Piestar

Your posts make me cry with mirth and embarrassment.

“THE DESKTOP IS A SOLVED PROBLEM!”
Oh yes, and all these weired experiments/”improvements” in Gnome just remind me of this entry:
http://elliotth.blogspot.com/2009/05/confessions-of-windows-hater.html
Conclusion: “I will admit that, all in all, I’ve made my desktop look a lot more like Windows than it did. But I’m man enough to admit it’s better for it. And I’m grumpy enough to wish I didn’t have to waste my declining years pissing about with stuff that shouldn’t be getting in my way in the first place.”

I don’t quite understand this post nor other I’ve just read. The headline of your blog is “A pragmatic look at the state of FOSS” but it sound like your definition of “pragmatic” is different than the one I know.

Quote of the day !

“Monty” Widenius is still grumbling about MySQL/SUN/Oracle. Post in a (german) forum, loosely translated:
Why is he still bitching ? He sure didn’t bitch when he took the money. And with the sale, MySQL became TheirSQL.
http://www.heise.de/ix/news/foren/S-Re-Warum-mault-der/forum-181957/msg-18786368/read/

@kklimonda
“but it sound like your definition of “pragmatic” is different than the one I know.”
Which is?

“And why is he trying to maximize vertical space? On a netbook you want to maximize horizontal space (…)”
Oh come on ! Clearly, if you port a desktop layout optimized for big displays to a small display, you are being innovative, and what could be wrong with that ? *rolleyes*

Here’s a nice gem for you:
http://www.h-online.com/open/features/OpenOffice-at-the-crossroads-1023702.html
Synopsis: Real developers don’t like specs, they don’t like QA, they don’t like updating help files just because they changed something, they love to do just whatever they want, oh – and would SUN / Oracle please write our paychecks ? In geek speak:
“The virtue of successful open source projects is supposed to be that they are developer driven – open, democratic, noisy, argumentative, divisive, chaotic and productive.”

I think too that Ubuntu seems converging to Mac Os X look, but I’m unsure if it is done with the explicit intent to copy it or it’s just converging evolution (http://en.wikipedia.org/wiki/Convergent_evolution).
I don’t think that the desktop is a solved problem: maybe it seems so just because now we are so accustomed to some (initially) arbitrary paradigms that we think that we can’t do better.

But play for a while with next KDE 4.5 (and in particular with plasma, nepomuk, akonadi), and you’ll see that a lot of things aren’t obvious any more, and we can still do things interesting with the OS and with the desktop environment! It’s not just a matter of applications.
Applications don’t live alone isolated on the desktop, they can communicate between each other, exchange data, and interact with us and with the system. If the desktop environment let these data flow, you can do more and unexpected things with your desktop environment.

Stil about the apps: I don’t think that we really need to get a new version of an application as soon as it go out in the wild. For the typical user updates are just annoyances. I think that a set of updates every six months is quite reasonable. This slower pace of distribution let people write documentation, discover issues, and fix them for most of us.
Remember also that there’s not just applications, but also libraries, and most of them are shared between apps. Sometimes if you want to update an app you need a new version of a library, but chances are that this new version is incompatible with other applications, and you get problems.

Also you seem to ignore that out there there are also quite a number “rolling release” distro, like debian unstable and arch linux, for example. They give the user the possibility to have always the newest version of everything as soon as is released. So why don’t people use them instead of all that six-month-release distros? I think that it’s because it requires a lot of user work to maintain a rolling release system, and because documentation about issues become old and inapplicable in a very short time, so it’ difficult for people to get and give help.

But still, if you want you just can compile yourself an application you care for. I usually do it for digikam: it’s released almost every 3 or 4 weeks, and since I’ve already an environment ready for compilation, I just need to type the magical “cmake && make && sudo make install” to get my new version. Not difficult.

But if it’s too difficult, if you are on Ubuntu you can just use one of the thousands of ppa that give you updated versions of the apps that matter for you. You seem to ignore this. For example, the KDE team gives us a release every month, and the kubuntu team puts this release on their ppa almost a couple of days after the release, so you get always updated applications.
So, if you don’t like the default, more conservative, approach, you can go (almost) bleeding edge using your preferred ppas.

So please, don’t be so superficial when you judge the gnu/linux world!

I also agree with the sentiment that the desktop was essentially “solved” with Windows 95 and disagree with the above sentiment that its paradigm was arbitrary. Microsoft spent considerable amounts of money writing the interface from the ground up utilizing tons of interface designers and testers. In its day Windows 95 was immediately intuitive to both new users and those who have already been indoctrinated, and it was hailed as the most significant UI breakthrough since the original Mac OS, which at the time had been stagnant for nearly a decade.

But so called “traditional copyright” is the problem. The fact that billions of people relinquish their rights to faceless corporations amassing power and wealth for one percent of the world is the problem. The fact that amazon is erasing books that can help people protect themselves against being enslaved is the problem.

I love the “security” and “stability” arguments because they’re always thrown out as if one is expected to take such heavy claims at face value. When it comes down to it, there’s nothing behind them because either the research hasn’t actually been done or that reputable sources dispel the claim.

When most advocates talk about “stability” they’re really referring to “uptime”, which is kind of a crock because they know xorg has grown increasingly unstable over the years and that when it goes down, it takes all the work with it. But, hey, I still get a 1972-esque DEC terminal blinking cursor! Now that’s “stable”! But, hell, even if you engage the “uptime” argument head on, there are plenty of Windows servers up against the Linux ones, although extreme uptime is a questionable goal in any case due to missing out on security and bug fixes and facing inevitable memory fragmentation. Contrary to popular belief, random access isn’t as fast as sequential access in RAM and fragmentation does present allocation problems.

Then there’s the “secure by design…cuz it’s UNIX” argument, which doesn’t even stand up to casual scrutiny. UNIX’s design goal was a time sharing system. Full stop. Its “security” considerations ended with the users being mutually exclusive. It was introduced in an era where high tech criminals were unfathomable. Claiming that Linux strictly adheres to this philosophy (which it doesn’t because it breaks out in arbitrary ways) doesn’t do it any service because the concept is laughable that the world’s most secure system was invented five decades ago when “operators”, as they were then called, were predominantly using punch cards and mechanical teletype terminals.

Then there’s all the Securina advisories, but when that one comes up the “Windows is just an OS” and “Linux is all in one” argument conveniently switches. Best part is that even if one were to throw in all of Microsoft’s stuff it still doesn’t work out.

“Beyond rotating themes from Microsoft’s stock, I never bothered to “tweak” Vista much and have even less inclination with 7. There was much todo about Aero “wasting” memory (though, curiously, nothing with regard to Mac OS), but these days 512MB video RAM is considered small. I don’t know how much Aero eats up, but I very much doubt it’s that much. Then I find out that video-hungry applications (e.g. games) can shut off Aero and other effects on the fly! That’s just awesome. It’s like having your cake and eating it, too, whatever that means.”

I seem to recall that 7 optimized their pipeline to halve the memory usage. I seem to recall on of the Engineering Win7 blog posts talking about how in Vista you keep a buffer in a main memory and on the graphics card while 7 can keep it only on the graphics card.

@AnonGuy

Totally agree, although they are a bit of a pain on windows if you leave it in default mode (install automatically), easily fixed.

One find I find seriously terrifying is the “release early, release often” motto for many distros. I’m a Windows user primarily but I enjoy using Linux when I’m coding software etc so I keep one disk with Arch Linux on it and another drive just for testing new distro-versions.

Earlier this year I installed Ubuntu 10.04 about two or three days after it’s release and was genuinely surprised when Update Manager popped up on my first boot. It wouldn’t be too bad if it was like two updates, one of which being the release notes. But it wanted to update about forty apps, we’re talking 350MB data.

Seriously? Is their QA process really that shite? Considering how Ubuntu currently is the premier average user distro and backed up by the bottomless horn of money called Shuttleworth one thinks they’d check things like that before actually releasing. They do have betas and RCs but do they actually do work themselves?

No matter how awesome a community is there’s no way it can replace proper in-house testing. Where many betatesters download the early versions to check them out and perhaps report one or two bugs, real companies employ people who get paid to sit eight hours a day trying their hardest to break the software. Not to mention things like automated testing etc.

Doesn’t people take pride in their work anymore?

@Delano

Yeah, I suspected as much. I just refused to admit it to myself. I’m a geek, I take pride in doing a good job when it comes to software and making the best solutions. But I guess that’s the difference between one developer and a clusterf*ck, many eyes lead to many “ok” instead of one great solution.

It’s really pathetic and if they bring up the “but we do it for free” I suggest taking a look at the various BSDs and Haiku. There you also have people working for free but the difference is QA and strict coding rules. The results speak for themselves, it may not be as flashy but atleast it works and works well.

Amen.

Linux in itself, without the GNU userland or tons of crap piled on top of it is a really nice piece of software. It’s not flawless by any stretch of the imagination but the basics of it is sound. It’s portable, supports lots of different technologies and generally behaves well. It’s when you start piling crap like X.org, a DE and the tons of different extra daemons on it things start falling apart.

Take for example the sound system. Back in the day when it was using OSS it worked well then ALSA came along for ideological reasons but it worked well enough too. But then people started creating additional frameworks like PulseAudio, Phonon etc and things going downhill.

A lot of so called progress is being put into motion by “cultists” and often because of ideological reasons. Apparently it’s OK to break working software if it’s deemed immoral by their Great Leader.

I’m looking forward to seeing more progress in the evil, immoral software project clang. It’s one of the more exciting things that’s happened lately in the compiler world, it’s technically superior to GCC with the added bonus of being properly licensed and not directed by a fanatical hippie.

Also, I totally agree with David on Clang being the shit. Watch the session on LLVM from WWDC (free to watch you just need a free Apple Developer account. It doesn’t give you the good stuff line App Store access but you get access to all the dogs and videos for free), freaking amazing. The Xcode 4 preview is just jaw dropping.

@Richard Querin:
“I think the whole distro system *could* be a real advantage ”
How so ? Do you think that one distro uses / should use one kernel version that “handles like a sports car” and another a kernel optimized for “12 people and plywood” ? There’s a reason why other use the same kernel whereever they can – it makes so many things so much easier, and I don’t think the performance pennalties are anywhere like with cars. (In windows, it’s mostly one switch: “Give more priority for foreground / background apps.”)

Imagine the Linux devs getting their acts together and actually cooperating and standardising: ONE installation mechanism, ONE desktop, ONE audio stack etc. ONCE a year (for the time being, the 1st of April would be appropriate new versions of mayor components – beta-tested in combination – would be released, then for a year it would be only bugfixes and security patches that DON’T break things. Everybody would have the same working and tested baseline. Where would that leave the need for different distros ?

OK, you could bundle different apps. But since with a working, identical baseline, every app developer could just create an install file and put it on SF (for example) where the installer on every machine could pick it up, “bundling” would be less of an issue. E.g., the OS setup programm could include a screen asking “Which bundle would you like to start with ? Standard – Office – Gaming – Education – Developer – Geek – Music recording – …”

Would you really prefer today’s mess to this solution ? (Of course it ain’t gonna happen – it would be the sensible thing to do, you know, not the “free” thing …)

@Carsten

Bundling and organizing is definitely a way to go but there are some issues that need addressing first. Biggest one would be about who decides what goes in. Looking at the software available you’ll find that there’s very little software that is actually complete. Not just complete as in “not beta” but also in complete featuresets and so on. You can have a bundle for “Video Editing” but that implies that it actually can edit video which is sadly lacking.

When you go buy Final Cut or Premiere there are some features you expect, like importing, timeline editing, filters and decent compatibility with formats. With today’s OSS climate you won’t find one single app that does all of this but you’ll have to install several applications to create a workflow.

So lets say you bundle a decent working video editor, you’ll still need additional software to get work done and who decides what that work entails? Can someone make the decision that importing h.264 video isn’t needed and therefore exclude some pieces of software?

Withóut a paying customer base it gets very hard to gauge what people actually want so other ways are needed to figure that out. With Premiere for example you can be damn sure that if a needed feature is lacking people will find another solution whereas with todays OSS landscape it’s not that clear cut.

In conclusion:
Your vision seems unlikely without centralised control and controlling today’s Linux users is like herding cats. They don’t like authority and “the man”. Adding the fact that many Linux users have an insane aversion of paying for stuff doesn’t help either.

@Carsten

It wasn’t meant as a response, counter-point or anything. Just my thoughts.

Somewhat of a continuation or something, elaborating on your reasoning.

@David: Thanks.

(IMHO, first we need the baseline – which won’t come -, then we need the apps – which won’t come -, and then we’d have the luxury of talking about bundling.)

Dr Loser,

Your response meshes perfectly with the Joel on Software article recently linked on jerkface comments:

http://www.joelonsoftware.com/articles/Stupidity.html

Basically, you’re turning a blind eye to the deadline slipping and general lack of practicality in favor of “cleanliness” and a geek’s sense of perfectionism. Even if ISO/OSI were “better” eventually, it took so long that everyone lost interest and overloaded TCP/IP to fit the workload. Is it a mess? Absolutely, but if ISO really wanted to avoid that, they wouldn’t have spent two decades “designing” it before ultimately throwing their hands in the air.

The comment about seven layers was strictly about how decision makers tend to get their own piece. That should have been quite obvious given the context. However, while I did not pass judgment on the technical aspects, I do find it unlikely that seven is the optimal number given the political circumstances.

Another one that’s somewhat obscure is a video format you’ve probably never heard of called NUT. When Matroska (MKV) was released, the ffmpeg/mplayer guys shit all over it calling it a hack, claiming C++ was inappropriate in any scenario, and saying that it’d bring video players to their knees (apparently ignoring the effect of high def H.264). So they went to work on an “efficient” container format, a programmer’s dream. I stopped checking in on it after 2-3 years but basically they spent their time arguing over whether headers should be two bytes or three and bitching about perceived inefficiencies measurable in kilobytes over a two hour movie. Meanwhile most of the stuff that made MKV appealing was labeled as “you don’t really need that bloat” so that even if NUT were implemented it’d be little more than a modernized AVI.

End result? MKV is a stone’s throw away from being the de facto standard for high def and multi-lingual content while no one has even heard of NUT. I certainly have never seen a NUT file in the wild.

@Tux Sux:

OK, on re-reading, you’re clearly suggesting that “the basic four are enough” rather than comparing numbers. Absolutely, and I apologise for (inadvertently) misrepresenting you. I’ll accept that, in general, the ISO committees got carried away above the transport layer. (Though the session layer did still work!)

I’d still suggest two things: the basic four worked perfectly well for ISO/OSI, and they worked pretty much out of the gate in 1980 (the Red Book in 1984 added a lot of nice features, but didn’t substantially change things — and where it did, the negotiation phase of the setup handled it transparently). Hard to see a committee issue at that level, then.

Additionally, it’s a little unfair to compare a protocol stack designed for telcos and referencing work from 1978 and before against a protocol stack designed for DARPA and mostly for use on LANs (thus the streaming nature of TCP/IP).

ISO/OSI performed its function extremely well between 1980 and … well, I last used it circa 2000. TCP/IP is fine now, but you’d have had to have thirty years worth of foresight in terms of hardware and network speed even to imagine something like VoIP working *at all*.

Not sure what your comment about geeky perfectionism is meant to mean. I’m saying that the (implicit) layers above TCP or UDP are a total and revolting and completely unco-operative mess, and you’re saying that the (explicit) layers above ISO/OSI transport are a total and revolting and committee-driven mess.

Well, so what? Neither one of us appears to be a perfectionist. Probably we’re both geeks, but it’s lice’n'flea territory at this point. Basically I agree with you about ISO/OSI upper layers (particularly 6 and 7, which were ridiculous). You appear to agree with me (“Is it a mess? Yes.”).

But that’s hardly the point. The marketplace decided, as we both pointed out, and that’s the way to go, as we both agreed; but it doesn’t necessarily mean that the committee-driven decisions of ISO/OSI were any better or worse than the committee-driven decisions of VOiP, SIP, etc etc etc.

Markets tend to be driven by costs versus functionality. Technical excellence is neither here nor there. By 2000, you could buy an ethernet card for about $20 and get the stack thrown in for free. For an X.25 card? Probably nearer $200. Number of socket programmers out there? Tens of thousands at the minimum. Number of X.25 programmers? Hundreds.

I’d hardly knock IP-based protocols. They’ve been my bread and butter since 1990. However, as a historian (sub-class, geeky), I kind of object to sweeping observations about why one technology beat another because of perceived organisational failures — at least, without reasonable proof. I’m fairly certain that timing, strong backing from the DoD, and consequent cost factors were rather more important.

And if I’m even partially correct, then it’s not really as good a metaphor as the Edsel, is it?

@David:

Ignoring pointy-heads on committees, the *engineers* at Honeywell (etc) who came up with the ISO/OSI model were not substantially different from the *engineers* who came up with the xCP/IP equivalent. I’d assume they swapped notes, papers, and ideas.

Elegance be damned. Perfection be damned. In this case we’re talking about standards. The ISO/OSI one worked in 1980, as advertised, and on 1980 hardware. The TCP/IP one has always worked, on similar terms. It doesn’t suck. It’s a perfectly fine streaming protocol. It’s a fucking pain in the arse to work with for anything else, which is what I’d expect. “At least it works” is a pretty crappy reason for burning hundreds of millions of man-hours fiddling around with Yet Another Socket Abstraction and shoe-horning Yet Another Session Layer Abstraction.

It’s not a question of whether it *works.* Lots of things *work* — even Token Ring *works.* It’s a question of whether it’s *saleable*. Show a CTO a network interface at a tenth the price — instant sale. The amount of crap required to back that $20 up is amortized into the future; rather like Linux, really.

NOTE: I am NOT saying that ISO/OSI is better value in this respect. I am merely saying that market decisions are made across many other dimensions than “at least it works.”

@Dr Loser

Just to clarify, I didn’t mean that TCP/IP sucks. I was being rhetorical in that even if it sucks it doesn’t matter now because it got done. I rank “it works now” higher than “it will work great once it’s finished”. It’s easier to build upon something that is out in the wild.

@David

Your rhetoric needs a little work, then. It’s difficult to read that entire paragraph any other way … never conclude with a statement that means the reverse of what you want it to mean.

I don’t think any of us would disagree with the high ranking of “it works now.” It’s a non-argument in context. Reiterating the history once more:

X.25 (ISO/OSI layers 1-3) worked from 1980 onwards. TCP/IP worked from about that point (although, initially, restricted to Unix platforms. Which was a big deal in those days).

The layers above, in each case? In a sense, not very important. In another sense, shit, in both cases.

If you were working with network protocols, as I was in the late ’80s, that didn’t matter anyway. Either your end-to-end protocol was ISO/OSI, or it was TCP/IP. You built stuff on top, and ad hoc. The fact that there were “standards” (which you ignored) was neither here nor there.

TCP/IP won out because of many advantages, few of which were technical (although few of which were entirely non-technical).

Off the top of my head, and in order:

* Support from the DoD. Great as a cost-reducer, amongst other things.
* Support over NetWare. This was actually a big thing in the ’80s and early ’90s (which is ironic, since Tux Sux has compared me to the “geeky perfectionists” in In Search Of Stupidity. Yes, I’ve read it. Not just the introduction). You could run practically any LAN protocol over NetWare (even AppleTalk, I believe, and I still miss the layers above AppleTalk 1-4). You couldn’t really run X.25, and why would you want to?
* Third-party hardware support. X.25 is essentially end-to-end, as you would expect from telcos. IP benefits from bridges and routers, and if you can get somebody to chuck those out the factory door at low prices, the network effect (the real hardware one, not this soft six degrees of Bacon thing) is pretty powerful.

My conclusion is quite simply that the merits and demerits of one solution or the other have little or nothing to do with committees. (On a brief note, however, you should be aware that the ITU-T is fundamentally a European organisation, which didn’t help much.) They have little or nothing to do with “it just works,” either, because when “it just didn’t quite work,” you built “it just works” on top of either one.

It’s all down to market realities and pricing and the ability to source “expertise.”

Your point that “It’s easier to build upon something that is out in the wild” is, I’m afraid, moot in the context of the ’80s. More to the point, it’s more convenient to use an end-to-end protocol if other people are using it 100x more, and that *already* has things built upon it “out in the wild.” (Incidentally, this mostly doomed SNA, even *between* IBM machines. Serves the stupid thing right.)

@Tux: Comparing ISO/OSI with the HURD, btw, is a step too far into unreality. Even if I accepted that ISO/OSI failed because of naughty committeeitis, I would like to point out that the HURD failed at the totally opposite end of the scale. No matter how often Richard talks to his own toenails, that doesn’t really constitute a committee in anything other than the “This Little Piggy” sense.

Cool, but as I already have tons of video in MKV and hate remuxing I’ll stick with MKV.

Yeah, and what do you do with it after it’s in NUT form? Those dozens of kilobytes saved will look great when unplayable files are sitting around on your hard drive.

@David:

Me, serious? You’ve got to be kidding.

@Tux Sux:

Sorry, that was David Weizades. I was desperately trying not to attach the wrong name, and whaddya know, I did.

@David Weizades:

It’s not a formal debate, but it’s not a Ubuntu-style forum either. (Despite the rather messy contretemps between John Stamos and JoeMonco on the previous post.) Sorry, but if you’re going to claim something is rhetorical when it looks like nothing of the sort, I’m going to point that out.

I didn’t comment on your quote:

“Linux in itself, without the GNU userland or tons of crap piled on top of it is a really nice piece of software. It’s not flawless by any stretch of the imagination but the basics of it is sound. It’s portable, supports lots of different technologies and generally behaves well. It’s when you start piling crap like X.org, a DE and the tons of different extra daemons on it things start falling apart”

… because all of that has been suggested by Linux devotees, incessantly, for what now seems like a lifetime. Sometimes they arrive as fully-formed Adam Kings. Sometimes they claim to be Windows users, but … (and the *but* is Arch? I think I’d rather spend my spare time fixing 1920s motorbikes).

I’m not sure about light-hearted, but I can manage light-headed:

* Linux without GNU is, basically, nothing. If it helps, I think that is also true of *BSD. *BSD is a lot less vocal about these things, however.
* There is no such thing as a GNU userland. Gnu provides a tool-chain, a set of programming tools (mostly CLI), and some interesting projects here and there. Some of them are worth pursuing; some not.
* It is impossible to set up a Linux system without the “tons of crap piled on top of it.” *All* Linux sysems come complete with “tons of crap piled on top of them.” Even if this *were* possible, it would not be “a really nice piece of software.” It would be an operating system without a purpose.
* The basics of it are sound in the sense that the basics of a steam-roller are sound. Most people have evolved beyond steam-rollers.
* It isn’t particularly portable. The kernel does OK, but what’s the point of a portable kernel? Have you ever tried to port X? Is there a point to a ported Linux without a video stack? Does Linux have an alternative to X?
* It doesn’t support “tons of different technologies.” It mimics some, botches others, and plays around with concepts which it changes arbitrarily every six months or so. It is *ideologically opposed* to technologies (if you can call them that) like stable ABIs and stable APIs. It is *ideologically opposed* to technologies where the source doesn’t come free, _even if the actual object library comes free_.
* It’s the only OS in the whole known universe where, after ten long damn years, people are still bickering about the DE (desktop environment). I believe that was the point of Kerberos’ post here.
* If I had to pick a system that leans excessively on daemons, it would be Solaris. Daemons are *not* part of the Linux disease. I can’t think of a single Linux daemon that is more obviously broken than the rest of the system. YMMV.

You would presumably be the same David Weizades as in http://www.mail-archive.com/xorg-prop-drivers-testers@lists.launchpad.net/msg00037.html and https://lists.launchpad.net/xorg-prop-drivers-testers/msg00037.html and various other links. Launchpad and Ubuntu and maybe if I could be bothered I could track down Arch. Doesn’t sound much like a “Windows user, primarily.” Sounds more like a troll who has been pulled over by Kerberos’ demolition of Ayatana.

But then, not only do I have my panties in a twist, I have a suspicious mind.

@Tux Sux:

Sincere apologies for confusing you with this waste of space.

There’s the prevalent attitude that Linux is “already there” and just needs some tweaking before it’ll be “noob-friendly” enough for everyone to use. That’s why Shuttleworth and co. are obsessing about themes, wallpapers and buttons.

The crazy part is how they are convinced that Linux is ready for ISVs and in their minds porting from Windows just involves a few additional steps compared to OSX. To them Linux just needs more marketshare and then ISVs will flock over.

I’ve gone over this issue 10k times with Linux advocates to no avail. In their minds Linux is perfect and because some indy 2d game developer brought his game to Ubuntu means it is only a matter of time because EA starts porting their full library to Linux.

There’s absolutely zero interest in making it easier to port from Windows to Linux. The Linux loons behind distros in their hearts believe that everything should be open source so helping proprietary companies port their software is not even a consideration. Most are playing the long game against proprietary software, expecting that the GPL programmer’s army will eventually provide enough alternatives and turn the tide.

It’s a shame really because there are Linux users that would pay for software and ISVs that would like to sell it to them. However distros design around open source applications and basically tell ISVs to go fuck themselves. You would think with the success of the app store that Linux advocates would re-think their twisted moral system that Stallman created for them but sadly not.

Cool, but as I already have tons of video in MKV and hate remuxing I’ll stick with MKV.

This remark reminds me of all those audio formats that were created to dethrone MP3. They compressed better, had more features, had 24 or 32 bit audio rates, etc. Yet they all died, or, if they’re still around, play second fiddle to MP3, the format they were supposed to usurp.

I’ve never even heard of MKV, so good luck finding widespread support for that while iTunes movie rentals eat their lunch.

A parallel of the Linux rebuttal to usability: “Anyone who doesn’t want to hand crank their car in the pouring rain is lazy and stupid”.

Anyone who doesn’t want to pedal their car like the Flintstones is lazy and stupid.

@JoeMonco

OK, one fudge and banoffee cheesecake for you.

Actually, I felt that John had one or two decent points — he just didn’t express them very well. At least he was defending the slightly weird (but consistent) edges of the Mac desktop.

I still maintain that a true troll is someone who voluntarily comes on to a site that is avowedly critical (but in Kerberos’ case, politely) of Linux, and then claims something like “I’m mostly a Windows user, but…” And then talks FUD.

@DrLoser

Excuse me while I feign frustration and refer to anecdotes.

- Users want shit to work with a minimum of hassle. Most Win and MacOS X users don’t want endless fiddling with their OS to get their job done – they want things to work right away. Noobs as well as professionals.

- Around the year 2000, and in the beginnings of Windows XP (particularly before SP2), Linux users always had two more pro-arguments going for them: stability and security. It really was more stable and secure, because in Windows at that time, both things were an afterthought. But Microsoft got both of that fixed very fast and very well over the coming years.
Some people also mention speed – but Linux GUI usually feels slower, more laggy and unresponsive. Benchmarks of games etc. usually ended in a tie. Linux console-only is of course fast, but irrelevant on the desktop.
Now *all* Linux has left is a debatable “ideological advantage” of the good David vs. the bad corporate Goliath (and there are numerous arguments about why this is of almost no relevance), or maybe some obscure applications you got used to on Linux that don’t exist or don’t work that well on other platforms, which no one else cares about.

- Aesthetics/design. Linux GUIs always have looked, always look and always will look amateurish, quirky and just plain “bad” in comparison. Design differences between single applications or GUI toolkits make it even worse. It always looks plain bad. Even if it’s just tiny things like stupid font sizes/spacing, stupid icon placement, badly readable/recognizeable icons, bad font anti-aliasing, bad fonts in general, or one app having a completely different feel than another. It all adds up to awkward usability and looks.
Sure, the OS should first of all work and not look like a great piece of art, but it should be aesthetically pleasing when you look at it all day. Design is also important because it most often also has an impact on usability, which is the most important reason that you need to get the design right.
Linux people *never* get GUIs right. Ubuntu’s attempts might be noble in intent, but always crap in comparison to the pleasant, unobstrusive, readable, consistent, well-structured and well-thought-out Win or OSX GUIs. Whenever I see a new screenshot of a new Ubuntu GUI “project” the first word that comes to my mind is: “fail” – and that’s even before checking all the functions/features. It’s the first impression. One that always counts. What has Desktop Linux to counter this? Nothing of relevance.

In any case, one sentence to sum it all up: The OS is there to a) support all hardware optimally (automatic and completely hasslefree for the user), b) launch applications which is the only thing users care about (one uses applications, not the OS) and c) get as little in the way of the user as possible while doing that.
Linux always gets in your way and requires constant attention, fixing, tweaking or working-around, and unless you’re still in school or university, you won’t have the time or will to constantly mess with Linux anymore. Do not waste your free time by fiddling with an OS that constantly gets in your way. That alone already makes Win or OSX worth their price.

It’s really that easy to see why people continue to pick Windows or OSX over Linux, and to see why many former Linux users leave ship called Linux (which never sinks, but is still broken beyond repair).

More like a severe case of I L i k e M y T e x t A W i d e A n d B l u r r y M e s s ™

Dammit

@Declination said: 2010.07.08 23:39

Yeah that’s exactly what I mean. Still too wide. It might be a slight improvement over the old default font but it’s still not there yet (and probably never will be).
The font in the window title bar (is it the same?) has inconsistent and amateurish spacing between single characters.
Even more dramatic though is the terrible icon bar, with bad icons that look like they were thrown together from various sources (as usual) and completely ridiculous spacing and size. There must be either a stupid technical reason within GTK (which they should then fix) for why this never gets fixed. Or these guys are complete loons with no sense of aesthetics whatsoever. Even if you should like how it looks, it’s still a waste of screen space!
And of course the elements of the status bar at the bottom look misplaced, the tabs are HUGE (really ridiculous) (and why is there even a tab by default when just one document is open?), the close icon on the tab is ridiculous. And the window controls on the left side makes no sense in Ubuntu because it’s easy to hit the close button accidentally. But I think this blog already mentioned this a while ago. These guys have *zero* clue, and they make the default UI. Great. A recipe for success.
In almost every single Linux GUI screenshot there’s something that looks plain bad.

If Linux could be fixed, someone would have done it already. After 20 years, whoever can not see that is too much of a zealot to care about things like usability anyway.

Linux is a world of fail. Its value lies in the fact that it teaches every conceivable lesson on how not to develop a desktop operating system.

Hem hem, FiveStagesOfLinux(TM)

They don’t really love CLI. They only say that because they wouldn’t know how to create a decent GUI even if their lives depended on it. It’s a trick they use quite frequently. Another fine example is the lack of a stable, standard platform, which they call abundance of choice and deem it a good thing.

I love how all the politically correct freetards are going batshit over some guy hitting on a girl. I mean, if some spanishese or negro criminal had done that he would be “hip” and a “playa.” But since it wasn’t one of them, he’s a rapest.
http://freeishsoftware.org/index.php/the-news/115-alert-sexual-assault-at-southeast-linuxfest

Looks like EugenicsGuy finally has a name. Welcome aboard Michael, please regale us with your poorly-spelled tales of how other races are keeping you down somehow, despite you considering them inferior to yourself.

A “player” wouldn’t get much play with that goofball’s moves; trying to force his hand into one girl’s crotch and trying to force kiss another. Really smooth.

@Michael:

None of us are assuming anything. We’re just reading your posts and drawing the obvious conclusion that you are a bigoted racist troll.

Why do I say troll? Well, your avowed subject matter is nothing whatsoever to do with the PieStar site.

Why do I say racist? Because your set of criminals includes “spanishese and negro,” but conveniently excludes, say, the Aryan nation. Or other white criminals. Or slanty-eyed, oriental, untrustworthy criminals. Strangely, it even excludes those hideous Yids … you can probably work Michael Douglas and Catherine Zeta-Jones into your argument if you want to.

Hitting on girls? Horrors! Only the master race is allowed to do this.

Normally I would recommend going back to Ubuntu Forums. In your case, I would suggest going back to the boonies and stocking up for the Rapture.

Now. Have you downloaded Ubuntu 10.10 Alpha 3 yet? I hear it’s nigger-free!

I’m not saying that the races are genetically inferior or any of that garbage. In fact I applaud them for harnessing white guilt to allow them to get away with anything

So, is this guy now considered a benevolent dictator for life or what?

@Thomas B.
Songbird? Pidgin? Deluge?
You’re almost there Thomas, almost there! Just a few more kinks to iron out .

Ever tried digsby? It’s very nice, kinda like Adium for windows.

I actually tried Digsby, I didn’t really like it. Pidgin suits me just fine.

On the topic of Miranda, something the original ICQ did, which Miranda inherited, that every other instant messenger missed, was a message based queue system. Every other IM system utilized pop-up alerts but left prioritization to the user. Tabs were eventually introduced, but this became an even bigger burden. When ICQ was popular, I found I could juggle about 6 simultaneous conversations or 7 with transient participants but with AIM, MSN, etc. I’d only accomplish 3-4 before people would start dropping off due to disinterest because of my slow response time.

ICQ was also extremely economical on screen real estate. Unlike AIM, MSN, etc. ICQ didn’t basically require you to keep a window open in fear of closing it at the wrong time. Messages could be freely dismissed, leaving nothing but a systray icon as evidence of the program’s operations. While the rest were designed for 100% focused, hardcore chatting, ICQ was far more flexible in facilitating those conversations where only a handful of things are said over the course of several hours.

These days I guess Facebook and the like fill the void, but I always felt something was lost with ICQ’s demise. Its base premise was solid, but the client was atrociously designed and amateurish looking (and sounding–I’m sure someone here remembers the dreaded default high-pitched “UH OH!!” sound effect that accompanied incoming messages). The protocol itself was a joke. Back in the day, Linux actually had an exclusive in LICQ, which was a clone designed to exploit the flaws in the base system. Once could literally masquerade as anybody without realistic chances of being caught. It was trivially easy to initiate arguments between two parties. The whole thing was mismanaged, all the way down to the joke of a website mirabilis.com, which was probably the original Wikipedia-esque link farm portal thingie.

I use Pidgin for AIM, MSN, Skype, and IRC (I’d probably use Xchat or MIRC, but I’m not big into IRC and am only really active in one channel). I find that AIM, Skype, and MSN all have cluttered interfaces. Pidgin keeps it simple and works just fine.

@TMR
“I love the “have you tried X” when it comes to IM clients. How about installing the one, maybe two, IM clients it takes to cover your friends”

Why would I install 2 when I can keep it simple and use just 1?
MSN is an ad infested crap, prone to crashing and keeps getting worse with each new version. Gtalk is just plain shit.
Trying new things is good when what you have sucks balls :S
(I miss the drawing thingy though)

Delano, you forgot that they were successfully sued by AOL over the name. Not only was it a lousy choice, it was legally prohibited! The irony is that they actually chose to stick with “Gaim” even though they also had “Everybuddy”, which is actually catchy and relevant.

As for TMR’s comment about the value of multi-protocol, I found my contacts basically migrated en masse over the years: ICQ -> AIM -> MSN -> Facebook. Sprinkle a little Yahoo for flavor. While I used stuff like Gaim, Trillian, and Pidgin until somewhat recently, I probably could have just picked a given service over the years without losing anything. The worst it got was when ICQ was actually competitive with AIM, but the AOL buyout ended that in a hurry.

As for TMR’s comment about the value of multi-protocol, I found my contacts basically migrated en masse over the years: ICQ -> AIM -> MSN -> Facebook.

That was basically my point. I don’t have contacts on more than one protocol really, unless you count Facebook as a 2nd protocol. Everyone I knew went from ICQ to MSN. Anyone not on MSN is on Facebook at this point and I hardly care that I have two clients (Live Messenger and a browser) open at this point. Hell, often have two different browsers open at any one time.

@Carsten

Interesting link. Even some of the comments are sane … although the idiot who thinks that a “show-stopper” is a good thing should go back and consult a dictionary.

Of course it’s a honeypot. Proper usability studies do not start with puffery like

‘Overall, first impressions are good. Typical remarks include:
“It is bold and different.”
“Ubuntu is fresh and accessible ….”
“This is good. People are getting tired of Windows.”’

(I’d seriously expect the comment “WTF?” to outnumber any of these, handily.)

Proper usability studies also go to great pains to avoid preselection bias, preferably documenting how they do so. This one looks like they’ve just walked down the corridor in the Isle of Man (on the assumption that nobody in Canonical corporate HQ actually uses the stupid thing).

Well, it sure wasn’t a comprehensive usability study. But if they just set out to find the most problematical things for new users, then they sure succeded – and I’d say that’s sensible enough. So they identified issues worth fixing, the comments mostly agree with very little stupidity – so what’s next ?
Point 1: Should be able to fix this in Ubuntu by simply setting the default format for OOo.
Points 2 and 3: Probably involves every other application – good luck
Point 4: Can’t they just install Flash as default ? What’s the problem here ?
Point 5: Once again in Ubuntu’s realm – and improving the descriptions of 30.000 packages shouldn’t be that hard, right ?
Point 6: I had thought that CUPS had been fixed bý Apple, but forgot that of course the GUI is different in Linux.

BTW: Quintin took a look at the latest Mandriva, his first (!) impression: “Mandriva 2010 Spring is awesome. The Gnome edition is as polished and complete as you can get.” Sigh. Compared to that, the CD “study” is still gold.

This is good. People are getting tired of Windows.

It’sAlways1998(TM). Those using computers over a 15+ year span have seen Windows improve dramatically while newer users basically associate Windows as “the computer” in a neutral fashion. What’s left are teenagers who lack experience and perspective and write dreck like the above.

Personally, I think people are getting more sick of Mac than Windows. In terms of improvements, Mac OS took a real nosedive ever sense Apple pulled engineers (permanently?) to work on the iPhone, which delayed 10.5 by ~6 months. They’re not even supporting flashy new technology anymore, which used to be their forte. Back in the 90s and early 2000s it was always “Mac is the first to support X” where X was USB, Firewire, 64 bit, etc. Now they’re dragging their feet on SSD support, which is nearing third generation, and is about the only “exciting” hardware technology these days, unless you count USB 3 or Blu-Ray, which they also don’t support.

“The funny thing is that it is all because of mac zealots and Apple itself, who convinced them to switch, using the same old false argument : “It is so perfect! It is the eighth wonder of the world!”.”

Mac Zealot – “Macs just work”
Me – “macfixit.com”

My pet hates are the “Macs have viruses and don’t crash.” mantras.

I await the virus that in on one fell stroke, wipes out the collective Mac smugness. It’s coming; Mac market share must be getting to the point where it’s worth it for the blackhats – go for the most market share, most of whom are prepared, or the rising marketshare with users who think they’re immune and take no precautions? It’s going to get to a point where the big slice of the small pie is bigger than a small slice of the big pie.

As for not crashing, Apple only did that tasteful screen telling you to restart your Mac because it sensed that you were tired and needed a rest. Because you were tired, the Mac surmised that what you were doing wasn’t up to your own high standards, so didn’t bother saving it.

I bet OSX users make up a large percentage of the people getting phished simply because they’re some of the most connected and potentially least aware.

This is a reasonable hypothesis. Windows users often say things like, “I never open emails/attachments that look suspicious” whereas Mac users say, “I don’t need to worry about opening suspicious emails and websites because my choice in technology makes me impervious to trickery.” In these cases, Mac users are automatically more susceptible simply because, as a group, their eyeballs tend to be on scamming material more often.

@Roboticus
“The fit-to-content/last custom size button makes me want to kick Steve Jobs in the balls.”
The green one? Me too, here’s a gift:
http://www.macupdate.com/info.php/id/30591/right-zoom

The redmund mafia strikes again with strongarm and fear.
http://www.atomicgamer.com/news/2855/lara-croft-discovers-microsoft-summer-of-exclusivity
*hatetards stick heads deep in sand*

Wait. I thought Adam was running an IE9 preview before? The trolls are so hard to keep track of.

Or he’s just a moron who doesn’t know how to properly spoof his user agent.

I’ll gladly stump up for this once-in-a-lifetime opportunity … but only if I get a full refund when I send them back to their maker.

Ah, the Ronco Pocket Troll(TM) — it has a nice ring to it. Through the nose, probably.

In all fairness to gzip and friends, the container format may not lend itself well towards parallelization. Similarly, unless you have a quad core, if you are only getting 25% utilization, then the task is i/o starved and throwing more threads at it won’t actually help.

Well, maybe it comes from the absolute fear of duplication that paradoxically pervades Linux. Its great to have untold swathes of distributions and programs that do basically the same thing but have more than 1 copy of a shared library that performs a specific task and its doom and destruction. I figure it makes sense that the lib*zip’s just want to focus on providing something simple for a program that needs to write compressed data, but you would think there would be some interest in high-performance implementations.

Hey Kerberos have you check your email lately? I already send you some letters and havn’t gotten a response yet.

Frak

Yep, for a project that keeps squealing about “freedom”, they certainly seem fixed and pushy in their ways.

We will tell you which apps you need and don’t need. We will determine which dependencies you need. We will decide what goes into the repositories.

It’s like Robert Mugabe claiming that he’s giving his people freedom of choice with a single-party election that sports different-colored ballots.

That indeed looks like the iPhone app store, but I am not aware of an iPhone app that needed to be re-compiled after a minor system update.

@TMR

It’s not walled if you have to build your own bricks…

@teel deer

Have you try http://seashore.sourceforge.net/The_Seashore_Project/About.html ? I’m not running the mac at the moment so I can’t tell it has the feature that you been looking for.

Holy Face-Palm, Batman! That has to be one of the most consistently cretinous comment threads I’ve ever seen. Twenty eight pages of ignorant off-topic drivel, ranging over almost a full year.

All by itself, it redefines the concept of “passive aggressive” behaviour.

Hi there, obvious troll!

Drats, they caught meh. Also, funny irony is funny.

Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10_5_4; en-us) AppleWebKit/525.18 (KHTML, like Gecko) Version/3.1.2 Safari/525.20.1

Adam does indeed useOSX; I’m not sure where the GNU came from, probably his imagination.

It’s been renamed GNU/OSX since GCC was stripped out in favour of LLVM/Clang, Jobs being big on irony and all.

@Thomas

There’s probably a 1940s black and white movie about this. Elegant Englishwoman with nice hat stares into screen. Large square warning dialog with unspeakably ugly font pops up: “You forgot about usability!” But darling, how could I forget about something I never had?

It’s getting far too late on this thread, btw, but Kerberos is as usual completely wrong with his primary assertion. There are far more than two things that an Operating System has to deal with, and consequently the OS is not a solved problem.

Desktops? Yup, I’ll go along with that.

Well.. and also, all the app relevant buttons (close, mini/maximize, zoom) are placed top left…

I think the most impressive thing about the screenshot, is how terrible it looks compared to the OSX version. Seriously, it looks pretty damn bad.

(And for once I’m not speaking as a fully-paid-up anti-Linux bigot. I can live quite happily with a number of Linux desktop designs — Red Hat and Centos spring to mind, although I seem to recall Mint being fine, and even the early Live CDs like Knoppix more than meet expectations.)

Every single, and I mean every single, desktop that Ubuntu has produced verges on giving me anaphylactic shock of the eyeballs. How on earth is this distribution so popular with newcomers? It’s like the Ford Edsel, but without the redeeming features.

“It did not add a dock to imitate mac osx.”

Before anyone mentions the new taskbar in Windows 7, it still works like the Windows taskbar. It just has less text and bigger pictures. Which is a plus in my book.

And that Ubuntu “dock” is horrific. I’ll bet the Firefox instance in that screenshot quits as soon as all the browser windows are closed. Let’ see: looks like OSX (or tries to), works like Windows… sound familiar?

I dunno. Mint is too green for me; it looks like a cartoony, parody desktop.

Ubuntu Satanic Edition is not only ugly but kitsch, but I suspect that was their intent.

Beautiful Linux desktops are certainly possible, but default setups seldom are beautiful.

Green is not my favorite color (in fact, it’s my least-favorite) but yeah, green desktops can look very attractive and are soothing to the eyes.

Mint, on the other hand, uses green far too much and is far too bright. It looks like an amateur overgreen desktop from the screenshots at gnome-look and deviantart.

But yes, Ubuntu took the cake. The ugly, feces/rock/dirt brown never looked appealing.

If Canonical wanted to make an African-themed distro, they could have done stylishly without windows borders that look like chocolate bars and without wallpapers that look like the inside of a glass of Coke.

By *workshop* I mean a bunch of artists and draftsmen sitting in a loft and drinking lattes (or probably, in context, a rooibus skinny). By “giving back to the community” I mean “giving back to the community.” By “shit brown,” I allude to the fact that the colour most naturally associated with Africa is “red earth,” and no, changing the thing to 1950′s Kansas Bathroom Purple doesn’t quite make the nut.

Still: aesthetics, morals, genuine idealism, or simple good taste have never been a huge motivator in Loon land, have they?

No no no. You need expert marketing advice for this sort of thing.

I’m thinking 10cc, backing

“I’m Anthrax! Fly Me”

Well, here is another site for the “community”:

http://penguinday.wordpress.com/

That blog contains most of my posts from Linsux, I know that especially the “archeology” and anti-Stallman posts were quite popular.

I’ve tagged all the postings that feature the “Linux will take over”-predictions, you can find them here:

http://en.wordpress.com/tag/freetarded-predictions/

I guess this will come handy, when SJVN and friends will flood the net again with exactly the same rhetoric.

More to the point, KEEP POSTING. Whenever the spleen is about to explode.

So one more antilinux luddites’ blog is dead. The last word left for History.

@TMR
Yeah that’d be cool.

@Lucid Lynx (Adam King)
Kerbie hasn’t been around, you can still go read http://www.binplay.com, linsux and penguinday.wordpress.com. The last having some of the best freetard owning i’ve ever read. I’m sure you’ll like it.

Hi all.
I read, and read many things here…
There is a lot of useful information. If you can collect and arrange all that information in a book, that will be great.
Please write a book. Please.

I’ll try again.. Is this blog *dead*?

First this blog goes silent, now the TM Repo is down? It seems linux hate is losing steam folks. *snicker* I knew this would happen. M$ can’t win when the enemy is the purest and most altruistic form of freedom.

Maybe the background color should change to purple, just to give the impression that it’s still alive and work is being done. Or switch the Search box to the left side.

I piss on this blog’s bloated rotten corpse *snicker*. Linux hate is Microsoft love.

Scared of what? Mafiasoft gunmen killing me during the night? It won’t happen. I have no alts. Only bad imposters.

This blog is dead. Move on, people…

Don’t do this to us kerberos! You haven’t answer my email yet!

Wow, I thought you all would be smarter than that. The last “kerberos” comment is obviously a fake. Here are two proofs:

@Tux Sux:

-When writing comments, Kerberos always has a capital K in his name, which isn’t the case for the fake commnt.

-The Gravatar ID of the real Kerberos is e8d1db0e58f664fdae3e291611b34347 whereas the Gravatar ID of the fake Kerberos is d54d755b886de9b0a99fb157d2314989.

So it’s someone impersonating Kerberos.

The farewell post is fake, because I wrote it. Of course none of you know me personally, and that’s perfectly normal. I’ve been an avid follower of this blog for the past year, although I had never commented. The fake post was a deliberate freetard-bait to create some ruckus and (hopefully) draw the real Kerberos’ attention long enough to shock the blog back to life – I guess that didn’t work out so well.

The fake was supposed to be easy to see through for the keen-eyed followers, with the lowercase “k” and a jumbled-up e-mail address, but somehow the gravatar system screwed up big-time and did place Kerberos’ own avatar next to the fake post, which, while completely unintentional, did in fact contribute to the whole ruckus thing.

A new Bugbuntu is coming. Perhaps that might revitalize the blog though in all honesty, nothing has changed and all the old arguments still apply. I believe they’re even keeping the theme colour this time.

Unity is GNOME <_<. You should research stuff before you post it.

Gotta say, I miss Piestar postings. Good to see Jerkface taking the helm at Binary Playhouse though.

Same here, I wish he would keep posting.

“why not use windows instead, it simply offers alot more and it just works.”

There’s Mac too.

The latest comment on linux hater finally reached 5000 comments, the maximum.

Well, quite an event.

GIMP is fine, I have used it for years. It is perfectly capable and efficient.

OO is fine. Its charting is worse than Excel’s, but Excel’s charting fails for anything but trivial cases anyway. Base is worse than Access, but I have no need to use Access since I can use a real database.

The Mac game market is really as terrible as ever.

If OS X has good design, then a copy also has good design. Ubuntu is available to everyone and in many languages for free and it does not have the Windows virus problems that come along with the games.

Not everyone can afford to pay 200% for Apple computers, and many of those who can do not prefer to. It is a matter of many factors including preference. You and your preferences are not holy and special.

Ubuntu is available for free because it isn’t worth crap.

When are you people going to accept that users are willing to pay for a quality product?

A browser is not a desktop operating system.

Oh, and btw, I certainly consider IE9 much better than Slowfox, and tied with Chrome.