Vote Up
6
Votes
Vote Down

Before we start dissecting this interesting announcement, let us revisit one particular tidbit from the official Linux kernel cheerleading squad:

“The Linux driver model is different. For users, the goal is to provide the 'Just Works’ experience. The Linux model is that IHVs get the source code for their driver accepted into the mainline kernel. This entails a public peer review process to ensure that the driver code is of sufficient quality and does not have obvious bugs or security risks.” (Italics mine)

So, obviously, what this “Just Works” experience entails is, as the lkml announcement states, “corrupt linked lists, corrupt page tables, and just plain 'weird’ crashes”. Otherwise, could someone from the kernel development team please step up to the plate, be honest for once and just admit that whatever “model” you are following simply does not work instead of lying to the public about the non-existent correlation between “open-source” and “sufficient quality” or continually shifting blames to unaffiliated third parties for things that you yourselves have promised but failed to deliver?

This is not even some arcane, technical thingamajig. It’s just good ol’ fashioned Adulthood 101. If a stable ABI is what you are incapable to deliver, just say so. Making an excuse about “Just Works” despite evidence to contrary is simply irresponsible and, quite frankly, an insult to humankind’s collective intellect.

#1 Posted by ReverseControllerSE on Oct 12, 2011 6:08 PM

Is it possible that the Idiots of Durden actually believe what they say?

That it’s not an excuse, but rather a complete failure to understand what’s going on?

#2 Posted by administrator on Oct 12, 2011 7:07 PM

This would be akin to Adobe asking all plugin developers to submit their plugins for review and then compiling them all into the main Photoshop binary.

#3 Posted by DigitalAtheist on Oct 12, 2011 7:16 PM

I would be happy if they could actually reach the “Just works” phase. As is, most of the time it is more like “just sorta kinda works in a half-assed way if you dangle your dong over a hungry alligator and squint your eyes really hard and tight at the computer”.

Linux is a laugh riot when it comes to the “just works” deal since it never “just works” but demands that YOU work your ass off looking for ways to make it “kinda sorta work” in a manner that almost, but not quite, resembles what the hardware maker originally intended.

#4 Posted by imgx64 on Oct 12, 2011 11:29 PM

“This entails a public peer review process to ensure that the driver code is of sufficient quality and does not have obvious bugs or security risks.”

Which they just did in the LKML email you linked. They reviewed it, decided it’s not up to their sufficient quality, and tainted it so that they don’t get a barrage of automatically-sent bugs. Theoretically, this would force the vbox driver maintainers to fix the bugs to get their driver untainted.

I’m still a bit worried about the actual patch, it seems that the file “kernel/module.c” is full of lines like:

if (strcmp(mod->name, “modulename”) == 0) add_taint(TAINT_REASON);

I’m not a kernel developer, but wouldn’t that be better as, say, a couple of arrays and a loop?

#5 Posted by JoeMonco on Oct 13, 2011 12:47 AM

“Which they just did in the LKML email you linked. They reviewed it, decided it’s not up to their sufficient quality, and tainted it so that they don’t get a barrage of automatically-sent bugs.”

If you ignore the time between the driver went into the mainline and it getting the “tainted” status, sure. Otherwise, it’s just another blame-shifting excuse made by the same people who ought not act like a bunch of 10-year-old children.

And, honestly, I have seen 10-year-olds acting more responsibly than that.

#6 Posted by JoeMonco on Oct 13, 2011 12:51 AM

“I’m still a bit worried about the actual patch, it seems that the file “kernel/module.c” is full of lines like:

if (strcmp(mod->name, “modulename”) == 0)
add_taint(TAINT_REASON);

I’m not a kernel developer, but wouldn’t that be better as, say, a couple of arrays and a loop?”

You must be new to this Linux thing, then.

#7 Posted by ChrisTX on Oct 13, 2011 2:35 PM

“I’m not a kernel developer, but wouldn’t that be better as, say, a couple of arrays and a loop?”

No and neither is to use a hash value instead of a full string compare.

#8 Posted by Adam_King on Oct 13, 2011 4:49 PM

It’s amateur kernel developer hour in this thread.

#9 Posted by administrator on Oct 13, 2011 5:06 PM

Yeah Adam, not like your professional kernel development, right? Which lines did you contribute again?

#10 Posted by administrator on Oct 13, 2011 5:07 PM

Ironically, Microsoft has contributed more to the Linux kernel than Adam King ever has.

#11 Posted by DigitalAtheist on Oct 13, 2011 5:34 PM

Adam, Please highlight the lines of code you have contributed to ANY project. Oh wait, you are still in high school pretending to be in college and pretending to have a job that pays for it all and pays you you and actually lets you work… riiiiiiiiiiiiiiiiight!!!!!!!!!!!!!eleventy1111.

#12 Posted by Adam_King on Oct 13, 2011 6:09 PM

DigitalAthiest has the wit of a sewing machine operator.

#13 Posted by administrator on Oct 13, 2011 6:45 PM

I didn’t realize sewing machine operators were known to be witless. Where can I look up this fact, Adam?

#14 Posted by Adam_King on Oct 13, 2011 10:17 PM

DigitalAthiest,

Oh snap. It looks like Administrator just laid the smackdown on your ass! You gonna take that?

#15 Posted by JoeMonco on Oct 13, 2011 10:19 PM

It looks like Adam has borrowed his sister’s computer again.

#16 Posted by ReverseControllerSE on Oct 14, 2011 1:16 AM

Borrowed?

I think she doesn’t know Adam’s playing with her stuff again…

When she finds out…

#17 Posted by administrator on Oct 14, 2011 2:11 AM

Poor kid. No wonder he wants everything to be free. He lives in a matriarchal household where he has to make due with the scraps of computer time his sister isn’t using.

#18 Posted by Gesh on Oct 14, 2011 4:32 AM

“It’s amateur kernel developer hour in this thread.”

Isnt that the level of the most kernel developers anyway? Wasnt Con Kolivas an anaesthetist or something (and his scheduler wasnt that bad actually)? They are like the Luxemburg’s national soccer team – bakers, dentists, salesmen – who pretend to play soccer and sometimes accidentaly score a goal. Or rather the competitions scores an own goal. (Yes, Im (somewhat) kidding)

And on the side note, I read somewhere (cannot remember where)that the kernel developers group is aging – apparently the younger people have something better to do.

#19 Posted by ReverseControllerSE on Oct 14, 2011 6:58 AM

Yeah, it’s called “not using UNIX or any UNIX like crap”...

... hm…. methinks it’s an anti-TM really

Something Better To Do With One’s Time™

—————————————

Also, will the whole UNIX cretinism simply die out over time?

Is there hope for a brighter future?

#20 Posted by KimTjik on Oct 14, 2011 7:19 AM

“Also, will the whole UNIX cretinism simply die out over time?

Is there hope for a brighter future?”

What represents a brighter future? Has something new, whatever it is, a chance to compete with either traditionally established or monetary strength? Dennis Ritchie recently died – I haven’t even seen one comment about it here – but do we even have a trend suggesting that the totality of Unix or unix-related is stagnating?

#21 Posted by ReverseControllerSE on Oct 14, 2011 8:15 AM

A brighter future without the cretinisms inherent in the UnixWayOfDoingThings™.

See the Unix-Haters Handbook:
http://simson.net/ref/ugh.pdf

————————————

Has something new, whatever it is, a chance to compete with either traditionally established or monetary strength?

You mean something like Windows NT(?), it’s the other way around, really – UNIX cannot compete with NT… at all; but it lingers on, when it should die in fire already.

#22 Posted by KimTjik on Oct 14, 2011 11:07 AM

Windows NT? Do you seriously mean that the future becomes brighter if everything runs on top of Windows NT, like it represents the ultimate code base for every device? I don’t think Linux is the ultimate answer either, and Linus Torvalds himself has stated several times that he expects something better to come. Even if you hate the competition by so called unix-like systems, without competition, even imaginable, progress will become painfully slow.

#23 Posted by Linsuxoid on Oct 14, 2011 2:10 PM

@KimTjik
I believe that he told about something new on the market. NT is not something new. Yes, NT is surprisingly robust and resilient (almost no signs of design rot over 20 years – I don’t know ANY OTHER project of this scale and this quality), but it will show its age sooner or later.

So it would be nice if some new OS research projects wouldn’t stick to concepts and designs that have already shown its age 20 years ago. Ironically, Microsoft seem to be the only one to do this (with Singularity), every other OS research project I could remember of bakes POSIX stupidity right into low-level OS concepts.

#24 Posted by Linsuxoid on Oct 14, 2011 2:32 PM

Oh, I remembered another example: microkernels. I’m looking at L4. It seems that they’ve fixed most (all?) performance problems (or rather trend to virtualization/consolidation helped them: Intel and AMD implemented tagged TLBs on x64 which solves biggest microkernel problem – TLB flushes) and are ready for prime time.

#25 Posted by JoeMonco on Oct 14, 2011 5:57 PM

“Even if you hate the competition by so called unix-like systems, without competition, even imaginable, progress will become painfully slow.”

So, what do you actually care about? “Unix-like systems” or the competition? And as long as there is competition, why should anyone care about if it’s “unix-like” or not?

Tether your horse to the cart. Now move!

#26 Posted by DrLoser on Oct 14, 2011 6:12 PM

@Linsuxoid:

L4 and its variants are the competition. In fact, given the parlous state of OS development at the moment, they’re the only game in town.

Slathering mulch upon mulch, which is the Linux kernel way, is hardly going to inspire a fresh new wave of OS development.

I’ve said it before and I’ll say it again. A capability-based OS on an L4 micro-kernel. If I had a tenth of Shuttleworth’s money, that’s the way I’d go.

But apparently Mark wants to be a world-famous twat rather than a hero. His choice, and I can only applaud him for looking inside himself and discovering the Inner Loon.

#27 Posted by KimTjik on Oct 14, 2011 6:59 PM

@ JoeMonco

You missed my point, or I didn’t make it clear: I don’t care if it’s unix-like or not. When something better comes I wouldn’t hesitate to support/use it. If it’s unix-like, adhere to POSIX, or not, is frankly quite irrelevant. On the other hand it doesn’t make sense either to pretend that NT is the answer.

#28 Posted by JoeMonco on Oct 14, 2011 7:06 PM

“On the other hand it doesn’t make sense either to pretend that NT is the answer.”

It’s not the answer (whatever that’s supposed to mean). It’s nonetheless a solution that I most definitely prefer to any *nix alternative at the present. Unless your point is “anything but Microsoft“, I honestly have no idea what you are trying to get across here.

#29 Posted by KimTjik on Oct 14, 2011 7:21 PM

If you want to you understand what I tried to say in the initial post you responded to. A claim was made that if only unix and its “cousins” disappeared everything will become brighter, with the addition that NT practically is all what we need. I don’t agree with that. Sure my opinion differs in that I see merits in unix-like systems, but neither them or NT is a final solution to the quest for a good operating system.

You choose to pick one piece of what I wrote and turn it into some kind of argument. In your responses to me it looks like you would prefer me to follow the line of “anything but Microsoft”. It would be so much easier, wouldn’t it? I’ve to disappoint you: I don’t.

#30 Posted by JoeMonco on Oct 14, 2011 7:32 PM

“A claim was made that if only unix and its “cousins” disappeared everything will become brighter, with the addition that NT practically is all what we need.”

Seriously, I wouldn’t care if all *nix systems drop off from the face of the earth tomorrow. That would arguably be one less thing I need to worry about.

“I don’t agree with that.”

Sure, given the fact you don’t even know what the right tools for managing Windows machines are as a Windows admin, you are totally not begging for ridicules here.

“Sure my opinion differs in that I see merits in unix-like systems, but neither them or NT is a final solution to the quest for a good operating system.”

No one says it this, but at the present, you do you have to suggest as an improvement to NT?

None? Figured.

#31 Posted by JoeMonco on Oct 14, 2011 7:32 PM

*what do you

#32 Posted by DrLoser on Oct 14, 2011 7:32 PM

What does anybody?

#33 Posted by DrLoser on Oct 14, 2011 7:43 PM

@KimTjik:

“A claim was made that if only unix and its “cousins” disappeared everything will become brighter, with the addition that NT practically is all what we need.”

Really? When? Where?

And while you’re digging that up, let me say that NT is far from everything we need … although you might have picked a less favourable adverb. “Practically” is pretty much it, at the moment.

I’m going to come straight out with this and admit that my fundamental belief is that Linux (on the desktop, demonstrably) is sh1t is precisely because it is a bad clone of Unix that depends heavily on its Unix roots. X. No more need be said.

(Although the weird Linus fixation on gotos in the kernel is another argument, m’lud, that he really doesn’t have a clue on the real aims of an OS.)

I have never yet seen anybody on any site I have ever posted on claim that Unix should “just disappear.”

First of all, this would be farcical.

And secondly, if you actually work with computers every single mother fonkin day of your life, you work with hopelessly compromised systems that you have no real control over.

Unix is by no means the worst of those … but, even on a server, it is certainly one of them.

#34 Posted by DrLoser on Oct 14, 2011 8:01 PM

Oh, and just to make this as obvious as I possibly can, Joe Monco is 100% right here.

What sort of funkin moron would insist on the code for device drivers being given over to a bunch of unqualified nobodies and “put into the main line”?

Can you think of a single other OS that has ever done this?

And, while you’re thinking about that, do you really, seriously, think that this is how IBM builds their Linux boxen?

Frack, you people are sodden with what Adam apparently thinks is alcohol, but I believe otherwise:

You were born without the ability to think straight, weren’t y’all?

#35 Posted by DrLoser on Oct 14, 2011 8:06 PM

Everything else is just (my) petty insults, btw.

If you cannot make your drivers work correctly, you have no claim to being a useful operating system.

The clue is in the gerundive.

#36 Posted by ReverseControllerSE on Oct 14, 2011 8:24 PM

It’s called autism, Doctor.

And yes, autists can’t think straight, sadly.

#37 Posted by imgx64 on Oct 14, 2011 11:11 PM

@KimTjik
“You choose to pick one piece of what I wrote and turn it into some kind of argument.”

I concur! I’ve had exactly the same argument with him before.

——
@DrLoser
“Really? When? Where?”

http://www.tmrepository.com/fudtracker/well-look-after-that-for-you-we-swear/#comment_6996
“Also, will the whole UNIX cretinism simply die out over time?”

http://www.tmrepository.com/fudtracker/well-look-after-that-for-you-we-swear/#comment_6998
“A brighter future without the cretinisms inherent in the UnixWayOfDoingThings™.”

http://www.tmrepository.com/fudtracker/xkcd-knows-the-truth-about-linux/#comment_6965
“Will abandoning *nix save [us] all?”

#38 Posted by JoeMonco on Oct 15, 2011 12:51 AM

@DrLoser
“Really? When? Where?”

I am not Dr.L, but I believe I know enough of what’s going on here to clue you in… Or, at least, I care just exactly enough to read past the first sentence.

“And while you’re digging that up, let me say that NT is far from everything we need … although you might have picked a less favourable adverb. 'Practically’ is pretty much it, at the moment.” (Emphasis mine)

Nice work pointing out the obvious fact that those who hate Unix hate Unix. And since you have got so much time digging up all that pointless crap we already know well about (since, you know, those words you have cited there have always been our opinions), why not just spend a bit of it ruminating on the following instead?

http://tmrepository.com/fudtracker/well-look-after-that-for-you-we-swear/#comment_7035

“Seriously, I wouldn’t care if all *nix systems drop off from the face of the earth tomorrow. That would arguably be one less thing I need to worry about.”

#39 Posted by imgx64 on Oct 15, 2011 2:40 AM

I don’t know what’s all this is about. DrLoser asked where has anyone claimed that a brighter future will come if Unix disappeared, and I answered.

Consequently, where have I ever pointed out that people who hate Unix hate Unix? (Yes, that’s a trick question; if you answer, I’ll say that you’re just stating the obvious)

——
Anyway, I don’t see any comparison between Unix and NT. Unix does not exist anymore, now it’s just a historical set of ideas, conventions, and a half-baked standard.

So, I have to ask (and please feel free to state the obvious to get through my thick skull), what exactly is this Unix you want to disappear? And what do you suggest that operating system developers outside Microsoft (these will always exist) do? They can’t actually implement NT, and any non-Unix OS will have to have all software written for it from scratch (not like POSIX automatically makes things cross-platform, but it provides a starting point).

All current Unix-based operating systems have added lots of non-compatible improvements, so it’s not like they’re stuck because they depend heavily on their Unix roots.

Actually, DrLoser, can you explain why Linux’s Unix roots are preventing it from becoming better? And what do closed source drivers have to do with Unix?

Few (possibly obvious) points before JoeMonco misinterprets a small detail and attempts to rebuke it:
1- I do not think that Unix is all that great. It has more warts than a frog.
2- Like KimTjik, I would love to see a successful operating system not based on Unix. Plan 9 and Haiku come to mind. They’re very interesting, although not practical for daily use.
3- NT is a good operating system, I never claimed otherwise. If I was a decision maker at some company, I would consider it along with other options (luckily for me, I’m not a decision maker).

#40 Posted by imgx64 on Oct 15, 2011 3:05 AM

“And what do closed source drivers have to do with Unix?”

Aah, ignore this. I forgot the topic of the page, sorry.

#41 Posted by Linsuxoid on Oct 15, 2011 4:15 AM

> So, I have to ask (and please feel free to state the obvious to get through my thick skull), what exactly is this Unix you want to disappear?

Unix has UGLY and inefficient fundamentals: security, file operations, process management etc.. API-s are not only ugly but also often non-existent (CLIIsAPI).

But it’s not the reason for it to disappear. Reason is that there are many smart people that could actually take OS design forward if only they don’t make laps around this ugly design.

> And what do you suggest that operating system developers outside Microsoft (these will always exist) do?

And now THAT’s why. There is NOTHING outside of NT.

XNU actually comes close, I actually like both Mach and IOKit parts (I prefer IOKit over WDM/WDF). But they not only have BSD thing in between, but also expose all that ugliness outside of kernel.

Plan 9 is not designed to be better OS, it’s designed to be better Unix (and failed to do so). As for Haiku, there are lots of things I don’t like in it (and surprisingly most of them came from Unix), but I must admit they’ve got pretty far from most of Unix, so it’s one of the best choices for Linux-on-desktop replacements.

#42 Posted by JoeMonco on Oct 15, 2011 5:04 AM

“I don’t know what’s all this is about. DrLoser asked where has anyone claimed that a brighter future will come if Unix disappeared, and I answered.”

Then maybe you should start learning how to be patient and read past the first loving sentence. Or even just read the previous post perhaps starting with “A claim was made” rather than digging yourself into a deeper hole by lumping together opinions from three different people into one.

“Consequently, where have I ever pointed out that people who hate Unix hate Unix? (Yes, that’s a trick question; if you answer, I’ll say that you’re just stating the obvious)”

Genius. Do you even know what “trick question” is supposed to mean? Maybe you should go back reading the previous posts starting with the word “Really” to get a better idea of that (and perhaps past the point of “And while you’re” if your attention span permits).

“Actually, DrLoser, can you explain why Linux’s Unix roots are preventing it from becoming better?”

So you have all the energy to spare for all that tangential crap that you have dug up but can’t even spend a minute piecing together all that stuff that have been discussed about POSIX, pipes and all from within here?

Anyway, if you want an answer from Dr. L, I think you can find a barrel-load of it from his blog if not here in the comment sections:

http://drloser.blog.co.uk/2010/11/06/back-to-the-1970s-future-9926795/

“Aah, ignore this. I forgot the topic of the page, sorry.”

Why did I even bother responding to this?

#43 Posted by ReverseControllerSE on Oct 15, 2011 6:16 AM

Thanks for posting that link to Doctor’s blog, Joe. I just had a blast rereading that article.

(I’ve completely forgotten some of the old posts our good Doctor wrote…)

#44 Posted by KimTjik on Oct 15, 2011 6:48 AM

DrLoser, you spin far beyond what’s necessary. As I already wrote as a response to JoeMonco, it looks like you prefer someone to be a radical of any side, which makes the process so much easier: either the person is with you, or against you.

I see that JoeMonco still repeats his own interpretation of something I probably wrote at Piestar (keep in mind that my less good intention at that time was to explore this unfamiliar rudeness and hostility), like I’m doing all administrative tasks on none-Windows computers (did, since I’m now a project leader in a totally different profession). That’s not really the case, but most of the time I work from some kind of unix-like PC connecting through terminal server while doing other network tasks directly from it. It’s totally out of context, but why not save you some time repeating the same thing over and over again.

So there you have some other sentences to clip and paste from as a basis for yet another argument.

#45 Posted by imgx64 on Oct 15, 2011 7:29 AM

“Then maybe you should start learning how to be patient and read past the first loving sentence. Or even just read the previous post perhaps starting with “A claim was made” rather than digging yourself into a deeper hole by lumping together opinions from three different people into one.”

So I read few lines and responded to them, the “claim” turned out to be “NT solves all problems” and not “the future will be brighter if Unix disappeared” like I initially thought (the former wasn’t what KimTjik said anyway). Big deal, you misinterpret and take things out of context it all the time too. I don’t know why you’re using such an angry tone.

By the way, they are only two people, not three.

——
“Genius. Do you even know what “trick question” is supposed to mean? Maybe you should go back reading the previous posts starting with the word “Really” to get a better idea of that (and perhaps past the point of “And while you’re” if your attention span permits).”

I just looked it up, and now I know the formal definition, hooray.

My question didn’t fit in the formal description, which is unfortunate. But that didn’t prevent you from not seeing my point (which is: I pointed the obvious because someone ignored the obvious, few comments above it (the whole thing turned out to be nothing but a big confusion and people putting statements in other people’s mouths, so it’s all water under the bridge now)).

——
“So you have all the energy to spare for all that tangential crap that you have dug up but can’t even spend a minute piecing together all that stuff that have been discussed about POSIX, pipes and all from within here?”

I didn’t spend that much time. Two of these are on this same page, and I had the third open in another tab anyway.

——
“Anyway, if you want an answer from Dr. L, I think you can find a barrel-load of it from his blog if not here in the comment sections:
http://drloser.blog.co.uk/2010/11/06/back-to-the-1970s-future-9926795/”

From the blog: “There are well-known and better ways of doing this sort of thing, but sightings are few and far between in the Unix world. It’s not because they’re impossible. I told you: it’s a matter of philosophy. They are simply unconscionable.”

Looks like we agree here. While Unix is not great, it can be fixed. The problem is that people refuse to fix it out of the belief that it’s perfect. Does that mean OS developers should just abandon Unix and have zero compatibility with everything that already exists? No, and people who understand this already break compatibility with Unix as I said.

Some (like Lennart Poettering) take it a bit too far and break too many things while attempting to fix Unix. Many Linux users and developers hate him, but I totally appreciate what he’s trying to do (being happy with the resulting software is a different matter though). Now if only Ubuntu moves to systemd instead of Upstart, I’ll be much happier.

#46 Posted by imgx64 on Oct 15, 2011 7:29 AM

“Why did I even bother responding to this?”

We already went through this: http://xkcd.com/386/

#47 Posted by ReverseControllerSE on Oct 15, 2011 7:40 AM

“So I read few lines and responded to them, the “claim” turned out to be “NT solves all problems” and not “the future will be brighter if Unix disappeared” like I initially thought (the former wasn’t what KimTjik said anyway).”

No.

You see, both of your assumptions are wrong.
You need to go, and read, again.

#48 Posted by ReverseControllerSE on Oct 15, 2011 7:57 AM

“Looks like we agree here. While Unix is not great, it can be fixed.”

Well, no, actually, it can’t be – the problem you see is in the UNIX culture.

The fixes required would not be tolerated by the UNIX faithful; those who could tolerate them, have migrated off of the Plane of UNIX Stupidity long ago.

Wasting time on UNIX, or rather, on fighting the UNIX (Durden probably) zealots, is neither a good idea, nor a productive cause.

————————————————

People interested in creating a bold new thing™ should not be using UNIX/POSIX/Durden-crapola as a starting point.

It’s like saying you want to create an awesome new car, but it has to be fully Yugo compatible and similar to it in every way.

Yeah, good luck with that plan (you’re gonna need it).

#49 Posted by imgx64 on Oct 15, 2011 8:28 AM

Just for kicks, I’ll say where I think I disagree about the topic of this page (which feels neglected with all the talk about Unix):

1- I do not think that Linux developers are as incompetent as you believe.
2- I do not think that the documentation of a piece of hardware (if there is any confusion about what this means, this is an example: http://realtek.info/pdf/rtl8139d.pdf ) should be treated like a trade secret. Competitors can’t steal anything from it, it’s just documentation on how the hardware interacts with the outside world. Hiding it makes supporting the hardware on other operating systems much harder.
3- I do not think the drivers of a piece of hardware are a trade secret either. If the manufacturer doesn’t want to open source its drivers, then fine, release documentation and let the OS developers do it themselves, they’re more than happy to do it.
4- I agree with the Linux Foundation link that JoeMonco posted ( http://www.linuxfoundation.org/collaborate/workgroups/technical-advisory-board-tab/linuxdevicedrivermodel#Linux_Device_Driver_Model ).

#50 Posted by ReverseControllerSE on Oct 15, 2011 8:59 AM

1. They are incompetent on a systemic level, whether or not they can “hack well” is entirely irrelevant.

2. It’s up to the manufacturer.

3. It’s up to the manufacturer.

4. Durden driver model is insane – a bunch of gits that know next to nothing about the hardware are supposed to make good drivers, really, you sure about that?

Oh wait, you mean they just get vote on whether your drivers get into the kernel… riiight, I can just see the manufacturers lining up for this. :/

—————————————————

Here is what it really comes down to: they know, that there is no reason for anyone to support their crazy little hobby, so they are trying to force all manufacturers and programmers and authors in general to release all of the source for all things, so that the Durden-Hackers can then support themselves.

Unfortunately, that’s a model from the late ’70 UNIX – it presupposes a much simpler capability set and very few requirements.

It just doesn’t work (well) for a modern system.

———————————————-

So if they want to play with their hobby OS, they can do so in a VM specifically made for that purpose (the name DurdenBOX comes to mind).

As far as the actual OS running the hardware goes, Windows will do just fine, they can even install a Dreck-Brown visual style (I can post links if anyone wants them), bash and even an alternate shell.

Then they can pretend very hard that it is actually DurdenOS and be merrily on their way.

#51 Posted by DrLoser on Oct 15, 2011 12:45 PM

@IMGX64:

“Looks like we agree here. While Unix is not great, it can be fixed.”

A pretty pallid form of agreement, if you ask me. Here are a few fundamental problems with Linux (I’ll get back to Unix later):

  • Ludicrous device driver model. I’m sorry, but it is. No other OS does this, and for good reason. Sort that out (grit teeth, accept blobs, produce a stable ABI) and you begin to have an Operating System. Continue as is, and you have a hacker’s paradise. The two are exclusive.
  • X. Seriously?
  • Security. Yes, you can subscribe to the LSB, or whatever it’s called, but you’re only going to do that if you’re the DoD. It’s very limiting for normal uses. Security should be fundamental; it should not get in the way.
  • Fundamentally broken tool chain. Difficult to fix. Absolutely everything depends upon the fundamentally broken tool chain.
  • User support. Genuinely lousy. As has been pointed out, if you don’t make money building stuff, but you do make money supporting it, you have a massive incentive to get it wrong in the first place.

And a few other things, but that’s good for starters.

How much of this applies to Unix, in a greater sense? Well, a depressingly large amount. The access-rights model is broken and cannot be wished away. You’d have to put an insane amount of effort into drivers and X and the rest to make them workable, and then you’d break everything going.

The semi-solution, which in reality was not much of a solution at all, was supposed to be Java. But Java is still designed around the CLI (classpaths, ugh) and is a prime example of sclerotic inflexibility (“our way or the highway!”) and is, frankly, inelegant. Inelegance is a key feature of Unix.

So, yes, you could start with Unix and build something better. As RC has pointed out, you’d have to throw just about everything out (apart from the Posix interface, also available on Windows) and start again.

Now. Why would you want to do that?

I can just about see a commercial Unix vendor getting somewhere with this mess. Any given combination of a Linux kernel, a Linux upstream, and a Linux downstream doesn’t stand a snowball’s chance in hell.

RPM, Linux.

#52 Posted by DrLoser on Oct 15, 2011 12:50 PM

And I wish to point out that I have made a very good living out of Unix, and even (occasionally) Linux. I would hate to see that part of my living disappear. I am no shill, but I genuinely wish that somebody could come up with a better alternative to Windows, if only for variety.

I think you could probably start with a Mach/L4 etc kernel and build out from there.

But you cannot possibly start with this cringing mess that cowers in a corner with a tin cup and begs everybody else to help it get a room for the night.

#53 Posted by DrLoser on Oct 15, 2011 12:58 PM

Oh, and also: when I said that “security should be fundamental,” I missed out the obvious requirement that “security should be pervasive.”

This nitwit insistence on “mechanism, not policy” is at the heart of why it is impossible to build a decent system over a networked *nix.

One simple stupid farkin KDE or Gnome program that depends upon rwxrwxrwx* and you’re screwed.

Kernighan, Ritchie and Plauger deliberately built Unix on shifting sands back in the early seventies, because the sands were shifting and nobody can predict where the next dune is going to be.

They’re stable now. Time to start again.

From scratch.

#54 Posted by DrLoser on Oct 15, 2011 2:21 PM

@KimTjik:

“Windows NT? Do you seriously mean that the future becomes brighter if everything runs on top of Windows NT, like it represents the ultimate code base for every device? I don’t think Linux is the ultimate answer either, and Linus Torvalds himself has stated several times that he expects something better to come. Even if you hate the competition by so called unix-like systems, without competition, even imaginable, progress will become painfully slow.”

Lousy argument. Try again.

Unix might have worked. Linux does not.

Competition requires competition. There is no *nix competition.

It isn’t about “the ultimate choice.”

It’s about people not screaming at you for no obvious reason.

It’s about using what you have now, and hoping for something better.

“Better” is what we who played with Linux in about 2000 or so were looking for. Even the possible “future” of better.

Didn’t happen, won’t happen, best to start again, possibly with another kid in the duck-pond in a basement, and at least find a way to ignore the intrusive morons this time.

It’s a dream. Maybe even Torvalds wakes up in the middle of the night, sweaty, with this dream.

But at this point it isn’t going to be a mirage like the Perfect Unix.

Sorry 'bout that.

#55 Posted by KimTjik on Oct 15, 2011 3:21 PM

It’s quite difficult to follow you general argument DrLoser, because it’s floating without addressing any certain implementation of for example the Linux kernel. I didn’t even mean that unix-like systems have to represent the competition, just that it’s important that something is. Notice I that I wrote “even imaginable” competition.

Let’s say we agree that Linux is flawed, but you yourself had some hopes at a certain time. That was a dream that some in the industry interpreted as a real threat, didn’t they? Would you deny that this “imaginable” competition hasn’t increased the pace in how certain things have improved?

Depending on implementation unix-like systems represent competition, or did you retreat to only talking about traditional PCs? You quite often point out that some Linux users live in denial, but the “there’s no *nix competition” looks like the same kind of denial to me. What’s the difference between that and the “year of Linux” mantra?

I agree with IMGX64 that it’s quite difficult to start over totally since you need some kind application base. In a sense it doesn’t matter how clever and fantastic an operating system is, if there’s nothing to do with it. Fortunately and at the same time unfortunately exist few accessible choices except GNU ones. Who would have both the interest and the resources to build a whole ecosystem of software from ground up? It’s not like you’re all the smartest in contrast to the rest all being stupid, and if you got some spare minutes you would fix it.

Did I say that there’s an ultimate choice? No, I didn’t, but someone else implied something already is. Did I say that I wait for the perfect Unix? No, I didn’t, and I don’t even know if anyone here did. Is this where I should write “try again”?

#56 Posted by imgx64 on Oct 15, 2011 3:32 PM

To reiterate, you can’t expect a company to waste money doing R&D on a non-Unix OS that will surely fail. This is a job for researchers and hobbyists, which have sadly failed in doing so.

Here’s what Rob Pike has to say about it: http://doc.cat-v.org/bell_labs/utah2000/

tl;dr: The market wants Unix, the research follows the market, therefore there is not much innovation in OSs. The solution: researchers should stop following the market and start focusing on new and interesting, albeit unprofitable in the short term, projects.

#57 Posted by DrLoser on Oct 15, 2011 5:18 PM

“Notice I that I wrote “even imaginable” competition.”

Yes, I noticed that, too.

Absent Sir Thomas Moore’s Utopia, what exactly would “even imaginable” mean, in this context?

#58 Posted by DrLoser on Oct 15, 2011 5:26 PM

@IMGX64:

Obviously Rob Pike, a “Google Fellow” or some such, has become a senile worthless old man.

tl;dr? Jesus, mate, that whole onion-on-a-buckle rant summed up to two thousand words, max.

And it was hilarious.

“If you claim that’s not innovation, but copying, I reply that Java is to C++ as Windows is to the Macintosh: an industrial response to an interesting but technically flawed piece of systems software.”

What? What? Who? Did anybody mention an industrial response? Flaw me baby, flaw me!

What is Systems Research these days?

“Mostly, though, it’s just a lot of measurement, a misinterpretation and misapplication of the scientific method.”

Which is a shame. Once upon a time, it was about more important things. Just ten minutes ago, I had never heard about “Systems Research.” This is IMPORTANT.

Why oh why oh why can nobody explain to me what Systems Research is?

#59 Posted by DrLoser on Oct 15, 2011 5:28 PM

I’m sorry, but that whole mea culpa is so full of sh1t that I’m just going to have to throw it open to the lads.

#60 Posted by Linsuxoid on Oct 15, 2011 6:16 PM

I want to take on that “but anything new would die without compatibility with POSIX”.

Microsoft changed underlying kernels three times: DOS->Win9x->WinNT (yes, I know that WinNT has been created before Win9x, but real transition on the client started with Win2k and finished during XP). Those OSes share almost no underlying concepts, yet have almost perfect backward compatibility.

Apple broke with their back compat 3 times: MC680×0->PPC on classic, then transition to OSX and then PPC->x86 on it (it’s rumoured that they’ll move to ARM on OSX). Everything with reasonalbe, yet not perfect backcompat.

Isn’t it surprising that these two have biggest ecosystems?

Linux trying (and failing) to be compatible with ten-twenty thousands of half-assed clones of about a thousand of proprietary applications. And the price is stagnation in the OS development.

What’s worse, most new OSes develop arond same stupid Unix heritage. Yes, you could implement compatibility layer to leverage existing ecosystems (in a wait for your own to emerge), but for goodness sake, try to find something fresh and innovative – don’t stick to that superuser + 12 bits of security nonsense, don’t try to pretend that forking whole address spaces are any good (fork is one of the reasons Linux has OOM Killer – both are quite stupid), try to make I/O asynchronous (all async I/O is non-blocking, but not all non-blocking I/O is asynchronous) – even so called web programmers understand benefits of async I/O (NodeJS is kind of hot new thing these days), try not to use unstructured (human-readable) octet streams to pass structured data around, try to make consistent resource management (no timers and events are NOT files) – that will also help to make centralized access control (concept of reference monitors is of the same age as Unix itself, and no LSM doesn’t even come close), and so on, and so on.

#61 Posted by JoeMonco on Oct 15, 2011 6:38 PM

“It’s quite difficult to follow you general argument DrLoser, because it’s floating without addressing any certain implementation of for example the Linux kernel. I didn’t even mean that unix-like systems have to represent the competition, just that it’s important that something is. Notice I that I wrote 'even imaginable’ competition.”

You still don’t get it, do ya? At this point, in this fabric of space and time (unless you are from some weird alternate universe – that’ll explain a lot about the “points” both you and IMGX64 are trying to make here), there is no competition. Maybe something in the near future – something unforeseen – will take over the world by storm. But given how technology evolves throughout history, that’s unlikely. And that brings us to our next point.

“Depending on implementation unix-like systems represent competition, or did you retreat to only talking about traditional PCs? You quite often point out that some Linux users live in denial, but the 'there’s no *nix competition’ looks like the same kind of denial to me. What’s the difference between that and the 'year of Linux’ mantra?”

If you are talking about market shares, with a mixed bag of price- and non-price competitions, availability of human resources, availability of applications and such, sure. Otherwise, take a look at this:

http://www.mysql.com/common/images/marketshare/gartner-database-deployment.png

Bear in mind that SQL Server would not run on anything but Windows, and it was the most popular thing next to Oracle DB, which had always had fully featured version for Windows, and as we all knew, Ellison the Larry would not do anything if there’s no money in it. So much for “traditional PCs” or whatever.

Having said all that, we are talking about competition technology-wise. We know that historically, matters of this nature have always turned more along the line of VHS vs BetaMax than anything else, but, regardless, this has always been my point (and perceivably Dr.L’s). You are just missing it here in its entirety.

#62 Posted by JoeMonco on Oct 15, 2011 6:38 PM

“To reiterate, you can’t expect a company to waste money doing R&D on a non-Unix OS that will surely fail. This is a job for researchers and hobbyists, which have sadly failed in doing so.

Here’s what Rob Pike has to say about it[...]”

We get it – Rob Pike is you hero. Whatever he conjures up from atop a toilet seat has to be right then, right?

And, hobbyists? Seriously, is anyone up for hiring a hobbyist for anything? I sure ain’t.

And, honestly, what you think people mean by “R&D”? “R&D” stands for “research and development”, and merely repackaging something that already exists hardly constitutes “research” or “development”. That’s what they call “making a distro”, and I have seen plenty of it coming and going for more than 10 years now.

Even if you manage to come up with yet another flavor of Unix, will that put you in the position to compete with NT? Not a chance. NT has always proved to a late-comer in every facet of the computing world from netbook to supercomputer, yet it has always managed to eat a good chunk out of what Unix has. The moment you decide to build upon Unix is the moment you have set your venture on a dive head-first into a market segment that is saturated with the same thing – you simply can’t win. The reason Linux won out in the first place over other *nix system was that people could do Linux on the cheap. If your focus is differentiation, the Unix route is a dead end.

In short, you can’t do differentiation with Unix. There is no more you can differentiate your product from the rest of the *nix stuff out there than you can tell one twin brother from another by just looking at their mere appearances. People don’t care how much your *nix is different from the rest of the *nixes – they just want something that looks like System V and nothing more. As ironic as it sounds, that’s the market for you.

#63 Posted by JoeMonco on Oct 15, 2011 6:47 PM

“I’m sorry, but that whole mea culpa is so full of sh1t that I’m just going to have to throw it open to the lads.”

Ever been to a high-school debate? Whatever sketchy idea you have about whatever subject matter in question counts as a good point, I tell ya.

#64 Posted by KimTjik on Oct 15, 2011 7:44 PM

Who you’re arguing with JoeMonco? What’s new about the database statistics? I have no argument with that, and haven’t even commented on it. Still the statistics is a mixed bag of… sorry they aren’t competitors, they just by coincidence exist in an alternative universe. The most comforting thing is that you’re like a tail, never allowing me to feel alone! That’s nice of you.

If I understand you correctly: – this time we’re talking about competition on a technology level – according to you and DrLoser there’s no competition to the NT kernel – whoever share a different opinion is a loon or liar

That then leads to a chain of arguments: – *nix isn’t a competitor because it has X market share – if the above doesn’t apply it’s still not a competitor because of NT kernel superiority technology wise

Would that be a layman’s summary of what you mean I don’t get?

Since you refer to Larry you also know that while he’s a screwed businessman his rage against Microsoft is quite unique. I even think you guys here are pussy cats in comparison. Do you think he cares whether you think the NT kernel is better designed, and/or that Linux/*nix is at best a badly designed compromise? If I understood you correctly – as described above – you don’t think so. He simply doesn’t care as long as Oracle makes more money. I suppose your analogy about VHS and BetaMax hence suggest that we still have to accept, while protesting (by means of TM Repository?), that sometimes bad beats good.

#65 Posted by JoeMonco on Oct 15, 2011 8:27 PM

“Still the statistics is a mixed bag of… sorry they aren’t competitors, they just by coincidence exist in an alternative universe.”

Interesting. Define this “alternative universe” you speak of. And, more importantly, elaborate how this is supposed to support your previous argument about “traditional PCs” that this particular point is meant to debunk.

Seriously, elaborate on your sketchy arguments, or I’ll elaborate on them for you with whatever interpretation I see fit.

“- this time we’re talking about competition on a technology level – according to you and DrLoser there’s no competition to the NT kernel – whoever share a different opinion is a loon or liar”

Nice try. However, consider that the other side has already presented in details the why and the how they disagree with your view, it is up to you to step up to the plate and come up with something substantial to defend your position rather that regurgitating the usual “I don’t agree” three-word vocabulary found in every other of your comments in this website. Blaming people for calling you out on your blatant laziness/evasiveness like this is just intellectually more insulting than claiming that your dog has eaten your homework.

“- *nix isn’t a competitor because it has X market share – if the above doesn’t apply it’s still not a competitor because of NT kernel superiority technology wise”

*nix is a competitor to what? That’s something I think you need to clarify, market-wise and tecnology-wise, along with the why and the how. The other side has already elaborated on both parts of the picture, and it is up to you to fill in the blanks and tell us what the loving hell you were getting at in the first place. And if you can’t fill in those blanks, again, I’ll go and fill them in for you – in whatever way I see fit.

“Since you refer to Larry you also know that while he’s a screwed businessman his rage against Microsoft is quite unique.”

Interesting, but irrelevant. When one day Oracle stops supporting Windows altogether, tell me all about that in grand details. Otherwise, give us something more substantial than some perceived personal grudge over whatever.

“Do you think he cares whether you think the NT kernel is better designed, and/or that Linux/*nix is at best a badly designed compromise?”

You wanted to talk about the market, and I gladly obliged. People obviously use Windows outside “tradition PCs” by the drove, and the market for complements reacts accordingly. If that’s not clear enough for you, then I don’t know what is.

“If I understood you correctly – as described above – you don’t think so. He simply doesn’t care as long as Oracle makes more money.”

As a CEO, I think that’s pretty much his job, isn’t it?

#66 Posted by JoeMonco on Oct 15, 2011 8:27 PM

“I suppose your analogy about VHS and BetaMax hence suggest that we still have to accept, while protesting (by means of TM Repository?), that sometimes bad beats good.”

I just found that a pretty good way to draw the line between the technology and market sides of the picture – in case the blame of your free-floating style argument got shifted onto the other side, again.

#67 Posted by KimTjik on Oct 16, 2011 6:04 AM

“Define this 'alternative universe’ you speak of.”

I just quoted your words, so I suppose you know what you meant. Am I obliged to write dry short sentences?

“Seriously, elaborate on your sketchy arguments..”

Seriously, it wasn’t an argument to prove that I don’t agree, but an attempt to single out your main opinions from your quite lengthy responses (in fairness it’s DrLoser who usually post several close to 500 word responses, not you, but you understand what I mean). Maybe you and DrLoser are in agreement that the main factor here is technology design, and put bluntly I suppose my ruff chain of arguments was correct.

There’s no reason to hide opinions in lengthy elaborations proving ones superiority in knowledge. You know a lot more than me about certain things, and probably I do the same about other things. I’m here to learn about your arguments, not to rebuke you. How am to learn about it if I don’t provoke any answers? Frankly, this place has a tendency to act like a cheerleader team, so I don’t think it hurts to have folks here who don’t fully agree, even though agreeing with some views.

Anyway, I try to summarise some of my views about *nix as a competitor:

- it looks like *nix has a pretty good position when it comes to pedagogic; even if DrLoser has shown that he views some of the fundamental design choices to be impractical/flawed, this design combined with it being open-source attracts many to dig deeper.

- in some niches, like the extreme TOP500, *nix will continue to dominate. It’s specialized and it has no direct “consumer” implication. I actually expected Microsoft to do better, since they pumped in a lot of money several years ago into their HPC line.

- the infrastructure of the Internet, from root-servers to some of the more popular services run on *nix. If someone chooses *nix instead of NT, it’s a competitive choice even if no direct money transaction occurred.

- smartphones and related is dominated by *nix operating systems on low level. Let’s see what happens but there’s no trend suggesting it’s going to change any time soon.

That could have been cut & pasted from anywhere, nothing new. I’m not qualified to evaluate *nix technology compared to for example the NT kernel. I cannot understand the above circumstances though as anything else than there exist competition. Since you all know about these circumstances, my question is why you don’t view that as competition?

#68 Posted by ReverseControllerSE on Oct 16, 2011 7:52 AM

“- it looks like *nix has a pretty good position when it comes to pedagogic;...”

No, the academia is filled with autists who are completely useless but are allowed to remain and hinder future generations.
Autists can’t stand change (any change, that is), they’re stuck in the ’70 and hate on NT, which they view as a threat to their lifestyle.

UNIX has no benefits to education, it has benefits for the autists-at-large employed there.

———————————————-

“- in some niches, like the extreme TOP500, *nix will continue to dominate. ...”

It doesn’t matter what they use, the customisations matter, they could just as well use NT, wouldn’t make any difference (except of course that they’re used to UNIX and don’t want any changes, and since the taxpayer’s money keeps flowing, no amount of issues caused by UNIX can force them to reconsider their choices).

———————————————-

“- the infrastructure of the Internet, from root-servers to some of the more popular services run on *nix. ...”

Either Linux, because it was for free, and/or some UNIX because their organisation has always used it (and either don’t want to change things if they don’t have to, or they simply can’t, since their solution is so tied to UNIX, that they are effectively locked-in).

———————————————-

“- smartphones and related is dominated by *nix operating systems on low level. ...”

Again, Android is used because it was available for free, also calling it UNIX is a stretch.

There is no competitive advantage in Linux, they can simply use it for free and offload all the financing to the society (which overpays products, which finance adverts, which finance Google, which then makes all kinds of stupid and useless products, and then pretends they are free and plays the role of St. Saviour).

———————————————-

“...my question is why you don’t view [UNIX/Linux] as competition?”

Because MS-DOS, later classic Windows and finally WinNT already destroyed UNIX, first on the Desktop, then on workstations and finally on the server (this last transition is not yet complete).

Linux on phones is a disaster, whether we consider security, usability, performance or just about anything else.
Were it not for the fact that Metro is one ugly POS, WinPhone would have already dealt a killing blow to the Durden-on-phones nonsense.

What you don’t understand is the fact that it is UNIX that had the upper hand, and Windows was the disruptive new technology that competed against it – and won (two decades ago).

The idea that UNIX or Linux could somehow be considered a replacement (a new challenger) for MS Windows, is patently absurd (even more so when you consider that they are stuck in the ’70 way of thinking).

#69 Posted by KimTjik on Oct 16, 2011 8:27 AM

“What you don’t understand is the fact that it is UNIX that had the upper hand, and Windows was the disruptive new technology that competed against it – and won (two decades ago).

The idea that UNIX or Linux could somehow be considered a replacement (a new challenger) for MS Windows, is patently absurd (even more so when you consider that they are stuck in the ’70 way of thinking).”

What you don’t seem to understand is that I understand this historical movements. I grow up in a family business moving between different kinds of unix, OS/2, MS-DOS, NT and finally a mix of NT, Linux and BSD. That was my playground in the 70’s as a kid.

I haven’t written anything about who’s challenging who, as who’s the new kid on the block. I only refer to how it still looks, and then it doesn’t matter what’s stuck in the 70’s thinking or not, or in which decade Microsoft is stuck. As someone else kind of wrote here, you use what’s available and affordable.

I haven’t invested anything in seeing either win, and I’ve no pride to defend, as if a bunch of code mean anything to me personally. Hence I don’t care for arguments like “it doesn’t matter what they use, the customisations matter…”, because whatever they use it affects the landscape of computing. If they don’t use NT they use something else, and if they would have used NT you for sure would have pointed out that as relevant, wouldn’t you?

If “free” is a factor is actually quite irrelevant to the question of whether it competes as well. “Free” is only a factor to understand why, not as a pretext to pretend it doesn’t exist. If “free” makes someone to not chose a product tagged with a price, it’s competing.

Don’t you see that you use the exact same logic as the one you criticise.

you say: “it looks like this, deal with it”

they say: “ok, but that’s because of reason A and B and probably also C, so even if it looks that way it isn’t like that”

Sure, if Microsoft wouldn’t have lacked common sense and in time understand what’s happening on the smartphone market, they probably would have had a better position. Microsoft will – at least to my expectation – improve they position, especially since they can target their existing customers with a solution integrating with other Microsoft products. Microsoft was stuck in old thinking. Code is one thing, and even if it’s theoretically beautiful beyond comparison, you got to sell it as well.

I think you should add one thing to why for example Android is successful. Sure it’s free, but manufacturers also needs something to put pressure on companies like Microsoft and Apple. It gives the leverage to cut better deals. In that context it doesn’t matter what flaws it has.

#70 Posted by KimTjik on Oct 16, 2011 8:29 AM

“... you got to sell it as well”

Should have continued:

“and with timing”

#71 Posted by JoeMonco on Oct 16, 2011 12:54 PM

“I just quoted your words, so I suppose you know what you meant. Am I obliged to write dry short sentences?”

So what exactly are you trying to suggest? That’s SQL server exists only in an “alternate universe” (not “alternative universe”, dipstick). Because applying those words in your context, that pretty much what they mean.

“There’s no reason to hide opinions in lengthy elaborations proving ones superiority in knowledge.”

I am sorry that you are hopelessly illiterate. Do I need to reword my argument along the line of your hollow, free-floating, three-word sentences so I, too, can pretend I have said something substantial without actually bringing anything to the table?

“I’m here to learn about your arguments, not to rebuke you. How am to learn about it if I don’t provoke any answers?”

Honestly, I don’t care if you walk out of here empty-handed or otherwise. No one is obliged to give you anything at any given time, and if you somehow believe that your provocation won’t result in angry responses, then, again, I wonder which weird alternate universe you are from.

“- it looks like *nix has a pretty good position when it comes to pedagogic; even if DrLoser has shown that he views some of the fundamental design choices to be impractical/flawed, this design combined with it being open-source attracts many to dig deeper.”

For someone obviously not even qualified for his own job, you sure have a lot to say about researching – provided that that’s what you mean by “pedagogic”(sic) – don’t ya?

Again, you need to elaborate on your argument. Digging further into a hill is “deeper”. Digging further downwards into the ground is also “deeper”. Digging harder into your nose with your finger is also “deeper”. Every research has an aim and a method. If your idea of “R&D” is along the line of aimlessly messing around with “open-source” code, then, I am sorry – you have missed the purpose of R&D by exactly the entirety of it.

Besides, there is nothing to be further understood about Unix. It’s not some mystery of the universe waiting for someone to unlock. It’s an operating system, a tool, a means to an end… Anything but a discovery. And if you somehow believe that there is anything to be discovered about Unix, then most certainly you don’t have what it takes to make anything useful out of the “source”.

Seriously, don’t make me laugh.

#72 Posted by administrator on Oct 16, 2011 1:47 PM

Actually, he was grammatically correct. It’s “alternative universe”. The word “alternate” is a verb; “to alternate between A and B”.

However, the term has been so thoroughly misused that I wouldn’t be surprised if dictionaries have simply picked it up as an alternative meaning.

#73 Posted by JoeMonco on Oct 16, 2011 2:09 PM

“- in some niches, like the extreme TOP500, *nix will continue to dominate.”

A little elaboration from you will sure help. I have some pretty good idea about what going on in that picture, but unless you can tell me what exactly you think of it, I ain’t gonna tell you nothin’.

“- the infrastructure of the Internet, from root-servers to some of the more popular services run on *nix. If someone chooses *nix instead of NT, it’s a competitive choice even if no direct money transaction occurred.”

Again, I may just have the answer on that one, but until you tell me what your view is, that’s not a single word you are going to get from me.

“- smartphones and related is dominated by *nix operating systems on low level. Let’s see what happens but there’s no trend suggesting it’s going to change any time soon.”

I have a Nokia 2730 classic, and I have absolutely no idea what you are blathering about.

“Since you all know about these circumstances, my question is why you don’t view that as competition?”

Again, what is your bottom-line? If what you care is technical merits, than give an example wherein such may be relevant to the whole picture and elaborate in details as to what your view is – I ain’t gonna do your damned homework for you. If what you have in mind is throwing frivolous questions at me for some good answers in return, then, I am sorry – please get lost.

#74 Posted by KimTjik on Oct 16, 2011 2:33 PM

Make fun of me if you so wish. I’m not sure if I use English terms correctly, but I for sure wouldn’t make cheap shots at someone trying to use some of the languages I master better than English.

You’re not obligated to give me anything, I agree about that. You’re not stupid, so it’s quite obvious you choose hostility instead helpfulness.

What concerns my main job and whether I’m qualified for it or not depend on skills to organise people and negotiate better deals for all involved. Maybe I’m not qualified, but it’s at least something you know nothing about.

No big deal, I find other sources of information.

#75 Posted by JoeMonco on Oct 16, 2011 2:45 PM

“You’re not stupid, so it’s quite obvious you choose hostility instead helpfulness.”

Or, you can just state your view in black and white and thus get the ball rolling in no time. It’s not really all that difficult to a functional adult. Hell, it doesn’t even to be in perfect Queen’s English.

Either way, my friend, either way.

And, until then, computer says no:

http://www.youtube.com/watch?v=2n0lL0vFvcg

#76 Posted by JoeMonco on Oct 16, 2011 2:46 PM

*need to

#77 Posted by ReverseControllerSE on Oct 16, 2011 3:18 PM

“What you don’t seem to understand is that I understand this historical movements. I grow up in a family business moving between different kinds of unix, OS/2, MS-DOS, NT and finally a mix of NT, Linux and BSD. That was my playground in the 70’s as a kid.
Well now, that’s a lot of nonsensical bullsh*t.”

I wasn’t aware that Linux, BSD and NT were around in the 70’s; are you sure you have good (or indeed any) grasp of this “history” thing?

Also, why in the world would a family business (a small company, I presume) overload their IT with three separate operating systems, two completely different usage paradigms and all of the problems that ensues?
I’m having a hard time believing you, but if it’s true, then your family’s company has some seriously twisted attitude when it comes to IT solutions.

————————————————————-

“I haven’t written anything about who’s challenging who…”

You have actually – you proposed that UNIX/Linux is the challenger to NT.

NT was the challenger to UNIX and it already won. None of the reasons behind that changed – UNIX/Linux is still stuck in the past, and Windows is still evolving.

————————————————————-

“ ...because whatever they use it affects the landscape of computing.”

Except that it doesn’t – nobody cares what OS runs on super computers (or washing machines for that matter). Operating system matters on the desktop and portable computers (where the user sees it and shops acordingly).

————————————————————-

“If “free” is a factor is actually quite irrelevant to the question of whether it competes as well.”

Actually it’s at the very core of it, nothing is really free, someone always pays – what happens when Google can no longer support Durden-phone development? What happens if they simply lose interest? What is the real cost of crappy Google projects that are always offered for free?

Can Durden-phone bundlers support what Google wrought? I think not.

-

“...If “free” makes someone to not chose a product tagged with a price, it’s competing.”

Except that it isn’t – working for free is not a sustainable business — you’re either price dumping (which is illegal), or you’re headed for bankruptcy.

Unless you’re arguing for authors to work for free (i.e. be slaves), like majority of freetards and pirates and the like.

————————————————————-

“It gives the leverage to cut better deals. In that context it doesn’t matter what flaws [Android] has.”

The only “leverage” that one can get from Android is patent infringement, and yes Apple took full use of that “leverage”, I don’t see how that helps Durden-mobile in any way shape or form.

Unless one takes to mean “harms” while saying “helps” (or, in your case, “better”, which means “worse”), we already have a name for that: DoubleSpeak™.

#78 Posted by KimTjik on Oct 16, 2011 3:56 PM

I answer one piece of your response:

“I wasn’t aware that Linux, BSD and NT were around in the 70’s; are you sure you have good (or indeed any) grasp of this 'history’ thing?

Also, why in the world would a family business (a small company, I presume) overload their IT with three separate operating systems, two completely different usage paradigms and all of the problems that ensues?”

If you want to you understand what referred to the 70’s. Don’t get childish, as if you didn’t understand that versions of Unix was used during the 70’s. Unix wasn’t in itself a choice, it was all decided by needed applications. Transitions were therefore more a necessary pain because of applications, not as a choice of platform.

Later in the 80’s it become quite uncertain as to which platform would get the best support for needed applications. OS/2 wasn’t really deployed, but at that time I had practise at IBM so I got exposed to it quite much. My father got involved in development of software for accounting, not so much coding but design. Anyway the platform become MS-DOS and later NT. I didn’t care so much; during elementary school I coded simple games for myself and later on I found other interests while simply working with economics.

Later on Linux and BSD were added for certain tasks. Main network and applications depend on Windows though.

I suppose we can close that chapter now, can we?

#79 Posted by JoeMonco on Oct 16, 2011 8:10 PM

“If you want to you understand what referred to the 70’s. Don’t get childish, as if you didn’t understand that versions of Unix was used during the 70’s.”

I didn’t realize you used to play around with things the size of bleeding wardrobes as a kid:

http://www.videointerchange.com/pdp-9.htm

Or are you by any chance referring to something along the line of Commodore PET?

http://en.wikipedia.org/wiki/Commodore_PET

Seriously, the only thing that you could (remotely) consider an “OS” was the BASIC ROM inside. You most definitely would not want to consider Unix with that amount of memory, that amount of computational power and that non-existent multitasking ability, and you most definitely would not be running anything without loading from tapes (or some giant 8” floppies) an app that would take up the entirety of the memory first. Anything relevant to your “experience” simply would not happen until at least the mid- to late 80’s. Even then, Unix was rarely present outside enterprise “3M” workstations and certainly would not be something you could play around with on a shoddy home computer.

This leaves us with the 90s onwards, and unless you are trying to tell us that you are some young prodigy having access to everything the corporate world had to offer in your line of work (which sounds rather unlikely), circa 1990 is pretty much as far as your “experience” can take you back to. Seriously, I has seen all kinds of bluffing, but this has got to be one of the most blatantly obvious thus far.

“Later in the 80’s it become quite uncertain as to which platform would get the best support for needed applications.”

I think it was pretty certain for those who actually needed those applications. Lotus 1-2-3 comes to mind:

http://en.wikipedia.org/wiki/Lotus_1-2-3

It was THE spreadsheet application even well into the 90s, internationally. Think of it as the MS Office of the time and you won’t be too far off from the reality.

“I didn’t care so much; during elementary school I coded simple games for myself and later on I found other interests while simply working with economics.”

BASIC. Sure some useful skill you had there, eh?

#80 Posted by KimTjik on Oct 16, 2011 10:33 PM

I’m not bluffing, but I apologize for referring to the 70’s when it must have been in 1980. Playground was meant literally, not like I as kid run anything on those computers; but I spent most of my days with father. As you correctly say, there wasn’t much to do. Hence you probably swept the field in vain for what I possibly could have done on those computers. Sorry, but I wasn’t intentionally misleading.

Remember Xenix? Not all related to Unix was that big. There was also some kind of odd operating system called Bridos, which had some in common with Unix structurally. I don’t remember any tapes, but 5.25” floppy disks were used, one more evidence of it being 1979-1980. Bridos could handle programs written for CP/M.

Lotus 1-2-3 wasn’t used so much, since accounting software were quite specialised and not integrated with third party applications even years after its release. Good for you if it was very obvious whether OS/2 or Windows was the right choice.

What’s so silly about a kid in school fiddling with BASIC and Pascal? If we are in the same age, what did you do at the time?

#81 Posted by JoeMonco on Oct 16, 2011 11:16 PM

“Remember Xenix? Not all related to Unix was that big.”

Still, hardly a hobbyist thing. What’s more, it was more popular that you think it was.

“I don’t remember any tapes, but 5.25” floppy disks were used[...]”

And roughly 5-100 times the capacity of a home computer’s memory circa 1980.

This is unless you are talking about machines like PC XT, which were by no means average household items:

http://en.wikipedia.org/wiki/IBM_PC_XT

And the damned thing runs Xenix.

#82 Posted by KimTjik on Oct 17, 2011 3:49 AM

Of course it wasn’t “a hobbyist thing”. It was for professional use, and not something we had at home.

You referred to Unix hardware’s physical size, and that’s why I said Xenix wasn’t big. It wasn’t a comment about how widespread its use was.

No the IBM PC XT came of course later. The computers used in the company in 1979-1980 looked a bit like early versions of Apple Lisa with double 5.25” floppy drives (released about the same time as the IBM one you mentioned).

I might be wrong, but I think we had some ITT computers running Xenix later on as well, alongside MS-DOS. However that was well into the 80’s.

#83 Posted by JoeMonco on Oct 17, 2011 6:03 AM

“You referred to Unix hardware’s physical size[...] However that was well into the 80’s.”

So it wasn’t exactly the “70s” like you previously claimed, was it? And that’s 10 years of difference at least we are talking about here – between time-sharing on a machine the size of a casket and personal computing on an enterprise desktop machine. It’s simply preposterous to confuse the two together in any way.

Seriously, why not tell us your experience in the first Woodstock you had had before you were born in 1970?

“I might be wrong, but I think we had some ITT computers running Xenix later on as well, alongside MS-DOS. “

The way I understand this is that your wheeling-dealing daddy had them, not you. I would honestly wonder what interesting substance you were on if you didn’t even know the purpose of the machines that your were supposed to work with. And MS sold Xenix to SCO and stuck to DOS for a good reason, but I am not sure about it thought. Maybe – I don’t know – it had something to do along the line of Lotus 1-2-3 and other apps that people actually gave a crap about using?

#84 Posted by KimTjik on Oct 17, 2011 7:52 AM

Since you made this personal JoeMonco you still seem to think you’re onto something, and you just cannot let it go. Is it because you publicly announced that I’m a bluff and it’s not comfortable to retreat? Since you’re so interested in my personal life I’ll continue this “biography”.

I thought, and it might still be true, that the “playground” time I referred to was sometime around 1978-79 (I took my bicycle to my father’s work after school). That would be the 70’s, right? Instead on swearing on my grandfather’s grave I give it some doubt and say 1979-80. Oh, forgive me holy JoeMonco if I’m one year wrong and it actually wasn’t until 1980! It doesn’t make it 10 years wrong, but it for sure is a unforgivable mistake!

You’ve accused me of putting to much emphasis on semantics some time ago, didn’t you? So tell me now why you try to tear apart a totally normal expression “we had”, because I’m referring to a family business I’ve been part of for many years, to mean something very specific? Since you ask, I started to work a bit before I even ended elementary school; the school director(?) even gave me some days off to attend lectures at the company.

You’ve persistent, I give you that. I don’t understand the purpose though.

#85 Posted by ReverseControllerSE on Oct 17, 2011 8:44 AM

Well Kim, given that you’ve misplaced technologies by over 10 years, it’s not very surprising people don’t believe you.

While it is possible that you’re telling the truth, the simple fact is that many Zealots out there are prone to lie through their teeth to push their agenda.

If you’re simply confused as to how exactly, or rather when exactly, certain events took place, then fine (however it is hard to understand how one misjudges the time by a whole bloody decade).

You should strive to express yourself more clearly, some of the things you wrote made little sense – you seem to have trouble applying English grammar (i.e. you use proper words, but the sentence follows some other language logic).

#86 Posted by KimTjik on Oct 17, 2011 9:07 AM

Tell me is 1978-79 the 70’s or not?

If yes, then there’s nothing to be upset about. OK, I understand that folks tend to lie, maybe to look better, or as you say push an agenda. Some probably only try to pick a fight for the fun of it.

Would it even be logical to believe I was talking about a family business running accounting applications in the beginning of 70’s. If no, you could figure it out yourself.

No, I’m not a liar and I don’t even have an agenda.

#87 Posted by JoeMonco on Oct 17, 2011 9:12 AM

“Since you made this personal JoeMonco you still seem to think you’re onto something, and you just cannot let it go.”

I think no one would need to be personal about anything if you are not utterly vague and dishonest about what you claim, not wanting to own up to the responsibility of your own words and constantly trying to weasel your way when accusation gets thrown your direction.

“Ah… I mean this actually.”

“No… I mean that actually.”

“No… It’s definitely this this time. Honest!”

I think I have seen the pattern here enough from you to be thoroughly sick of it.

“Is it because you publicly announced that I’m a bluff and it’s not comfortable to retreat? Since you’re so interested in my personal life I’ll continue this 'biography’.”

You are not just a “bluff”. You are intentionally misleading on your every claim about yourself and what you know about a subject. Seriously, just observe the same pattern of weaselling you have repeated in the last 10 comments – it’s just something way too obvious to be ignored.

“I thought, and it might still be true, that the “playground” time I referred to was sometime around 1978-79 (I took my bicycle to my father’s work after school). That would be the 70’s, right?”

With practically none of the knowledge relevant to the seventies. That’s what makes your “70s” claim stink like a dead trout.

And, on top of that, your claim about the eighties is also dubious at best due to your lack of knowledge on the subject matter. Really, am I supposed to consider playing Bubble Bobble on your dad’s computer a serious usage relevant to the history of operating systems?

http://en.wikipedia.org/wiki/Bubble_Bobble

#88 Posted by JoeMonco on Oct 17, 2011 9:12 AM

“You’ve accused me of putting to much emphasis on semantics some time ago, didn’t you?”

I think you are the one who have accused me several times of putting too much “emphasis” on semantics. Probably something to do with your supposedly broken English or whatever. But, who knows – I am obviously too new to this juxtaposition business that you have got going there anyway.

On the other hand, I do however definitely recall myself asking you numerous time to clarify your viewpoints and statements and with each request either completely ignored or rejected outright at each instance. I am sorry, but this spiel of yours is getting really, really old.

“because I’m referring to a family business I’ve been part of for many years, to mean something very specific”

Is it specific enough to offer you an position of authority on the subject of operating systems? After all, that’s what everyone here wants to discuss, I all I care is to get the ball rolling with as little pain that I need to go through as possible. Thus far, all you have been doing is undermining a potentially informative discussion by asking frivolous questions without clarifying as to why you are asking them or what your opinion is on the subject and persistently claiming knowledge of things that you are never actually familiar with all the way from Piestar to here.

I am sorry, but you can’t just come here and expect people to turn tricks for you without offering anything substantial and meaningful in return, and this game of yours is nothing more than a nuisance to anything wanting to read something worthwhile here. You either step up to the plate and give us a clear viewpoint to start a discussion with, or get lost.

#89 Posted by KimTjik on Oct 17, 2011 10:48 AM

You’re honest about your opinion and that I respect. I’ve already stated that you know much much more than me about these things, but I’m a curious person. Do I intentionally try to mislead anyone? No.

You see I left the family business altogether in the middle of the 90’s and for over a decade I travelled the world doing humanitarian work. There’s a big gap and there’s no way I can pick up all what I lost.

Yes, I at times totally ignored to respond to your questions. Sometimes because some of your peers had already written what I meant, and other times because of your choice of approach. I shouldn’t have engaged at all, and should have controlled my curiosity from showing up here as well.

I chose to “get lost” even though I passively will keep an eye on discussions here. However, as I stated before it has a tendency to become cheer leading. Maybe that’s partially the purpose of the site; was it you who said that it was a place to vent frustration about loons?

#90 Posted by ReverseControllerSE on Oct 17, 2011 11:25 AM

“I left the family business altogether in the middle of the 90’s and for over a decade I travelled the world doing humanitarian work.”

Really now… you know, that statement makes everything you said even less plausible.

:/

#91 Posted by KimTjik on Oct 17, 2011 11:27 AM

Why?

#92 Posted by ReverseControllerSE on Oct 17, 2011 12:10 PM

Because it’s a cop out.

Also, it makes some of your statements even less plausible, unless you were describing things that happened while you were gone.

#93 Posted by KimTjik on Oct 17, 2011 12:22 PM

If I understand the “cop out” term correctly, it means something like I would try to escape responsibility. OK, despite it’s truthfulness maybe it was a meaningless addition to the discussion.

I worked some years in the family company to in OS terms Windows 3.1. After that I only temporarily returned home. Sure I had access to and used close to every version of Windows after that, but not as before that.

When I returned to the company for real I made the transition to Windows 2003 R2. Nothing strange about that.

#94 Posted by JoeMonco on Oct 18, 2011 4:13 AM

“I’ve already stated that you know much much more than me about these things, but I’m a curious person. Do I intentionally try to mislead anyone? No.”

With all that blatant weaselling? Suspension of disbelief, anyone?

“Yes, I at times totally ignored to respond to your questions. Sometimes because some of your peers had already written what I meant[...]”

Not seeing that happening. Examples?

Or are you just going to ignore this like every other instance?

“[... A]nd other times because of your choice of approach.”

I am sorry for hurting your widdle feeling. Can we move on to the actual subject in question now, or are you just going to keep stalling for as long as possible?

“I shouldn’t have engaged at all, and should have controlled my curiosity from showing up here as well.”

Honestly, no one gives a crap about your “curiosity” just as much as the image at the center of the About page is not a lollipop.

Take the hint.

“I chose to “get lost” even though I passively will keep an eye on discussions here. However, as I stated before it has a tendency to become cheer leading.”

Or, you can just, as I said, clarify your viewpoint and get a proper discussion going. However, if you are, as you say, know as little about the subject matter as you claim, then I wonder if you are even capable to tell the difference between object opinion and “cheer leading”.

“Maybe that’s partially the purpose of the site; was it you who said that it was a place to vent frustration about loons?”

Say what you want, but until you are bothered enough to come up with a question that is not ambiguous at best, then I honestly don’t think there is anyone on earth that will manage to live up to your expectation.

You have a problem with this site, and I have provided a crystal-clear solution to it (i.e. clarifying your statements, which is not supposed to be something below your dignity – unless you are a total cretin – and something that you can do as many time as you please), but if your choice is to completely ignore this solution and instead keep on moaning about the same thing at every turn, then there is nothing honestly I can help you with.

#95 Posted by JoeMonco on Oct 18, 2011 4:15 AM

*Or, you can just, as I said, clarify your viewpoint and get a proper discussion going. However, if you do, as you say, happen to know as little about the subject matter as you claim, then I wonder if you are even capable to tell the difference between objective opinion and “cheer leading”.

#96 Posted by JoeMonco on Oct 18, 2011 4:53 AM

Hell, I’m feeling generous today, so how about I give our resident wet soap KimTjik an example on how to clarify his statements or question?

“- in some niches, like the extreme TOP500, *nix will continue to dominate. It’s specialized and it has no direct “consumer” implication. I actually expected Microsoft to do better, since they pumped in a lot of money several years ago into their HPC line.”

Well, obviously, Kim has a opinion on HPC, namely, “to do better”. How better though, you might ask? Well, I am curious about that, too, and since I ain’t Kim, I think he is the only person capable of coming up with an answer here.

What do you have to say about that, huh, Kim?

#97 Posted by KimTjik on Oct 23, 2011 5:22 PM

“Well, obviously, Kim has a opinion on HPC, namely, “to do better”. How better though, you might ask? Well, I am curious about that, too, and since I ain’t Kim, I think he is the only person capable of coming up with an answer here.”

1.) Microsoft pushed out quite a bit of information/promotion about their HPC line of products some years ago. Hence I without necessary insight expected to see more Windows systems represented among the TOP500. Simple as that. Call it user/market share or whatever you prefer to label it.

How much better? I’ve no answer on that since as I already wrote that my expectations weren’t based on any insight.

2.) Correct me if I’m wrong, but as far as I can find out Windows HPC is stated to support up to 64 cores (or threads), even though there are also figures telling that some certified configuration might support up to 4 times that figure.

I don’t know whether this really matters much for Microsoft, since this might not be their first priority. You have Linux products though certified for 4096 cores/threads. I cannot find any data proving how well these systems scale, or whether there are practical work loads demonstrating that this matters today. You said you had some idea about it, so maybe you know.

This was just an example about technical differences.

...

I suspect the Windows HPC line to do fairly well bellow the behemoth category. I tried to find some figures, but without luck.

(By the way: my name isn’t “Kim”, there’s another historic reason for the nick)

#98 Posted by DrLoser on Oct 23, 2011 5:42 PM

Thank God for the voice of sanity, at least for today.

And we’re gonna have to keep calling you “Kim” in much the same way as we call KurkosDR “Kurkos.” It might be a bit inaccurate, but it is at least convenient.

#99 Posted by DrLoser on Oct 23, 2011 5:46 PM

Oh, and the obvious reason that Microsoft is not in the Top500 is that they can’t make any money out of it. They tried, but it just ain’t possible.

If you had access to the OS for a PDP-11, you could probably scale that up to several petaflops over ten thousand cores. It’s the same with Linux.

Boil the OS down to its basics, throw the other 95% of irrelevancies away, spend your time making sure that it runs gigantic monolithic modelling systems in parallel, and you’re golden. It’s no secret, really.

But then, it hardly matters to me, or you, or anybody who doesn’t work in one of a select few research labs, does it?

You must be signed in to leave comments.