Well, first of all, I'm not sure my medical records need to last a thousand years...
But more importantly, this is completely irrelevant:
magnetically-stored data (or even optically-stored data) degrades far faster than a paper copy.
I assume you're either making some completely uninformed, Luddite claim, or you're talking about the physical media -- in which case, you're ignoring RAID, offsite backup, checksumming (and digital signatures), and other things that are very difficult/costly or even impossible to do with paper, and which make a properly managed digital copy much, much more durable than a paper copy.
Now, on the other hand, it seems far easier for people to screw this up than with paper. Everyone understands paper, to a reasonable extent -- though as mcgrew points out, most of us probably forget that paper will degrade in 50 years or so. But when I say things like RAID, most people either have no idea what I'm talking about, or assume it has something to do with WoW or cockroaches.
But this is more a sociological problem than a technical problem -- do it right, and there's no reason a digital copy can't outlast a print copy, and every reason a print copy can't outlast a digital one.
You're right, 4.0 was the tech preview -- the alpha.
4.1 was maybe beta quality
4.2 might count as a release candidate, but with no WPA, it sure as hell wasn't a release.
4.3 looks promising. But so did 4.0, 4.1, and 4.2.
These people really need to grow up and start calling them betas -- or take a clue from Linux, establish an obvious convention (odd numbers are unstable; don't use 2.5 until we release it as 2.6), stick to it, and clearly label it a Developer Preview.
I'm really starting to wonder if they'll make it to the level of functionality 3.5 had by the time they hit 4.5.
Also, no support for transcoding in the latest version.
They basically adopted KDE4's philosophy: "Let's break everything, release it as a dot-oh release, add some sexy new features (without fixing the old ones), and blame users for upgrading when stuff doesn't work!"...only, more so.
There is currently no one version of AmaroK which does everything I want. There are two versions, each of which does a different thing that I want. And they refuse to fix the old version, because they're too busy on the new one...
Re:The Fundamental Fatal Flaw Of Desktop Linux
on
KDE 4.2.4 Released
·
· Score: 3, Insightful
Disparate people/teams all working in isolation with no single controlling authority to enforce a consistent UI over the entire system.
No such single controlling entity exists which enforces a consistent UI over any desktop system.
Play with Windows for a bit. There's the standard way you're supposed to do things, and then there's the IE7/8 way, and then there's the Office "Ribbon" way (which is implemented several ways in several different apps), and then there's the iTunes "let's make it look OSX-y" way...
Or OS X. Mac users seem to be under some really weird illusion that X programs make the system inconsistent, when even among recent apps, you have one aqua-ish look, and one chrome/steel-ish look.
I could go on...
So you have Idea/Concept 1 and 2 that are both great in isolation but when thrown together they make no sense. Everyone dumps their own pet favorite UI ideas into the mix and you get one big mess.
A mess which somehow works everywhere else, but when it comes to Desktop Linux, this is the reason people ditch it.
Not lack of drivers. Not lack of application support. Not lack of vendor support, or of preinstalled options. Not sheer FUD about new things.
No, it's the lack of a consistent UI that's the problem.
And anyone who dares to question the fatal flaw gets modded as a -1 Troll and a heretic and unbeliever
Or as someone who brings up a tired old troll which has been discounted time and time again.
And that is why Android is exploding onto Cellphones and Netbooks
"Exploding"? Really?
How's it doing compared to the iPhone?
No, Android has exactly the same "controlling authority" as everything else. That is, it doesn't -- as soon as you install a third-party app, you get whatever you get.
while standard Linux has gotten whipped right out of the market by Microsoft.
Desktop Linux was ever in a position to be "whipped out of the market" by Microsoft? News to me.
No, Microsoft has always dominated the desktop market. Linux and OS X both seem to be growing lately, but not fast enough to make a real dent.
But at the moment, Microsoft dominates the market mostly because Microsoft dominates the market.
If it worked in KDE3, it must work in KDE4. It must either be obvious how to do this, or it must be in a FAQ somewhere.
better than 3.5 in some aspects, and worse in others.
Like having no bluetooth. The ways in which it was better are irrelevant when you're missing basic functionality like my fucking mouse.
4.2 was pretty much on par, with some things much better, and some missing pieces
Missing pieces like WPA support.
The things that are much better are, honestly, things I can live without. They're cool, they make me more productive, but I can live without them.
I cannot live without such obscure things as working wireless.
The things that keep getting dropped on the floor are not obscure, they are major pieces of functionality that you could not sell a computer, regardless of OS, without some support for.
4.3 is basically better than 3.5 in almost all respects.
It damn well better be.
Because frankly, this is like XP vs Vista. When Vista was in beta testing, the apologists said, "It's a beta! Expect it to be broken!" When it was released, they said, "Everyone knows you don't buy MS software until SP1!" Well, Vista SP2 is out, and many people seem convinced it's on par with XP in most ways.
The fact that KDE4 is behind Vista is just really fucking sad, and I want to like KDE.
Giant disclaimer: I run Kubuntu, which is widely acknowledged by the KDE people as being the worst KDE-based distro ever. It's served me well in the 3.x line, but for some reason, the 4.x releases have just pulled random experimental nightly builds, incorporated them into the release...
I mean, the bluetooth issue was known about, and they put it as a "known issue", and went ahead and released, and didn't fix it for at least, oh, two months. WTF?
To anyone who hasn't tried KDE4: Don't. Wait for 4.3, maybe it'll be ready then. Or use a distro other than Ubuntu, but expect large things to break.
Knetworkmanager, in Intrepid, was KDE3-based, and for some reason, KDE3 apps can't talk to a KDE4 kwallet. Ubuntu solved this by having them not even try, meaning it forgot all my saved passwords, and saved any new ones in the clear.
The network manager plasmoid looks potentially awesome, except:
- Add two of them, and you get two notifications for network events. WTF?
- The fonts don't fit. At all. This is a common KDE4 problem for me -- it's always fucking up and chopping off part of a chunk of text for no reason -- but this is especially bad in the NetworkManager plasmoid.
- No WPA support. WTF?! Does nobody test this shit?
Going back to Intrepid is not an option, as Intrepid broke Bluetooth, and had dozens of very ugly graphical glitches and performance issues that are fixed in Jaunty. And Jaunty broke WPA.
This is why things like Flash video make me happy.
Flash video makes me angry, because it works exactly where it works, as well as it wants to work -- still requiring an order of magnitude more CPU than the competition, on the exact same file.
But...
of course you need to get the *latest* codec pack, which requires a new player, and new libraries, and since we only write the codecs and not the encoder or decoder itself you'll have to get product X too,
I really haven't seen that... and the codecs generally do just hook into most players.
Yes, everything always needs the latest. The only difference is that Flash will silently update itself. There's another all-in-one solution, though: VLC.
Human readable also means human debuggable. It means when there's something wrong with the file, I can figure it out with a text editor. It also means textual, which means it can be stored in version control, among other things.
And slowness is completely irrelevant, when we're talking about tiny bits of script or menu layout. It only becomes relevant if you're encoding the entire file that way, and I guess I don't really see the point of that.
Anyway... I would probably choose something like JSON or Yaml as a starting point. On the other hand, XML namespaces are nice, too, and HTML+microformats will get you a long way.
EBML can be read much faster than compressed XML or JSON, while being smaller at the same time.
Granted.
Both of these advantages are critical for video streams.
The stream itself is in EBML?
Well, that's interesting. Certainly if we're picking a format designed to be on a hard disk (or SSD), I'd implement it today with something like zip, with some xml files inside it. That's right, un-mux'd, as muxing is completely unnecessary pretty much anywhere except optical storage.
No, I'm talking about things like the menus, which just aren't going to be that big. This is after working on HD-DVD, which did use XML -- basically, you'd read the XML and other related assets off simple archives on the disc, and keep them in RAM while you play the actual video data.
This has nothing at all to do with the speed of Java itself. It has to do with a poorly designed API -- the HD-DVD animations were implemented by the player, either in software (likely C or ASM) or in hardware. My speculation here is that the Java animations were implemented in Java, or that there was something else wrong with that program.
Also, desktop Java, in a nice, JIT-ed VM on 2ghz+ CPUs, likely multiple cores, is many, many times faster than the tiny, restricted Java that gets put on devices. My phone manages to lag while I press buttons, and I really can't press buttons very fast.
Either way, the fact remains: A tiny portion of the screen was redrawing itself pitifully, versus smooth, fullscreen animations in HD-DVD. If you want to pretend it has nothing to do with Java, fine, but it absolutely does have to do with how they designed the menus, scripting, and animations for Blu-Ray.
Java is nothing like C++.
Dude, I didn't mention C++. This was about Java vs JavaScript.
misinformation puts a bee in my bonnet.
Then look to the plank in your own eye first.
the storage difference between blu-ray and HD was barely anything. It's auighable to cnosider it a factor when talking about pressing movies to the disk.
I never said it was a factor in who won. I said I'm kind of glad the larger format won.
And 20 extra gigabytes, or an increase of 66%, is not "barely anything".
The real reason Blu-Ray won is it's packaging.
No, the real reason Blu-Ray won is most of the studios used it. Most of the studios used it largely because the people backing it paid them large sums of money, and enticed them with better DRM. Or perhaps there were other deals we didn't know about -- regardless, it was giant robots smashing into other giant robots, with customers and real developers alike getting trampled underfoot.
It had nothing to do with customer choice -- indeed, if the customers had anything to say about it, the fact that HD-DVD is region-free should've counted for something.
developing a menu system similar to that of DVDs based on EBML;
I still don't get why MKV bothers with EBML at all, instead of using compressed XML, or a better format like JSON. It seems to me that as soon as you go binary, you lose a major point for XML in the first place.
I'd also be really curious to know what they plan for this. Seems to me an obvious choice might be to just do html. With video tags, canvas, and all that other good stuff, you have most of what you need -- just add an API to change tracks, subs, etc, and a URI scheme for accessing chapters, titles, etc...
the Apple product only received one year of free OS support, whereas the PC received seven years of free OS upgrades.
Well, define "upgrade". You're effectively comparing "service packs", which are just giant bundles of upgrades.
Does Software Update still work on 10.2?
IMHO this could be fixed quite easily if Apple allowed me to upgrade from 10.2 to 10.5 for free, and only charged for major jumps like 10.x to 11.x. That would be equivalent to when MS charges to jump from 98 to XP or XP to Vista, but provides the incremental updates without charge.
No, now you're comparing version numbers to release names. So no, you can't upgrade from Jaguar to Leopard, any more than you can upgrade from 2000 to XP, although Jaguar is 10.2 and Leopard is 10.5 -- but then, look here -- Windows 2000 is NT 5.0, XP is NT 5.1, except certain editions which are NT 5.2. Shouldn't you be demanding a free upgrade from Windows 2000 to Windows XP, or from XP to Windows Server 2003, or Windows Home Server?
There generally aren't upgrades in the OS world that aren't incremental anymore, unless you're going to count DOS to Win9x, or Win9x to NT -- and of course, OS 9 to OS X, but not really any other Mac OS before OS 9.
In other words, "major" upgrades tend to be more a management/marketing/business decision, rather than a technical one, most of the time. You could make a case that Apple shouldn't be releasing so many of those, but then again, there was a long wait between XP and Vista, and the result was a Vista that sucked for a lot of people.
However, I do remember there being a point where the PS3 was the most expensive. And I think the point stands -- the HD-DVD players did get down to $99, and were consistently around $150. Blu-Ray never touched those prices.
I'm not entirely convinced about the superior capacity -- we'll have to see how it works out. See, both Blu-Ray and HD-DVD were working on additional formats with more layers -- a four-layer HD-DVD would be 60 gigs, for example. And it may well be that HD-DVD would've been cheaper enough per disc to be cheaper per-gig.
I don't know enough about the science to say absolutely, though.
I have a HD-DVD drive for my 360 and a PS3, and I'll grant that the early software for HD-DVD was much more polished.
Not just the early software. Remember, they still have to use Java...
But also, because all that stuff I mentioned wasn't required by the original spec, I believe it's still possible to sell a Blu-Ray player that will just play the movie, and won't even run the script, let alone give it network access.
In other words, if you do want that extra stuff, you'll have to shop specifically for BD 2.0 or whatever they end up settling on.
Even if things were missing from the original HD-DVD spec, there was at least enough there for a script to run, and connect to the network, and download a new version of itself (stored in local storage) -- in other words, enough to deal with whatever new, exciting things happened.
Once you have that, it's possible to support new features even on older discs, and to gracefully fallback with software detection -- two things you can't do if you can't even be sure your software will ever run.
So, maybe eventually Blu-Ray will settle on something they'll call 2.0, or 3.0, and they'll train customers to watch for it, and it'll be to the level HD-DVD was at launch. I'm not sure they're there yet.
I do love me some lossless audio, and that just wasn't a common feature on HD-DVD.
Meaning it wasn't capable of it, or most discs didn't do it?
it was more or less a wash as far as which was the better format from an end-user standpoint,
Yeah, unfortunately, neither camp had quite gotten its act together. We had an awesome demo we were pulling together for CES, and then the war ended days before. You'd look at the schedule and see HD-DVD meetings and parties canceled...
But for existing stuff, there were a few cool things on HD-DVD that I don't think we saw on Blu-Ray, but they mostly seemed... meh. For example, pop in Goodfellas, it'll just start playing, you pull up the menu if you need it -- need to go do something else? You can bookmark your current position, then turn the player off, and go find it later. That was about the most useful thing that ever made it to market...
About the coolest thing other than that was 300's build-your-own-trailer feature, which could upload a playlist of clips (defined as start and stop times within the movie) and share them -- the top rated trailer was called "penetration" (it was all sex scenes and people getting stabbed).
I can't say for sure, but I really don't think anyone was anywhere close to those on Blu-Ray. Even so, it was just the tip of the iceberg we were working on...
But the studios were giant robots smashing into other giant robots, and we got crushed underfoot.
It's actually the "birth certificate" issue that makes me question it, but that seems to be the work of one loud user.
The "legalize marijuana" suggestions are well written and well thought out. And there is the possibility of a change for the better. Dell created a similar website, and within days, the top voted issue was a suggestion to sell Linux pre-installed -- which was a wake up call for Dell, that so many people cared about it, so they implemented it.
Toshiba built HD-DVD on top of their existing patent portfolio, and unilaterally altered the rules of the trade association charged with helming DVD's future, the DVD Forum, in order to push through adoption of their arguably-inferior standard over Sony's more advanced, more open, less expensive competing proposal.
Ok, I used to work in the industry, and that is probably the most biased and uninformed opinion I've heard. Let's break this down:
Toshiba built HD-DVD on top of their existing patent portfolio,
And Sony didn't do the same with Blu-Ray?
and unilaterally altered the rules of the trade association charged with helming DVD's future, the DVD Forum,
Citation needed. The DVD Forum has 159 registered members as of 2008, according to Wikipedia. Looking at the structure of it, I have trouble seeing how any one company could alter the rules.
In fact, reading to Sehnsucht's post, it actually looks like a reasonable change. What is the point of counting an 'abstain' as a no?
in order to push through adoption of their arguably-inferior standard over Sony's more advanced
At launch, Blu-Ray had no implementations of any sort of network access, even on the PS3. Any players other than the PS3 had absolutely abysmal performance, due to the use of Java for everything -- a simple animation, sliding a menu in that would cover a tiny portion of the screen, had to be redrawn in chunks, painfully slowly. No mandatory network, no mandatory local storage, I'm not even sure they had picture-in-picture support.
By contrast, HD-DVD had most of the features Blu-Ray was planning, but actually required and implemented in the first Toshiba players. I'm talking about a small amount of local storage, an ethernet port, picture-in-picture, scripting always enabled, and menus were written in Javascript, wrapped around an animation API that was presumably much lower-level -- menus slid smoothly onto and off of the screen, with nice translucency effects. There was a drawing API if needed, but we didn't need it.
And yes, Javascript is a better language than Java. Javascript is very Lisp-y, whereas Java is like C++, only worse.
Oh, and there's the technological advantage that an existing DVD factory can be upgraded to HD-DVD, easily.
The only technological advantage of Blu-Ray was better bandwidth and storage. But with people producing for both, the HD-DVDs generally were shipped dual-layer (30 gigs), while the Blu-Ray discs were shipped single-layer (25 gigs). No one was using that extra space, and if they were using the extra bandwidth, I sure as hell couldn't tell.
more open,
HD-DVD used only AACS for its DRM, and had no region coding. Blu-Ray used AACS and BD+, and was region-coded. Given that I consider both DRM and region coding to be evil and anti-consumer, HD-DVD is certainly the more open in that sense.
less expensive
For the manufacturers? Maybe, but as I said, there's that advantage of being able to upgrade existing DVD hardware, so there has to be some advantage. But looking at the price of movies at the time, HD-DVDs were generally cheaper, and HD-DVD players were cheaper and better than Blu-Ray players. I never saw a $100 Blu-Ray player, ever -- indeed, as I understand it, the PS3 is the cheapest to this day.
This is why you only saw Toshiba HD-DVD players, while dozens of companies were making blu-ray players.
The Toshiba players were cheap, and there was also the Xbox 360 HD-DVD drive. I have no idea if it was Toshiba inside, but the Xbox itself certainly didn't use any code from Toshiba. And there seemed to be all kinds of third-party software players.
Contrast this to Blu-Ray -- cheapest was the PS3, and it still didn't have all the features the Toshiba player did (like network access -- even though the PS3 is wired, yo
I can't see installing Windows, and a bunch of proprietary crapware, and a new monitor, and losing all my nice mplayer keyboard shortcuts (skip 10 seconds, skip 1 minute, skip 10 minutes), then paying $30-40 a disc, and losing the ability to watch the movie if I scratch the disc, or can't find space to pack it when going on a trip...
Contrast this to:
I can rent a DVD for a few dollars, pop it in, rip it, return the disc, and watch it when I have time. I can rip five or ten movies and take my laptop on vacation -- which I'm sure the video store would prefer to me actually physically taking them with me, probably scratching them... I can watch them on Linux, with mplayer, with those nice keyboard shortcuts -- find my place in under a minute, and far more accurately than the appropriate "chapter".
Or, I can get all the same features, and a beautiful high def picture, with a torrent, with the added bonus that I don't have to leave the house.
So yeah, the other disadvantage of the cutting edge is that the DRM hasn't been completely obliterated the way it has with DVD.
The problem is, of course, that the traditional broadcasts will likely interfere with both the new digital broadcasts, and other technologies in that spectrum.
The NM plasma applet is still considered beta, if not alpha quality.
In which case, I should really be blaming Kubuntu for installing it and making it the default.
If your WPA wifi network doesn't work, then either NM itself broke or your controlling program sucks; it is probably the latter here. Install nm-applet (the gnome one) or knetworkmanager (kde 3.5 based) and you'll be off and running.
Indeed, both of these work. Which raises the question: Why did Kubuntu choose an alpha-quality app for this purpose?
You could have just left this out and you'd have a much more respectable (if offtopic) post.
It is getting frustrating, though. Either Kubuntu or KDE is making me focus entirely too much on just getting shit to work with every release.
Another example: Kubuntu 8.10 was released without working Bluetooth. Again, I had to use the GNOME applet to get any Bluetooth at all.
Another example: Every now and then, I get a notification (through the shiny new notification system) which won't go away. For example, a file transfer stuck at 0 bytes, can't be stopped or paused, the window can't be moved or killed, best I can do is hide it or restart Plasma.
Another example: A recent update to Kmail seems to have removed my ability to view messages in a threaded fashion, making it useless for mailing lists.
Another example: In KDE3, I could choose "hibernate" from the logout screen, meaning I could hibernate entirely with the keyboard. Now, I have to click and hold on the "Turn off" button to get options to hibernate or suspend.
Another example: Gwenview, using the zip:// KIO handler, crashes frequently. Hell, most of what I use crashes somewhat -- even krunner. How do you fuck up an "enter command here" dialog to where it crashes periodically?
Another example: The crash handler doesn't always appear, meaning I sometimes have to reproduce a bug many times (like 5 or 10) before I actually get a trace. Can you guess why I don't like to report bugs?
Another example: I've caught kded using 100% CPU. Now, kded is a program whose sole purpose in life seems to be updating config files. Again, how do you fuck this up to where it eats 100% CPU, and refuses to write changes?
Another example: It seems very difficult to find a widget that doesn't get screwed up by my font settings. Not sure what I've done that makes it this way, but according to my clock widget, it is now 3:38 pm Tue, 2 Jur
A few of these I can see being somewhat obscure, and not necessarily easy to reproduce. But the clock? WPA? Bluetooth?
I want to like KDE. I really do. I loved KDE3, and I'm loving most of KDE4.
But this is getting absurd.
Either Kubuntu or KDE itself is nearly giving me enough material to open a Daily KDE WTF. Every major Ubuntu release breaks something critically important to a lot of users -- like, oh, Bluetooth, or wireless. The fact that those are fixed in the next release doesn't really help, when the next release is likely to break something else also.
As its name suggests, CSP allows the description of systems in terms of component processes that operate independently, and interact with each other solely through message-passing communication.
Sounds pretty much exactly like Erlang. Your description only reinforces that:
There are plenty of CSP-style large scale projects that really are assembled from discrete single-threaded programs; this works well for distribution across loosely-coupled clusters (and even topologically distant sets of those) whereas most multi-threaded programming paradigms make strong assumptions on things like expected delays and bandwidths and their derivatives (like uniformity).
It depends on the program, of course, but Erlang itself makes no such assumptions. It includes both a simple RPC system, and robust binary processing and network libraries, making it quite easy to build loosely-coupled clusters -- but each program is already written in that style to begin with, as Erlang can also run thousands of simultaneous "processes" (actors).
By using that actor model pervasively, and thinking in terms of very long-running programs -- Erlang powers things like telephone switches which need truly minimal downtime, thus, programs are built to continue handling requests even while in the process of being upgraded -- you're already thinking and developing in the way you would have to in a cluster.
Now, granted:
most multi-threaded programming paradigms make strong assumptions on things like expected delays and bandwidths and their derivatives (like uniformity).
The paradigm has nothing at all to say about this, but individual programs might. Like anything else, really -- if I develop and test on a 2.5 ghz Core 2 Duo, my software might run sluggishly on a 450 mhz Arm.
But even so, I'd argue you're a lot closer to a truly cluster-able app when developing in this way.
The most significant drawback to CSP is that the inter-SP communication protocol must be appropriate to the job, and the risk is that it's too heavyweight compared to techniques like STM or lock-full shared memory.
Erlang solves this also, for local multithreading, by using immutable memory, making it pretty much automatically a lock-free system. Each process appears to have its own memory, and is unable to interfere with the memory of another process, but it is also effectively shared memory.
Personally, I'm a bit excited about Reia, which would provide sane syntax and hopefully some decent Unicode support on top of the Erlang VM -- but it's awhile from reality, of course.
Actually, you might have a point -- I honestly don't know how well OS kernels are implemented for this sort of thing. On the other hand, Linux has been ported to machines with more cores (and CPUs!) than that before. Worst case, the kernel-level stuff won't receive a boost -- your filesystem won't go much faster -- but how much of your CPU time is currently spent there?
No, most CPU time is spent in applications, as it should be. And that's where you have the issues you describe -- either there aren't separate threads, or there are, and they're synchronized with locking. And yes, Erlang does solve a lot of that, without needing to change the OS at all.
Slashdot Mirror
Well, first of all, I'm not sure my medical records need to last a thousand years...
But more importantly, this is completely irrelevant:
magnetically-stored data (or even optically-stored data) degrades far faster than a paper copy.
I assume you're either making some completely uninformed, Luddite claim, or you're talking about the physical media -- in which case, you're ignoring RAID, offsite backup, checksumming (and digital signatures), and other things that are very difficult/costly or even impossible to do with paper, and which make a properly managed digital copy much, much more durable than a paper copy.
Now, on the other hand, it seems far easier for people to screw this up than with paper. Everyone understands paper, to a reasonable extent -- though as mcgrew points out, most of us probably forget that paper will degrade in 50 years or so. But when I say things like RAID, most people either have no idea what I'm talking about, or assume it has something to do with WoW or cockroaches.
But this is more a sociological problem than a technical problem -- do it right, and there's no reason a digital copy can't outlast a print copy, and every reason a print copy can't outlast a digital one.
You're right, 4.0 was the tech preview -- the alpha.
4.1 was maybe beta quality
4.2 might count as a release candidate, but with no WPA, it sure as hell wasn't a release.
4.3 looks promising. But so did 4.0, 4.1, and 4.2.
These people really need to grow up and start calling them betas -- or take a clue from Linux, establish an obvious convention (odd numbers are unstable; don't use 2.5 until we release it as 2.6), stick to it, and clearly label it a Developer Preview.
I'm really starting to wonder if they'll make it to the level of functionality 3.5 had by the time they hit 4.5.
Um, no, they said it'd be usable on 4.1. They only started saying this about 4.2 when it became obvious how much 4.1 sucks.
And would it have been so hard to just label it 4.0 beta?
Also, no support for transcoding in the latest version.
They basically adopted KDE4's philosophy: "Let's break everything, release it as a dot-oh release, add some sexy new features (without fixing the old ones), and blame users for upgrading when stuff doesn't work!" ...only, more so.
There is currently no one version of AmaroK which does everything I want. There are two versions, each of which does a different thing that I want. And they refuse to fix the old version, because they're too busy on the new one...
Disparate people/teams all working in isolation with no single controlling authority to enforce a consistent UI over the entire system.
No such single controlling entity exists which enforces a consistent UI over any desktop system.
Play with Windows for a bit. There's the standard way you're supposed to do things, and then there's the IE7/8 way, and then there's the Office "Ribbon" way (which is implemented several ways in several different apps), and then there's the iTunes "let's make it look OSX-y" way...
Or OS X. Mac users seem to be under some really weird illusion that X programs make the system inconsistent, when even among recent apps, you have one aqua-ish look, and one chrome/steel-ish look.
I could go on...
So you have Idea/Concept 1 and 2 that are both great in isolation but when thrown together they make no sense. Everyone dumps their own pet favorite UI ideas into the mix and you get one big mess.
A mess which somehow works everywhere else, but when it comes to Desktop Linux, this is the reason people ditch it.
Not lack of drivers. Not lack of application support. Not lack of vendor support, or of preinstalled options. Not sheer FUD about new things.
No, it's the lack of a consistent UI that's the problem.
And anyone who dares to question the fatal flaw gets modded as a -1 Troll and a heretic and unbeliever
Or as someone who brings up a tired old troll which has been discounted time and time again.
And that is why Android is exploding onto Cellphones and Netbooks
"Exploding"? Really?
How's it doing compared to the iPhone?
No, Android has exactly the same "controlling authority" as everything else. That is, it doesn't -- as soon as you install a third-party app, you get whatever you get.
while standard Linux has gotten whipped right out of the market by Microsoft.
Desktop Linux was ever in a position to be "whipped out of the market" by Microsoft? News to me.
No, Microsoft has always dominated the desktop market. Linux and OS X both seem to be growing lately, but not fast enough to make a real dent.
But at the moment, Microsoft dominates the market mostly because Microsoft dominates the market.
See, "ready for users" is undefined.
I would define it pretty clearly:
If it worked in KDE3, it must work in KDE4. It must either be obvious how to do this, or it must be in a FAQ somewhere.
better than 3.5 in some aspects, and worse in others.
Like having no bluetooth. The ways in which it was better are irrelevant when you're missing basic functionality like my fucking mouse.
4.2 was pretty much on par, with some things much better, and some missing pieces
Missing pieces like WPA support.
The things that are much better are, honestly, things I can live without. They're cool, they make me more productive, but I can live without them.
I cannot live without such obscure things as working wireless.
The things that keep getting dropped on the floor are not obscure, they are major pieces of functionality that you could not sell a computer, regardless of OS, without some support for.
4.3 is basically better than 3.5 in almost all respects.
It damn well better be.
Because frankly, this is like XP vs Vista. When Vista was in beta testing, the apologists said, "It's a beta! Expect it to be broken!" When it was released, they said, "Everyone knows you don't buy MS software until SP1!" Well, Vista SP2 is out, and many people seem convinced it's on par with XP in most ways.
The fact that KDE4 is behind Vista is just really fucking sad, and I want to like KDE.
Giant disclaimer: I run Kubuntu, which is widely acknowledged by the KDE people as being the worst KDE-based distro ever. It's served me well in the 3.x line, but for some reason, the 4.x releases have just pulled random experimental nightly builds, incorporated them into the release...
I mean, the bluetooth issue was known about, and they put it as a "known issue", and went ahead and released, and didn't fix it for at least, oh, two months. WTF?
To anyone who hasn't tried KDE4: Don't. Wait for 4.3, maybe it'll be ready then. Or use a distro other than Ubuntu, but expect large things to break.
Knetworkmanager, in Intrepid, was KDE3-based, and for some reason, KDE3 apps can't talk to a KDE4 kwallet. Ubuntu solved this by having them not even try, meaning it forgot all my saved passwords, and saved any new ones in the clear.
The network manager plasmoid looks potentially awesome, except:
- Add two of them, and you get two notifications for network events. WTF?
- The fonts don't fit. At all. This is a common KDE4 problem for me -- it's always fucking up and chopping off part of a chunk of text for no reason -- but this is especially bad in the NetworkManager plasmoid.
- No WPA support. WTF?! Does nobody test this shit?
Going back to Intrepid is not an option, as Intrepid broke Bluetooth, and had dozens of very ugly graphical glitches and performance issues that are fixed in Jaunty. And Jaunty broke WPA.
When is Kubuntu going to be good again?
This is why things like Flash video make me happy.
Flash video makes me angry, because it works exactly where it works, as well as it wants to work -- still requiring an order of magnitude more CPU than the competition, on the exact same file.
But...
of course you need to get the *latest* codec pack, which requires a new player, and new libraries, and since we only write the codecs and not the encoder or decoder itself you'll have to get product X too,
I really haven't seen that... and the codecs generally do just hook into most players.
Yes, everything always needs the latest. The only difference is that Flash will silently update itself. There's another all-in-one solution, though: VLC.
Human readable also means human debuggable. It means when there's something wrong with the file, I can figure it out with a text editor. It also means textual, which means it can be stored in version control, among other things.
And slowness is completely irrelevant, when we're talking about tiny bits of script or menu layout. It only becomes relevant if you're encoding the entire file that way, and I guess I don't really see the point of that.
Anyway... I would probably choose something like JSON or Yaml as a starting point. On the other hand, XML namespaces are nice, too, and HTML+microformats will get you a long way.
EBML can be read much faster than compressed XML or JSON, while being smaller at the same time.
Granted.
Both of these advantages are critical for video streams.
The stream itself is in EBML?
Well, that's interesting. Certainly if we're picking a format designed to be on a hard disk (or SSD), I'd implement it today with something like zip, with some xml files inside it. That's right, un-mux'd, as muxing is completely unnecessary pretty much anywhere except optical storage.
No, I'm talking about things like the menus, which just aren't going to be that big. This is after working on HD-DVD, which did use XML -- basically, you'd read the XML and other related assets off simple archives on the disc, and keep them in RAM while you play the actual video data.
Jave on devices is not slow.
This has nothing at all to do with the speed of Java itself. It has to do with a poorly designed API -- the HD-DVD animations were implemented by the player, either in software (likely C or ASM) or in hardware. My speculation here is that the Java animations were implemented in Java, or that there was something else wrong with that program.
Also, desktop Java, in a nice, JIT-ed VM on 2ghz+ CPUs, likely multiple cores, is many, many times faster than the tiny, restricted Java that gets put on devices. My phone manages to lag while I press buttons, and I really can't press buttons very fast.
Either way, the fact remains: A tiny portion of the screen was redrawing itself pitifully, versus smooth, fullscreen animations in HD-DVD. If you want to pretend it has nothing to do with Java, fine, but it absolutely does have to do with how they designed the menus, scripting, and animations for Blu-Ray.
Java is nothing like C++.
Dude, I didn't mention C++. This was about Java vs JavaScript.
misinformation puts a bee in my bonnet.
Then look to the plank in your own eye first.
the storage difference between blu-ray and HD was barely anything. It's auighable to cnosider it a factor when talking about pressing movies to the disk.
I never said it was a factor in who won. I said I'm kind of glad the larger format won.
And 20 extra gigabytes, or an increase of 66%, is not "barely anything".
The real reason Blu-Ray won is it's packaging.
No, the real reason Blu-Ray won is most of the studios used it. Most of the studios used it largely because the people backing it paid them large sums of money, and enticed them with better DRM. Or perhaps there were other deals we didn't know about -- regardless, it was giant robots smashing into other giant robots, with customers and real developers alike getting trampled underfoot.
It had nothing to do with customer choice -- indeed, if the customers had anything to say about it, the fact that HD-DVD is region-free should've counted for something.
That's one advantage of this guy's product -- you buy one codec pack, install it, and never look back. Now any video player you use will just work.
Of course, my approach is to just install VLC and make it the default, which has a similar effect.
everyone is going proprietary
It seems Youtube is going somewhat open.
No, the more interesting problem to me is:
developing a menu system similar to that of DVDs based on EBML;
I still don't get why MKV bothers with EBML at all, instead of using compressed XML, or a better format like JSON. It seems to me that as soon as you go binary, you lose a major point for XML in the first place.
I'd also be really curious to know what they plan for this. Seems to me an obvious choice might be to just do html. With video tags, canvas, and all that other good stuff, you have most of what you need -- just add an API to change tracks, subs, etc, and a URI scheme for accessing chapters, titles, etc...
the Apple product only received one year of free OS support, whereas the PC received seven years of free OS upgrades.
Well, define "upgrade". You're effectively comparing "service packs", which are just giant bundles of upgrades.
Does Software Update still work on 10.2?
IMHO this could be fixed quite easily if Apple allowed me to upgrade from 10.2 to 10.5 for free, and only charged for major jumps like 10.x to 11.x. That would be equivalent to when MS charges to jump from 98 to XP or XP to Vista, but provides the incremental updates without charge.
No, now you're comparing version numbers to release names. So no, you can't upgrade from Jaguar to Leopard, any more than you can upgrade from 2000 to XP, although Jaguar is 10.2 and Leopard is 10.5 -- but then, look here -- Windows 2000 is NT 5.0, XP is NT 5.1, except certain editions which are NT 5.2. Shouldn't you be demanding a free upgrade from Windows 2000 to Windows XP, or from XP to Windows Server 2003, or Windows Home Server?
There generally aren't upgrades in the OS world that aren't incremental anymore, unless you're going to count DOS to Win9x, or Win9x to NT -- and of course, OS 9 to OS X, but not really any other Mac OS before OS 9.
In other words, "major" upgrades tend to be more a management/marketing/business decision, rather than a technical one, most of the time. You could make a case that Apple shouldn't be releasing so many of those, but then again, there was a long wait between XP and Vista, and the result was a Vista that sucked for a lot of people.
Points well taken.
However, I do remember there being a point where the PS3 was the most expensive. And I think the point stands -- the HD-DVD players did get down to $99, and were consistently around $150. Blu-Ray never touched those prices.
I'm not entirely convinced about the superior capacity -- we'll have to see how it works out. See, both Blu-Ray and HD-DVD were working on additional formats with more layers -- a four-layer HD-DVD would be 60 gigs, for example. And it may well be that HD-DVD would've been cheaper enough per disc to be cheaper per-gig.
I don't know enough about the science to say absolutely, though.
I have a HD-DVD drive for my 360 and a PS3, and I'll grant that the early software for HD-DVD was much more polished.
Not just the early software. Remember, they still have to use Java...
But also, because all that stuff I mentioned wasn't required by the original spec, I believe it's still possible to sell a Blu-Ray player that will just play the movie, and won't even run the script, let alone give it network access.
In other words, if you do want that extra stuff, you'll have to shop specifically for BD 2.0 or whatever they end up settling on.
Even if things were missing from the original HD-DVD spec, there was at least enough there for a script to run, and connect to the network, and download a new version of itself (stored in local storage) -- in other words, enough to deal with whatever new, exciting things happened.
Once you have that, it's possible to support new features even on older discs, and to gracefully fallback with software detection -- two things you can't do if you can't even be sure your software will ever run.
So, maybe eventually Blu-Ray will settle on something they'll call 2.0, or 3.0, and they'll train customers to watch for it, and it'll be to the level HD-DVD was at launch. I'm not sure they're there yet.
I do love me some lossless audio, and that just wasn't a common feature on HD-DVD.
Meaning it wasn't capable of it, or most discs didn't do it?
it was more or less a wash as far as which was the better format from an end-user standpoint,
Yeah, unfortunately, neither camp had quite gotten its act together. We had an awesome demo we were pulling together for CES, and then the war ended days before. You'd look at the schedule and see HD-DVD meetings and parties canceled...
But for existing stuff, there were a few cool things on HD-DVD that I don't think we saw on Blu-Ray, but they mostly seemed... meh. For example, pop in Goodfellas, it'll just start playing, you pull up the menu if you need it -- need to go do something else? You can bookmark your current position, then turn the player off, and go find it later. That was about the most useful thing that ever made it to market...
About the coolest thing other than that was 300's build-your-own-trailer feature, which could upload a playlist of clips (defined as start and stop times within the movie) and share them -- the top rated trailer was called "penetration" (it was all sex scenes and people getting stabbed).
I can't say for sure, but I really don't think anyone was anywhere close to those on Blu-Ray. Even so, it was just the tip of the iceberg we were working on...
But the studios were giant robots smashing into other giant robots, and we got crushed underfoot.
It's actually the "birth certificate" issue that makes me question it, but that seems to be the work of one loud user.
The "legalize marijuana" suggestions are well written and well thought out. And there is the possibility of a change for the better. Dell created a similar website, and within days, the top voted issue was a suggestion to sell Linux pre-installed -- which was a wake up call for Dell, that so many people cared about it, so they implemented it.
I would hope the same can happen here.
Toshiba built HD-DVD on top of their existing patent portfolio, and unilaterally altered the rules of the trade association charged with helming DVD's future, the DVD Forum, in order to push through adoption of their arguably-inferior standard over Sony's more advanced, more open, less expensive competing proposal.
Ok, I used to work in the industry, and that is probably the most biased and uninformed opinion I've heard. Let's break this down:
Toshiba built HD-DVD on top of their existing patent portfolio,
And Sony didn't do the same with Blu-Ray?
and unilaterally altered the rules of the trade association charged with helming DVD's future, the DVD Forum,
Citation needed. The DVD Forum has 159 registered members as of 2008, according to Wikipedia. Looking at the structure of it, I have trouble seeing how any one company could alter the rules.
In fact, reading to Sehnsucht's post, it actually looks like a reasonable change. What is the point of counting an 'abstain' as a no?
in order to push through adoption of their arguably-inferior standard over Sony's more advanced
At launch, Blu-Ray had no implementations of any sort of network access, even on the PS3. Any players other than the PS3 had absolutely abysmal performance, due to the use of Java for everything -- a simple animation, sliding a menu in that would cover a tiny portion of the screen, had to be redrawn in chunks, painfully slowly. No mandatory network, no mandatory local storage, I'm not even sure they had picture-in-picture support.
By contrast, HD-DVD had most of the features Blu-Ray was planning, but actually required and implemented in the first Toshiba players. I'm talking about a small amount of local storage, an ethernet port, picture-in-picture, scripting always enabled, and menus were written in Javascript, wrapped around an animation API that was presumably much lower-level -- menus slid smoothly onto and off of the screen, with nice translucency effects. There was a drawing API if needed, but we didn't need it.
And yes, Javascript is a better language than Java. Javascript is very Lisp-y, whereas Java is like C++, only worse.
Oh, and there's the technological advantage that an existing DVD factory can be upgraded to HD-DVD, easily.
The only technological advantage of Blu-Ray was better bandwidth and storage. But with people producing for both, the HD-DVDs generally were shipped dual-layer (30 gigs), while the Blu-Ray discs were shipped single-layer (25 gigs). No one was using that extra space, and if they were using the extra bandwidth, I sure as hell couldn't tell.
more open,
HD-DVD used only AACS for its DRM, and had no region coding. Blu-Ray used AACS and BD+, and was region-coded. Given that I consider both DRM and region coding to be evil and anti-consumer, HD-DVD is certainly the more open in that sense.
less expensive
For the manufacturers? Maybe, but as I said, there's that advantage of being able to upgrade existing DVD hardware, so there has to be some advantage. But looking at the price of movies at the time, HD-DVDs were generally cheaper, and HD-DVD players were cheaper and better than Blu-Ray players. I never saw a $100 Blu-Ray player, ever -- indeed, as I understand it, the PS3 is the cheapest to this day.
This is why you only saw Toshiba HD-DVD players, while dozens of companies were making blu-ray players.
The Toshiba players were cheap, and there was also the Xbox 360 HD-DVD drive. I have no idea if it was Toshiba inside, but the Xbox itself certainly didn't use any code from Toshiba. And there seemed to be all kinds of third-party software players.
Contrast this to Blu-Ray -- cheapest was the PS3, and it still didn't have all the features the Toshiba player did (like network access -- even though the PS3 is wired, yo
I can see the difference in quality.
I can't see installing Windows, and a bunch of proprietary crapware, and a new monitor, and losing all my nice mplayer keyboard shortcuts (skip 10 seconds, skip 1 minute, skip 10 minutes), then paying $30-40 a disc, and losing the ability to watch the movie if I scratch the disc, or can't find space to pack it when going on a trip...
Contrast this to:
I can rent a DVD for a few dollars, pop it in, rip it, return the disc, and watch it when I have time. I can rip five or ten movies and take my laptop on vacation -- which I'm sure the video store would prefer to me actually physically taking them with me, probably scratching them... I can watch them on Linux, with mplayer, with those nice keyboard shortcuts -- find my place in under a minute, and far more accurately than the appropriate "chapter".
Or, I can get all the same features, and a beautiful high def picture, with a torrent, with the added bonus that I don't have to leave the house.
So yeah, the other disadvantage of the cutting edge is that the DRM hasn't been completely obliterated the way it has with DVD.
Regarding your sig, chances are that old hardware will run 10.5. So, it's not the machine that doesn't last, it's the software...
The problem is, of course, that the traditional broadcasts will likely interfere with both the new digital broadcasts, and other technologies in that spectrum.
The NM plasma applet is still considered beta, if not alpha quality.
In which case, I should really be blaming Kubuntu for installing it and making it the default.
If your WPA wifi network doesn't work, then either NM itself broke or your controlling program sucks; it is probably the latter here. Install nm-applet (the gnome one) or knetworkmanager (kde 3.5 based) and you'll be off and running.
Indeed, both of these work. Which raises the question: Why did Kubuntu choose an alpha-quality app for this purpose?
You could have just left this out and you'd have a much more respectable (if offtopic) post.
It is getting frustrating, though. Either Kubuntu or KDE is making me focus entirely too much on just getting shit to work with every release.
Another example: Kubuntu 8.10 was released without working Bluetooth. Again, I had to use the GNOME applet to get any Bluetooth at all.
Another example: Every now and then, I get a notification (through the shiny new notification system) which won't go away. For example, a file transfer stuck at 0 bytes, can't be stopped or paused, the window can't be moved or killed, best I can do is hide it or restart Plasma.
Another example: A recent update to Kmail seems to have removed my ability to view messages in a threaded fashion, making it useless for mailing lists.
Another example: In KDE3, I could choose "hibernate" from the logout screen, meaning I could hibernate entirely with the keyboard. Now, I have to click and hold on the "Turn off" button to get options to hibernate or suspend.
Another example: Gwenview, using the zip:// KIO handler, crashes frequently. Hell, most of what I use crashes somewhat -- even krunner. How do you fuck up an "enter command here" dialog to where it crashes periodically?
Another example: The crash handler doesn't always appear, meaning I sometimes have to reproduce a bug many times (like 5 or 10) before I actually get a trace. Can you guess why I don't like to report bugs?
Another example: I've caught kded using 100% CPU. Now, kded is a program whose sole purpose in life seems to be updating config files. Again, how do you fuck this up to where it eats 100% CPU, and refuses to write changes?
Another example: It seems very difficult to find a widget that doesn't get screwed up by my font settings. Not sure what I've done that makes it this way, but according to my clock widget, it is now 3:38 pm Tue, 2 Jur
A few of these I can see being somewhat obscure, and not necessarily easy to reproduce. But the clock? WPA? Bluetooth?
I want to like KDE. I really do. I loved KDE3, and I'm loving most of KDE4.
But this is getting absurd.
Either Kubuntu or KDE itself is nearly giving me enough material to open a Daily KDE WTF. Every major Ubuntu release breaks something critically important to a lot of users -- like, oh, Bluetooth, or wireless. The fact that those are fixed in the next release doesn't really help, when the next release is likely to break something else also.
From Wikipedia:
As its name suggests, CSP allows the description of systems in terms of component processes that operate independently, and interact with each other solely through message-passing communication.
Sounds pretty much exactly like Erlang. Your description only reinforces that:
There are plenty of CSP-style large scale projects that really are assembled from discrete single-threaded programs; this works well for distribution across loosely-coupled clusters (and even topologically distant sets of those) whereas most multi-threaded programming paradigms make strong assumptions on things like expected delays and bandwidths and their derivatives (like uniformity).
It depends on the program, of course, but Erlang itself makes no such assumptions. It includes both a simple RPC system, and robust binary processing and network libraries, making it quite easy to build loosely-coupled clusters -- but each program is already written in that style to begin with, as Erlang can also run thousands of simultaneous "processes" (actors).
By using that actor model pervasively, and thinking in terms of very long-running programs -- Erlang powers things like telephone switches which need truly minimal downtime, thus, programs are built to continue handling requests even while in the process of being upgraded -- you're already thinking and developing in the way you would have to in a cluster.
Now, granted:
most multi-threaded programming paradigms make strong assumptions on things like expected delays and bandwidths and their derivatives (like uniformity).
The paradigm has nothing at all to say about this, but individual programs might. Like anything else, really -- if I develop and test on a 2.5 ghz Core 2 Duo, my software might run sluggishly on a 450 mhz Arm.
But even so, I'd argue you're a lot closer to a truly cluster-able app when developing in this way.
The most significant drawback to CSP is that the inter-SP communication protocol must be appropriate to the job, and the risk is that it's too heavyweight compared to techniques like STM or lock-full shared memory.
Erlang solves this also, for local multithreading, by using immutable memory, making it pretty much automatically a lock-free system. Each process appears to have its own memory, and is unable to interfere with the memory of another process, but it is also effectively shared memory.
Personally, I'm a bit excited about Reia, which would provide sane syntax and hopefully some decent Unicode support on top of the Erlang VM -- but it's awhile from reality, of course.
Yeah, I know I'm offtopic and trolling, and I should be filing bug reports instead...
But yeah, thanks for the NetworkManager widget. Now my WPA-enabled wireless networks don't work.
WTF, KDE?
I look forward to having my touchpad break in KDE 4.3, and my keyboard break in KDE 4.4, and maybe, just maybe, a beta-quality release by KDE 4.5.
It's the applications.
Actually, you might have a point -- I honestly don't know how well OS kernels are implemented for this sort of thing. On the other hand, Linux has been ported to machines with more cores (and CPUs!) than that before. Worst case, the kernel-level stuff won't receive a boost -- your filesystem won't go much faster -- but how much of your CPU time is currently spent there?
No, most CPU time is spent in applications, as it should be. And that's where you have the issues you describe -- either there aren't separate threads, or there are, and they're synchronized with locking. And yes, Erlang does solve a lot of that, without needing to change the OS at all.