROTECTED]>
> wrote
>on Fri, 22 Sep 2000 20:34:48 -0000
><[EMAIL PROTECTED]>:
>>On Fri, 22 Sep 2000 18:50:51 GMT, The Ghost In The Machine
><[EMAIL PROTECTED]> wrote:
>>>In comp.os.linux.advocacy, T. Max Devlin
>>><[EMAIL PROTECTED]>
>>> wrote
>>>on Sun, 17 Sep 2000 01:52:44 -0400
>>><[EMAIL PROTECTED]>:
>>>>Said The Ghost In The Machine in comp.os.linux.advocacy;
[deletia]
>>>
>>>And now we're more or less stuck with it. :-/
>>
>> This is more a side effect of bad MS engineering than
>> anything else. Nearly every other commonly known OS
>> has not suffered this problem (VMS, Unix, NextStep, MacOS).
>
>There's two other issues, unfortunately. Not only did MS engineer
>it badly, but MS tried to (and, for the most part, succeeded to)
>leverage its dominance in one operating system -- MS-DOS -- to
>give us all an even worse-engineered graphical subproduct, namely,
>Windows 3.1, and later on Windows 95, Windows NT, and you know
>the rest. :-)
>
>But the big problem? We bought it hook, line and sinker. (We may
>have felt that there was little other choice, but we did buy it.
>Or perhaps middle management in key corporate accounts bought it.
>I don't know.)
What's this "we" stuff? ppppffffttt!
>
>If that is indeed the case, that speaks ill for the American
>business consumer, and the home user as well; do we really prefeer
>surface shiny glitz to substance? I'm not sure I like the answer.
I'm not sure I would characterize it quite like this. It would
be more accurate to call it a herd mentality. Up until recently,
it was all the other potential Microsoft rivals that had all
of the glitz and shininess.
[deletia]
Also while the herd mentality is certainly there, I think the
nature of software interfaces and how they tend to interfere
with free choice is far more critical. It's not enough to merely
have the "biggest fraternity", you also need a way to trap people
in once they've made a bad initial decision.
--
Neglect of duty does not cease, by repetition, to be neglect of duty.
-- Napoleon
What's another word for "thesaurus"?
-- Steven Wright
------------------------------
From: [EMAIL PROTECTED] ()
Subject: Re: so what?
Date: Fri, 22 Sep 2000 23:21:22 -0000
On Fri, 22 Sep 2000 15:37:22 -0500, Philo <[EMAIL PROTECTED]> wrote:
>well folks...
>i seem to have seen a few hostile posts here over the last few days...
>now it is a good thing to have discussions etc. as to who knows the most...
>which operating system is best and so on...
>but don't you think we should be talking about more important things.?.
>for example...i dare anyone to post an example of anything more nerdy and
>useless than what i have done recently.
>
>booted to linux command prompt...
>then opened dosemu ...
>then from within the dos emulation mode i fired up my cp/m
>emulator...
>it worked but now what do i do?
>so rather than have heated arguements here...has anyone ever done something
>useful like me?
You mean like video capture, spreadsheets, creating word compatible
documents, games, ripping mp3's and creating a home audio server,
scanning and web publishing of photos, downloading and playing media
files of various kinds (image, music, video), balancing the books,
or just surfing the net?...
>
>
>Philo
>
>
>ps : have a sense of humor ...ok?
...not too different from what I used to do with my Atari ST.
--
In the Spring, I have counted 136 different kinds of weather inside of
24 hours.
-- Mark Twain, on New England weather
A new koan:
If you have some ice cream, I will give it to you.
If you have no ice cream, I will take it away from you.
It is an ice cream koan.
------------------------------
From: Richard <[EMAIL PROTECTED]>
Subject: Re: Because programmers hate users (Re: Why are Linux UIs so crappy?)
Date: Sat, 23 Sep 2000 00:01:08 GMT
2:1 wrote:
> > Define "incompetent" and please explain how the programmer
> > possesses this mythical quality.
>
> I assumed thet I was using your meaning of the word 3 posts back. Besides,
> the programmer is able to create a system, so he is able to simplify it
> etc...
And I put it in "quotes" precisely because I don't think it means anything.
> > has to do with *high-level architectural decisions* which "hackers"
> > never make consciously and explicitly and thus always do badly.
>
> I hack software. I also design things.
I eat. I also cook. And no matter how much I eat, it won't make me
a great chef.
> > "I like it" isn't a rational argument for anything. You've just
> implicitly
> > asked "Why should programmers be rational?"
>
> No I haven't. If a programmer likes a program, then why should he change it.
> Like/dislike is not rational, but you can never completely rationalise
> anything that is human. It is a perfectly good reason.
Not for engineering! Would you *want* to use a bridge whose only
argument going for it is because the engineer "likes it"?
> What about different tastes in art and music. Some people can stans Wagnerian
> operas, other people love them. Those do not coincide in any way.
A star looks completely different from a dog, but both are made up
of objects that follow the same laws of physics. Similarly, objects
perceived by humans as beautiful need not be perceived as the same
in order to follow the same laws of beauty and order.
> You're confusing tribes. You're also talking about only one aspect of beauty.
> Sticking to that aspect, (put bluntly) why do some men go for thin girls and
> others go for fat girls, if there is a hardwired reason to go for one type?
Actually, people don't go for fat girls. Some people are more willing
to *tolerate* obesity or maybe even have a fetish for it, but that's
not the same thing as being attracted to it.
> There is also no hatred of users among most programmers. I really have no
> idea how you got this in to your head. Find some OSS programmers and ask them
> if they truly, deeply dispise users.
Like they're going to admit this! And not having admitted one's
feelings to oneself is not the same as not having those feelings.
> I dont agree with this. Someone who is not an academic is not likely to want
> to invest the time learning the ins and outs of a system. And why should
> academics like unnecessary and unnatural contortions?
Because they already have a vested interest in them. If they don't
inflict unnecessary contortions on the new generation then that
proves they were morons to accept these same contortions from their
own teachers.
I know plenty of non-physicists who would like to learn physics
but they don't want to *do* physics and the moment they admit
this to a physicist, they're screwed.
> You don't know many academics then. I know plenty (I'm a student). Most of
> them are nice, agreeable people. Of course, there are ill-adjusted bigots,
> but then, there are plenty of non academic ones, too.
I know plenty of academics also. And quite a few are nice, agreeable people
who happen to hold that their field of expertise is all that matters and
that other people are doing things which may be necessary but just aren't
worthwhile ultimately. Take this attitude to an extreme and you get contempt.
> > Is there a reason why laypeople don't understand what a
> > soliton is instead of being presented with the oxymoronic
> > (and plain moronic) "duality of light"??
>
> That sentance makes no sense. Besides, what have solitons (in *general*) got
> to do with light?
In general, all fundamental particles *are* solitons.
There are three possible ways to structure the teaching of physics;
1) historically
2) mathematically, and
3) pedagogically (ie, according to the concepts themselves)
Now take a guess as to which one is NOT used.
> > You only think there is because society indoctrinated you
> > into being unquestioning of academics,
>
> Please withdraw that. I find it incredibly insulting. I have always
> questioned and discouvered, since a very young age.
Which is independent of indoctrination and propaganda. Academics
for example are some of the most propagandized, and most vulnerable
to it, people on the planet.
Most universities offer courses about Propaganda. It isn't the
domain of the stupid and credulous.
> Great teachers (1) do not assume that they are unquestionably right and (2)
> treat schoolchildren as people. I don't think teachers should assume that
> they are wrong, but assuming you are wrong and not assuming you are right are
> different.
And even the greatest teachers work in a system that encourages to act
against their instincts. The mass schooling system's sole purpose is
to indoctrinate blind obedience to authority through the use of ritual
humiliation (this is the only conclusion possible after you understand
that all of the harmful consequences of bad schooling which never get
changed are a feature and not a bug). It's very effective.
> > > Programmers are not responsible for the users mental well being at
> all.
> > > In fact, the OSS programmmers (esp. GPL ones) are not responsible
> for
> > > anything. No fitness for a particular purpose guarntees. Since the
> >
> > Blatantly incorrect, illegal, unethical and immoral.
>
> There are no fitness for a particyular purpose guarntees. Read the license,
> esp. GPL The programmers are making avaliable something that may be of
> utility if the user wishes to try it. There is nothing unethical illegal or
> immoral about that. Only malicious software is immoral, illegal and
> unethical.
Your last sentence is incorrect. And your first sentence is not the same
as your previous claim. There may not be any /particular/ guarantees but
there most certainly are some general guarantees.
> OSS is not a gift. Read above about that. You should use it with the caution
> that you would use something that you took out of a skip (with permission, of
> course). You get some great things out of skips.
What's a skip?
> If you were making HCL(aq) and someone asked you if they could have some, and
> then thay drank it, it's their fault if you gave it to them with a license
> saying that it might be dangerous--use at your own risk.
It depends enormously. Is the fellow a chemist or a first-year student
in his first chemistry lab? Or better yet, have you ever even seen him
around the chemistry department before? These things matter in lawsuits
and moral philosophy.
> I wouldn't complain about that, if I made no enquirey about the interface.
> I'd just dump them in the bin and have done with it.
I would complain proportionately to the amount of time and effort I had
invested which turned out to be completely wasted (which, if I had to
arrange for a car to pick them up would be a *great* deal of trouble).
If someone offered me hard drives, I would expect them to be IDE or SCSI.
> > Only by eliminating programs wherever possible.
>
> So how are these processes instructed to do what they have to do. You have to
> have a program at some point.
A process is the entity that results from converting a program into ...
something else. If you save and reuse the process every time, then you
only need the program the first time. If you don't discard processes,
you don't need to recreate them.
> f(x) = x^2 + x*4
>
> That function is not an operator because it's not operation on anything.
>
> y=f(10)
>
> Aaah, now it's an operator because it is operating on something.
>
> A program is an operator. A processor is an object used by a computer to make
> a program work.
A reasonable argument. Unfortunately, it's symmetric and you can use that
argument to say that "10" is the operator on which f(x) works. Your argument
fails because f(10) is not the proper analogue to a process. Instead, a
process is +10, a program is 10 and the compilation process that results
in +10 is the +.
And while this discussion is interesting, I don't know enough about the
related theory to continue discussing it profitably. This only came up
because someone wanted to argue that executable files and non-executable
files are completely different things, which is wrong.
> > I hate commercial software because it isn't free. But I don't think
> that
> > Linux is free in any meaningful way either.
>
> Free as in speech or beer?
I don't dispute that it's cheap and even gratis, but those are not
meaningful senses of the word. I don't think that Linux is libre.
If the only way to understand Linux is to spend a year reverse-
engineering all the architectural design decisions, then it's not
free in any meaningful sense.
> As for linux, you can do with it what you will
> except make it less free. Seems quite free to me (but people argue till the
> cows come home aboout how `free' the GPL is, but I won't do it here).
The GPL is necessary but *not* sufficient to ensure liberty.
The GPL defines "source" as "the preferred format for human
manipulation" (going from memory). Well, C/C++ is not source;
Smalltalk code compiles *down* to C code.
I'd define source code as "the highest language possible to
achieve the necessary task".
> If you don't care for compatibility then ytou an evil programer who doesn't
> give a monkeys for users mental health, because they can't run their programs
> on your system.
It's not *my* fault if no one wants to port crud to my platform. <grin>
> > OS design principles I've seen are consistent with each other (and
> that's
> > quite a long list).
>
> Compatibility is the big one that doesn't fit. And it's very important.
Between compatibility and "everything else", I'll pick everything else.
> You seem to insist that all programmers are evil failures, bitter and twisted
> through years of hating users. I think you're wrong. really good design and
> compatibility with a less good design are exclusive.
Is that how you justify anti-human user interfaces? Even when programmers
don't *have* to follow what came before, they still do it!
I don't deny that compatibility can be a reason for bad design, but all
too often it's just an excuse. And you can tell which is which by what
the programmer does when the excuse is removed.
> > It is. It's also a failure. Of the goals it even attempted to reach,
> it
> > failed half of them. It was working from Unix semantics after all.
>
> You speak like an expert on using plan 9. are you?
Hardly. I just went through the design documents and interface specs.
> The file system looks like unix. How do you know that unions are a kludge.
> Have you looked at the code?
Union directories are a kludge regardless of how nicely they're
implemented. The most eloquent lecture is still a failure if the
students fail to learn anything.
Union directories are a kludge because their function can be
performed through a simplification of the filesystem design;
relax the constraint that directories may only have one parent
thus turning the FS from a tree into a lattice (an infinitely
more powerful data structure).
> > special devices, or the superuser concept, or even apply the FS
> semantics
>
> What's wrong with devices as files? And, yes, they did get rig of the root
> concept (I read some docs that said so.)
They just replaced it with another concept that's equivalent
in bogosity to root. Now I remember: they invented "null" users!
What a load of crap!
There's nothing wrong with devices as files but there is something
wrong with devices in the /filesystem/. It's perfectly sensible for
two different components (say, the framebuffer and TCP) to export
objects that mean different things. What's wrong is for one component
to export objects that "really mean" something that doesn't even
exist in that component at all. And it is definitely wrong for one
special component to depend on kernel support to pull off this kind
of crap.
> Could you give a reference that says thay are definitely wrong?
I'm going from memory on all this; but they admitted that disconnected
namespaces made effective debugging impossible and made replication and
migration of process' environments to different machines completely
impossible; they were going to invent a new syscall to handle *just*
that. I don't know where you could find this in the docs, if it's very
important to you then I can check it out later.
In any case, disconnection is wrong from purely theoretical considerations.
1) it isn't necessary in order to have namespaces /local/ to each process
(what they actually wanted from the start), and 2) *connectedness* is one
of the fundamental principles of OS design. Plan 9 had only half-assed
connectedness since the filesystems were connected by mounts only from the
outside-in and never the inside-out.
> How are mounts not bidirectional?
If I mount yourMachine under /myMachine, then I can do 'list' and see
/myMachine/yourMachine, but when you do 'list' you can't see anything
different from your end.
> It depends on your definition of bidirectional. In one sense they are. You go
> up in one direction, you come back in the same direction, so , uhh yes, it
> does.
No, that's just the shell keeping track of where you've been to compensate
for the limitations of unidirectional links. You never actually "go back",
you just "forget" that you ever went forward and you suddenly find yourself
back.
> OK, so create 2 symlinks. Easy enough. So you can have unidirectional *and*
> bidirectional links.
Symlinks are a kludge in the first place, I don't plan to ever implement them.
And having two links is not the same thing as having a single bidirectional
link. If you remove a bidirectional link, both sides are gone. And since
there is no legitimate reason for users being able to create unidirectional
links ....
Additionally, how are you going to *find* the paired symlink? If you only
have one bidirectional link then there's a function that provides the other
side's name.
> > The only bidirectional links in Unix are directory links. The backend
> > of the link is uniformly named ".." so it's not very useful, is
> > inconsistent
>
> It's cnsistent in one way, much like saying `up one floor' is a consistent
> way to refer to the floor above whatever the number of the floor that you are
> on.
Now imagine that the *ONLY* way you could go up one floor was by saying
"up one floor" while when you go down you could freely say "go down to
floor X".
Calling all links that go up ".." is uniform, but very, VERY stupid.
Why can't a user rename ".." to "domain of he who thinks of himself
as a god"?
> The links are the same. One end is an entry in the directories information
> structure. the other end points to a file inode. All hard links work in the
> same way.
Just that sometimes you have two of them and sometimes you only have one
of them. Like I said, the fact that you have backends to directories is
inconsistent with the fact that you can't have backends to files.
> > Authentication of users is not a sensible system-wide security scheme.
> > Imagine having to provide your name and password every time you
> > want to access a process in some way. And conceptually, it just
> devolves
> > into Access Control Lists, which are inherently broken.
>
> What is wrong with ACLs? You are makeing an un backed up statement.
There is a lot of research on why ACLs suck. There are many,
MANY things that you can't do with ACLs that can be performed
trivially with capabilities (things like you want to give your
access to an object to process A but it needs to give its access
to that same object to process B in order to do its job), and
I don't know of any examples of the reverse. Additionally, user
names are not compatible across machines on the network. And in
Unix, the creation of a user is an act of divine intervention.
>From a purely theoretical perspective, "user" is a superfluous
concept that does not belong at the level of the filesystem. It
belongs at the level of the login process *only*.
In fact, eliminating the low-lewel concept of "user" is necessary
to be able to do stuff like creating subusers that are subordinate
to two or more users. VSTa has a tree of users but what you want
is a lattice.
> A general graph is probably the best abstraction.
Bingo. Though the difference between a graph and a lattice
(the ability to create cycles) is questionable.
> Unix provides a 3 level tree. It's not _totally_ flat, but t could be better.
>
> ________everyone___________
> | | |
> _____group1 group2 group3
> | | |
> user1 user2 user3
But every change to that structure, no matter how minute or insignificant,
requires an act of the system's god.
> > "anyone" /includes/ you.
>
> Er, yes. I find it easy to use. I don't want it changed.
That's where the "just haven't been exposed to the alternatives"
comes in. :-)
> Frogs are the most monmentally stupid ceratures to have graced the surface of
> this planet. they are worse than sheep. I don't see the relavence, though.
It's not possible to make a good judgement if you lack the perspective
and the knowledge required. Even after you check out VSTa and Plan 9, you
won't know the many things that have caused me to judge Unix intolerable.
> > particular piece of software doesn't have to be usable by everyone.
>
> Then why are you complaining that users find SW hard to use?
Because software doesn't have to be usable by everyone but it does
have to be usable by those people who have to use it.
> > I think that computing is the perfect field. Mathematics is a millenia
> > old field so it would be extremely difficult for me to contribute
> there.
>
> Computing is a branch of mathematics.
Maybe for the people who deal with Turing machines and P vs NP problems.
> > Are you talking about standard english or C/C++?
>
> I was talking about which ever lang you were refering to. I don't know which
> , but I still believe what I said.
Then you're just /assuming/ that the alternatives are inefficient ....
> C is very widely used, soit's going to have the most invested in it. Besides,
> some programs I write carefully don't optimize much.
>
> Also, do you mean efficiency of speed, ease or the best oveall compromise?
Since the purpose of programming languages is to express ideas to fellow
programmers, an efficient language is one that can be used to learn the
encoded ideas with the least amount of time and effort. C/C++ is a complete
failure for any large system since frequently it's impossible to reverse-
engineer software written in it.
> > > There are plenty of people who do like them and do understand them.
>
> > There are plenty of people who are into pain and submission as a
> sexual
> > thrill.
>
> It looks like a poor attempt at throwing insults.
Ohhh, no; you just haven't seen the dozen other things that came to mind!
> Eh? No what I'm justifying is a camera that some people find unuseable and
> some people really like. What is wrong with that?
The fact that nobody actually finds it usable?
> The purpose of english is to allow people who speak it to communicate.
The same can be said of *every* other language in existence, so you
haven't actually said anything.
> It is
> coincidence that it is used as the business language.
So Japanese people learning English is unrelated to the fact that English
is an international language of commerce?
> I expect there are as many broad minded programmers as normal people. If you
> don'tt like it, then you should become a hermit.
Normal people don't make for very broad minded programmers. And thank
you for the suggestion but I'd already considered becoming a hermit.
> I doubt there is an OS that is 1000 times as good as unix, if that could be
> judged at all.
It could be judged with an analogue of a function point metric.
> What do you mean by all that. WTF is wrong with enlightenment?
Sigh. All I'm saying is that my attitude towards Unix is exactly
the same as most Linuxers' attitudes towards Windows. And if
Linuxers have something personal against the dominance of Windows
then why can't I towards the dominance of Unix?
> > The same applies for *every* other OS project out there.
>
> If your OS is interesting then people will lend a hand with it. But having
> flame wars on an OS ng will not help the cause.
NOTHING will help the cause. That's why I get pissed off and start
arguing on a newsgroup!
Plan 9 is interesting. So is VSTa, EROS and more than a dozen other
OSes that are vastly superior to Unix. But they'll *never* get the
programmers to make it a fraction the size of Unix because the only
thing that supposedly matters is Compatibility.
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.advocacy) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Advocacy Digest
******************************
