Im not an computer expert or anything, but copying everything from RAM to HD could hurt performence and stuff.
Well, it's clear you're not an expert.
From what I understand, it doesn't copy all RAM, it asks objects to serialize themselves -- which is pretty much how programs would save state now, only it's done automatically.
Even if it was, how do you think hibernation works?
I think linux is way better then any windows
Irrelevant. He's trying to build a system much better than both of them.
Also his stuff is still vaporware
Maybe. It has been done before, though. Look up Smalltalk.
I would like this (concept) if it where open source...
Well, some of those Smalltalk VMs are... I do agree with you there.
Newton OS had the same thing. It caused me to lose data twice when I accidentally deleted a large part of a Newton Works document and then did something else. Undo only undid the something else; the deletion became permanent as soon as it passed out of the one-step undo buffer.
Two things:
First, your problem seems to be more with the fact that undo history was only one level deep than anything else.
And second, you do need revision control, and it needs to be easy enough for the masses, but more powerful than just "undo".
Keep in mind, the whole OS is designed this way, including all programs.
Let me give you an example of what happens when it's implemented as a library: GNOME and KDE sessions. At least in KDE, it's possible to save a session, or even to have it autosave when you logout. It will remember all open programs, and the geometry of their windows. It will even query the programs, asking them to save their state.
Now, this would be awesome, wouldn't it? It'd be a lot more efficient than hibernate/resume, if it worked -- for example, an ODF (plus some simple geometry and state) is much smaller than the entire virtual image of OpenOffice. If the programs were written well, to load only what they need on demand (and thus start much faster), the whole system would shut down and wake faster.
You could even start to have multiple sessions, maybe mapped to virtual desktops, maybe not, so that when you boot, you could choose whether to have it launch your web browser, text editor, and terminals, or have it launch your mail client, IM client, and softphone, or maybe have one that just launches whatever movie you were playing (which would resume from the exact moment it was at when you shut down)...
Problem is, too many programs don't support this. Some, like Firefox, seem to supply their own session management. Some don't even try, and thus, when the DE tries to resume them, it ends up launching a fresh instance. Some can't be persisted, due to their fundamental architecture -- how would you propose to save the state of a running terminal?
So, doing it as a library doesn't work, unless everything's using that library. If everything's using that library, that's pretty much what you get.
And sometimes, you do have to enforce sometimes performance-decreasing features in order to provide a better user experience. Imagine if filesystem access was just a library, and programs had access to the entire disk. It might be interesting to build an OS that way, but even if you did, I imagine you'd want to restrict most user-level programs to dealing with the POSIX API, and being bound by Unix permissions and POSIX ACLs.
Only the people who have past/present guilt about hiding something from someone will feel this is a horrible idea.
So, it's your opinion that stalking should be OK. Got it.
I don't think I really have to say much more than that -- if there are any women on Slashdot who have been raped or abused, they'll likely tear you apart right about now.
Remember, this is Opt-in folks.
Yes, that is true. But that doesn't seem to be your argument...
People in relationships, you all better wake up, your gonna have to be honest by choice or by technology.
People in relationships should also trust each other enough not to have to use things like this.
The fact that you have nothing to hide doesn't in any way imply that you'd never want privacy. Consider the most basic sense -- I'm reasonably secure about my body, but that doesn't mean I'm inviting people to set up a shower cam.
Nor does wanting to hide something imply that there is something shameful to hide. The simplest example there would be a surprise party. The more frightening example is when people really are after you -- witness protection exists for a reason.
Now, if you've opted in, fine. But what you seem to be implying here is that no one should ever have a reason to opt out, and you're flat-out saying that the only reason to ever want to hide is guilt, and that is simply not true.
What happens when they use your stuff anyway? I know that if I came home and found someone had broken in and used my stuff (eg, ate my food, slept in my bed, and walked off with my electronics, I'd call the police.
I think you just answered your own question.
I can do things (like keep the doors locked) to reduce the chances that someone breaks into my place.
And you can do things like keep your phone in your pocket, and not let people play with it. If it gets stolen, then they aren't tracking you, they're tracking the thief.
.. people are dumb, and lots of people will install this without thinking it through.
And how is it Google's fault that people are dumb?
Let's start addressing the real problem -- that of dumb people -- and stop trying to protect them from themselves.
Take another example: Teenagers are always posting shit on Myspace, Livejournal, Facebook, and everything else they can find. They aren't always obscuring their own name, meaning future employers can look them up this way. So is the right solution to "protect" these kids by blocking all access to social networking, or to block employers from using search engines to evaluate potential hires?
I have a better idea -- let's educate the ones we can, and let the rest suffer from their own mistakes.
Assuming you can make the case for doing so -- and it looks as though you have plenty of help there -- it might be worth clarifying what an "open standard" means.
For something to be considered an open standard, it must meet the following criteria:
- A comprehensive formal specification. (This should be obvious.)
- At least one reference implementation for which source code is freely available. (It doesn't have to be freely re-usable, so long as it's there.) OR, many very different implementations which can communicate. (There probably isn't a reference HTML/CSS renderer, but there are enough implementations that one isn't needed.)
- No legal issues for either of the above points, or the use of the specification. (Obvious example: No patents allowed, unless they've been turned over to the public domain.)
It should also meet the following criteria:
- A well-written, accessible, comprehensive formal specification. Or, both a formal specification and easier-to-read documentation.
- Both an official open source reference implementation, and several competing implementations.
- Corporate backing -- especially a corporate stake in it. This implies that said corporation has had their lawyers verify that there are no legal issues.
- Simple, clean design, especially relative to other standards providing the same thing. For example, if the choice is between SOAP and XML-RPC, you probably want XML-RPC -- and you might prefer REST to either of those, especially if your data is not XML.
- Popularity. This really matters the least, so long as the others are met -- it's more important that I can hold the ideals of REST in my head, and implement it from scratch in a few lines of code, than that there are probably more SOAP and XML-RPC implementations. But it shouldn't be ignored -- it would be insane to try to replace HTML with something completely different, for instance. (Both HTML5 and XHTML are incremental improvements, and are sane. Trying to replace HTML with a YAML-based format would not be sane.)
I'm not suggesting that policy has to follow these to the letter, but that's what I personally consider an open standard, and especially, what I consider to be a good standard. In the past, when I've called Microsoft's "Open" XML various names -- "Neither open nor standard" comes to mind -- these are the guidelines I was using.
It doesn't seem to stop bugs from getting into Open Source software.It doesn't seem to stop bugs from getting into Open Source software.
Stop all? No. It would certainly stop some.
But this was also about fixing bugs once they're there.
This idea that with OSS if you find a bug you can patch it is also bullshit for the most part. It assumes that the user is both a competent programmer
If we're talking about a security hole, I'd say that's a fair bet.
Even if they're not, they have options -- they can hire someone else to fix it, if it's important to them. To an individual, that might be unrealistic -- to a large organization doing a security audit, it's essential.
As far as most people are concerned, if they find a bug in OSS, they will just report it to the development organisation in much the same way as they would for proprietary software. At that point, they are in pretty much the same boat. Almost.
Almost. Because at that point, there's still the part where the community generally has the same priorities and the same motivations that they do.
In the rare case where that is not true, there is nothing stopping you (or anyone else) from forking it and attempting to build a community who does have your best interests at heart. I don't mean to imply that this is easy, only that it is possible -- with a proprietary product, you're going to have all the same challenges of gathering a development team and convincing them to work on your fork, with the additional challenge (or impossibility) of getting permission from the original vendor, let alone source code.
With proprietary software, quite often, the company's interests are not at all aligned with the consumer's interests. Often, the consumer is not the same as the customer. And if that's the case, there's nothing you can do about it -- again, forking is much harder, if it's even possible.
The whole development process is open to scrutiny from the customers. This is almost never the case with proprietary software.
any flaws on commercial applications tend to get patched a lot faster than on open source, as the vendors producing the software have a lot more to lose than an open source programmer
This ignores the "many eyes" factor, and the additional effect that anyone who finds a security vulnerability can also patch it, and can inform people of the patch at the same time as the vulnerability. Contrast this to proprietary software, where anyone who does find a breach will also find that the best they can do is report it to the vendor and hope for the best -- and when some of them take many months to be patched, it may be worthwhile for them to start exploiting it, if for no other reason than to get Microsoft to take them seriously.
All of those have been argued to death... Let's assume I'm completely wrong. There's still the fact that there are many corporations which support open source. If an IBM, or a RedHat, or a Canonical ships an insecure product, they have every bit as much to lose as a proprietary vendor -- often moreso, as they tend to have quite a lot more competition.
All of which has very little to do with the supposed counterargument:
We need to move in the direction of what are known as 'open standards' - in effect, creating a common language for government IT. This technical change is crucial because it allows different types of software and systems to work side by side in government.
Microsoft aside, there is plenty of proprietary software that not only supports open standards, but actually revels in them. Unless the argument about security implied that there's an inherent insecurity in ODF itself, I don't see what the relevance is.
However, this article unfortunately presents it as an argument of security against hot new stuff. I don't think anyone is urging the government to become less secure.
But on a modern system how fast most OS features act is the split between milliseconds and who really cares?
Depends how much you're using those features, and the split can be rather large.
For example: Even if you got a decent POSIX system on Windows, I doubt you'd get an efficient fork. On OS X, process creation time is relatively slow. On Linux, fork is very, very fast, uses copy-on-write, and actually makes Unix pipes a sane alternative to threads, under some circumstances.
Now, my information on that is all very old, so it's possible I'm completely wrong, and Windows and OS X had caught up.
The summary is highly misleading "Ubuntu as much as twice as fast!" At extremely short unnoticeable tasks which no human would care to measure except in a benchmark.
Yeah, that's probably true. I haven't actually looked at the benchmark, and I suspect that once you've got 5 gigs of RAM to throw at it, and a fast enough CPU, the bloat starts looking less important.
Bah. Reading the other post, I should stop using the word "throw".
It's got full SDL support, so it should work fine.
Does "full SDL support" include full OpenGL drivers? Can I play Doom 3 on it?
That's not really what Haiku is for. It's a desktop system. You should not use it for any headless tasks.
I consider that to be a detriment. It's nice to be able to deploy software to the same environment I use at home. It's one less thing to think about when I'm wondering why it worked on my development machine, but not the test server.
You're trapped in the X mindset. I can't think of any other windowing systems that behave like that
Which in no way means that it is a bad way to behave.
Or do you mean, you've never seen another windowing system get too screwed up to boot, either from a driver issue, configuration issue, or something's wrong with your desktop session?
Another example: If something has trapped all my input within X -- say it's a game, and it's capturing all keyboard and mouse actions, and has stopped responding -- I can still ctrl+alt+f1 and kill it. Or, worst case, the whole system's hosed, I can ctrl+alt+backspace -- without losing my network connections, or anything non-X I've left running -- and then re-enter my session, only this time, I have the benefit that the entire windowing system might still be in the filesystem's cache.
All of these are real, user-visible advantages to X. How is your system "better" by not supporting these features?
WTF does "boots cleaner" mean?
You really need to try it.
Translation: You can't explain it. Maybe it's a completely subjective feeling.
it won't be long before it quietly glides past Ubuntu in overall ease of use.
If that happens, it'll be worth thinking about, but don't forget the "worse is better" component of ease of use. Linux was here first. All my stuff works on Linux. Even if it is easier to boot, connect to a network, and get a browser, that doesn't matter at all if it doesn't have my favorite editor, or if I have to rewrite Amarok from scratch...
With a few more developers, I think Haiku has far more potential to beat Windows and Mac in desktop usability than linux does.
Linux has already broken both in several important ways. The rest is all about software support and commercial support.
linux is reaching certain limits in the usability and overall "smoothness" department that a nice ground-up redesign would do nicely to correct.
Maybe. However, I think it would be foolish to throw away the majority of open source drivers in order to accomplish that -- at least, not yet.
In commercial development, it's not strange to wipe out your code and rethink your work from the bottom up occasionally. The unix world has been incapable of doing this for decades...
Which "commercial development" are you talking about?
If it's OS X, you might have a point, if they didn't simply take NeXT and develop it into a usable product -- and remember, NeXT is a Unix, and so is OS X.
if it's Windows, well, Vista was a disaster, and that's the closest they've come since Windows 2000... which was taken from NT, and large chunks of the 98/ME line. And that's the closest since Windows 95... which just sucked in tons of stuff from 3.1, both of which ran on top of DOS.
If it's anything else, you're either talking about an entirely embedded system, which implies much less code and compatibility required, or you're talking about a piece of software, and if you truly believe the open source community has never started from scratch on any software, you clearly haven't been paying much attention. The obvious example would be KDE4.
I think I get what you've said here, though:
It's smaller and of a more sane design. It was designed by a professional ke
Haiku is not just a kernel, it's a whole operating system.
Even better! Port it! With the noise you've been making about POSIX, you'd think Haiku would be POSIX-compliant, right?
What is it that the Haiku kernel does that the Linux kernel doesn't? And if there is something, why couldn't the Linux kernel be adapted to do that as well?
All of the things you mention in linux are rather hackish and feel overall awkward.
Are they implemented at all in Haiku?
In Haiku, when I boot plug in a network cable, does the OS notice and automatically try to connect?
Can I connect to an open wireless network in two clicks, the way I can on OS X or Ubuntu? Is my wireless card even supported?
The speed, elegance, or efficiency of a nonworking or nonexistent program is irrelevant.
Linux always feels sort of heavy when compared to systems like Mac or Be that just do exactly what they want to do.
Oh, they do exactly what they want to, sure. And as long as you want to do exactly what Steve Jobs has thought of, they're great!
And as soon as you don't, you're out of luck. Something as simple as sloppy focus will likely never be implemented in OS X, never mind sane keyboard shortcuts.
I think the point is that a redesigned approach might provide similar functionality with reduced resources because there are fewer unecessary layers of abstraction.
Except it doesn't provide similar functionality. From what you've told me, I can no longer play games, use wireless, or watch HD videos, despite that Be was supposed to be a multimedia operating system.
Removing the networking aspect from the window manager and mode switching with the kernel is a simplification.
I'll agree with you there, only because the network design of X seems to have been broken to begin with and has only gotten worse over the years. I used to enjoy being able to run programs remotely using ssh's X-forwarding, but lately, the various X11 environments seem to crash when connecting to a remote system that isn't identical to their own.
Regardless, the networking aspect is a Good Idea, just poorly implemented. Why shouldn't I expect to be able to run any program, anywhere, as if it were local to my desktop?
The driver system is simplified
If by "simplified" you mean "missing major drivers", OK.
it's not designed to run headlessly so it runs natively with a GUI.
Ew. This is one of the things I always hated about Windows Server -- no debugging over a serial console, and you need a video card in the machine, even if it's never used. No thanks.
And on a desktop, you've now set yourself up such that, if the video driver ever has a problem, you're hosed. I can at least boot to text-mode and try to fix the problem -- depending on how serious it is, I could try various configurations, launching an X server, killing it, launching it again...
Serviceability is also good stuff for a desktop.
I'll grant you, mode-switching in the kernel is good. Video drivers really belong in the kernel, for performance reasons, as long as we have a monolithic kernel approach.
But what reason could you possibly have for requiring it? What do you gain by not being able to remove it, like any other kernel module? Like, I don't know, filesystems, sound drivers, RAID controller drivers? Easily most of what a Linux system needs at boot can be swapped out -- why is that a bad thing?
It boots faster
What are you comparing? There are versions of Linux which boot in under a second. Please be specific.
I don't think you understand- the point is that Haiku is designed from the ground up for the desktop based on BeOS's example.
And Linux was designed from the ground up to be a terminal emulator. Neither are being used for what they were originally designed for.
The point is that it's not linux
If that's your only point, there are lots of things that aren't Linux.
it's designed for the desktop.
So is Linux, for that matter. At the time, desktops needed terminal emulators.
The companies supporting linux are doing it for the server and HPC--
And you somehow think they control the direction of kernel development.
What's with this Katamari Damacy attitude in open source? I thought F/OSS was supposed to be more flexible.
And Linux is. It can be in the server, and in HPC, and on the desktop. There are many things other than kernels which contribute to that.
A trivial example: There's no reason the same networking system, at the kernel level, can't work for both server and desktop. Certainly the same firewall will be useful on both. But the Debian networking system is more useful for servers (and the loopback interface), whereas NetworkManager is more useful for desktops (and especially laptops).
The kernel really isn't a great place to differentiate these. The kernel really should just be a giant device driver -- which is the point. Does Haiku support my wireless card? Can it provide accelerated OpenGL on nvidia hardware? Since Be was so good at video, once upon a time, how well does mplayer work on Haiku, especially with, say, hardware-accelerated h.264 playback?
It's becoming increasingly clear that X11 will never be hacked into a usable local display option.
Not as long as nvidia is hacking it, no.
The open source community badly needs something more desktop centric.
Show me an open option. Until then, well, certain Intel cards seem to do better with Compiz -- mostly, I would guess, because of more open drivers.
Or, you know, you can turn off desktop effects, and let it use technology that was boring in 1995 to give you a smooth desktop experience.
I wish more developers would pay attention to Haiku for the desktop so we can have a POSIX compliant free desktop operating system that is built with the desktop in mind.
Far too late for that -- the Linux kernel supports entirely too much stuff, most of it very well. And are there any good nvidia drivers, open or not, for Haiku?
If you really want to make that succeed, start trying to port the more interesting Haiku features to Linux.
Neither runs the other's software (without proper tools, and even then it's not nearly perfect),
Close enough, though. Aside from Wine running similar tools on both platforms -- and, when Wine works, it's often faster than running the same app on Windows -- there's plenty of cross-platform development.
Let me put it this way: Suppose I'm a Java developer with Eclipse. That'll run fine on any platform I throw at it. But, even before I get to my own software, Eclipse is such a hog that I'll want every ounce of performance I can throw at it.
For that matter, if I'm developing a Java program -- or Ruby, or Python, or anything else sufficiently cross-platform -- I may well care when it gets to deployment time which OS is faster. If developing a new app, I may choose to support one platform over another for performance reasons.
It's probably not as useful as benchmarks within an OS (between Linux filesystems, say) or between POSIX-compliant OSes (but these benchmarks don't test the Windows POSIX layer, I'm sure), but it's worth mentioning.
Oh, and I like being amaturish -- a big HA HA to everyone who tried to convince me that Vista is fast, when you give it enough RAM.
I have no problems with corporations using P2P to cut costs.
Where I have a problem is when they also want to revoke rights.
If they just used BitTorrent to deliver streams, no problem, I applaud that. Maybe multicast would be better for some streams, but whatever. If it's really a problem, I can limit my upload, or block them from uploading at all.
However, when their EULA prevents me from reading my own logs (much less any of the above shaping), that's when I have a problem.
I think the point is, rather, that there's plenty of old code they could release now, as id has done with old games like Doom (1, 2) and Quake (up to 3)
What makes it slightly scary is that it claims to be a parking violation.
However, I would likely make a very loud noise about being required to not only have Internet, but also a specific browser and a specific operating system, and having to download their software.
For unemployment, at least here, the entire thing is done over the Internet. However, the website pretty much works in any browser (though the layout was slightly off in Konqueror), and if you don't have Internet (or a computer), you walk to the unemployment office, they sit you down at one of their computers, and you do it there.
For a parking violation to be so unaccessible has got to be violating some regulation somewhere.
Just goes to show that no matter how much protection you have on the tech side, there's always a social engineering way around it.
Also goes to show how a little paranoia goes a long way.
This always confuses me. It's legal to download DDLs at random?
My assumption has always been that since I do have a valid Windows license, it would be splitting hairs to try to find the exact version that's in my Windows (when I could probably download a version known to work with that combination of app / wine version).
However, this seems to happen less and less as time goes by. Used to be, you'd need IE or some IE DLLs. Now, most things will work just fine with the Gecko shim.
Ok. I always use a "test" user for downloaded binaries (other than the OS updates.) For example, to avoid compiling google-earth, just downloaded the binary and always use it on the "test user", so malware infection danger is minimized to that dummy account.
I tend to create one per piece of software, when I bother. If it's one "test" user, that could start to get like nobody -- because you run so many things under the same user, any one of them could blow up more than it should.
As an example, I got tired of waiting for Ubuntu to update Rubygems, and the Rubygems update itself didn't work well. Ruby 1.9 was always a bit behind, too. So I downloaded Rubygems and Ruby 1.9, gave them each a user, and had them install to their home directory. Now, if I need to completely wipe it and start over, I won't be picking pieces of Rubygems out of every corner of my system -- I just nuke/home/rubygems, and start over.
Since Wine is implemented in user mode, I think that that prevents me from liberal access to my partitions.
I see -- but, this was a response to you being afraid of installing Vista again. I wouldn't be afraid to have Vista or XP on this laptop, if it's properly backed up. If it's an emergency, I can pull out the SATA disk and put it in the fileserver where the backups are. Your "two full days" don't seem realistic, with good backups.
What I'd be more afraid of is a deliberate, specialized attack -- something which finds my home partition and grabs my ssh keys, for example. Ok, those can be encrypted, but realistically, I should be doing full-disk encryption to protect against that...
That's also the reason I hate IBM when forces me to run their installers as root (I think in order to install some secret-shitty-license-key-file)
I wonder if those will run chroot'ed? Or in a VM? Debian's "fakeroot" might also help you here.
KVM isn't going to be as nice as the bare metal, and an ntfsclone image is going to be easier to handle than a raw (even sparse) image.
No, the nice part about something like KVM or Qemu is when you run it in snapshot mode -- all changes are written, copy-on-write fashion, to a temporary file. If you need to make a change permanent, you'd boot a fresh version, make the change, then commit.
But my Windows is mainly for playing games, so my solution is to simply pipe the image back over ssh, then bzcat, then ntfsclone. Or, if I want to try something fun, ntfsclone to lzop to the local Linux partition, restore after the next boot.
I always felt plain text and HTML suffices for any and all communication requirements.
I feel that way now, although I would either severely restrict the HTML (to facilitate WYSIWYG editing), or use ODF.
But I'm curious -- did you actually get that multi-national company to use any open standards, or are they still doing Exchange and Word? I'm going to guess that buying the Office licenses will be cheaper, for many organizations like this, at least in the short term.
Receipts for snow shovels from your Australia office. Never mind that it doesn't snow there.
So, what, they don't write software in China? Or you're arguing that since they probably won't be caught or punished, they should do whatever they want?
Slashdot Mirror
So my system is running a dinosaur?
Most likely.
Im not an computer expert or anything, but copying everything from RAM to HD could hurt performence and stuff.
Well, it's clear you're not an expert.
From what I understand, it doesn't copy all RAM, it asks objects to serialize themselves -- which is pretty much how programs would save state now, only it's done automatically.
Even if it was, how do you think hibernation works?
I think linux is way better then any windows
Irrelevant. He's trying to build a system much better than both of them.
Also his stuff is still vaporware
Maybe. It has been done before, though. Look up Smalltalk.
I would like this (concept) if it where open source...
Well, some of those Smalltalk VMs are... I do agree with you there.
Many devs = Succes and a better(?) linux.
Linux != open source.
Newton OS had the same thing. It caused me to lose data twice when I accidentally deleted a large part of a Newton Works document and then did something else. Undo only undid the something else; the deletion became permanent as soon as it passed out of the one-step undo buffer.
Two things:
First, your problem seems to be more with the fact that undo history was only one level deep than anything else.
And second, you do need revision control, and it needs to be easy enough for the masses, but more powerful than just "undo".
Keep in mind, the whole OS is designed this way, including all programs.
Let me give you an example of what happens when it's implemented as a library: GNOME and KDE sessions. At least in KDE, it's possible to save a session, or even to have it autosave when you logout. It will remember all open programs, and the geometry of their windows. It will even query the programs, asking them to save their state.
Now, this would be awesome, wouldn't it? It'd be a lot more efficient than hibernate/resume, if it worked -- for example, an ODF (plus some simple geometry and state) is much smaller than the entire virtual image of OpenOffice. If the programs were written well, to load only what they need on demand (and thus start much faster), the whole system would shut down and wake faster.
You could even start to have multiple sessions, maybe mapped to virtual desktops, maybe not, so that when you boot, you could choose whether to have it launch your web browser, text editor, and terminals, or have it launch your mail client, IM client, and softphone, or maybe have one that just launches whatever movie you were playing (which would resume from the exact moment it was at when you shut down)...
Problem is, too many programs don't support this. Some, like Firefox, seem to supply their own session management. Some don't even try, and thus, when the DE tries to resume them, it ends up launching a fresh instance. Some can't be persisted, due to their fundamental architecture -- how would you propose to save the state of a running terminal?
So, doing it as a library doesn't work, unless everything's using that library. If everything's using that library, that's pretty much what you get.
And sometimes, you do have to enforce sometimes performance-decreasing features in order to provide a better user experience. Imagine if filesystem access was just a library, and programs had access to the entire disk. It might be interesting to build an OS that way, but even if you did, I imagine you'd want to restrict most user-level programs to dealing with the POSIX API, and being bound by Unix permissions and POSIX ACLs.
Only the people who have past/present guilt about hiding something from someone will feel this is a horrible idea.
So, it's your opinion that stalking should be OK. Got it.
I don't think I really have to say much more than that -- if there are any women on Slashdot who have been raped or abused, they'll likely tear you apart right about now.
Remember, this is Opt-in folks.
Yes, that is true. But that doesn't seem to be your argument...
People in relationships, you all better wake up, your gonna have to be honest by choice or by technology.
People in relationships should also trust each other enough not to have to use things like this.
The fact that you have nothing to hide doesn't in any way imply that you'd never want privacy. Consider the most basic sense -- I'm reasonably secure about my body, but that doesn't mean I'm inviting people to set up a shower cam.
Nor does wanting to hide something imply that there is something shameful to hide. The simplest example there would be a surprise party. The more frightening example is when people really are after you -- witness protection exists for a reason.
Now, if you've opted in, fine. But what you seem to be implying here is that no one should ever have a reason to opt out, and you're flat-out saying that the only reason to ever want to hide is guilt, and that is simply not true.
What happens when they use your stuff anyway? I know that if I came home and found someone had broken in and used my stuff (eg, ate my food, slept in my bed, and walked off with my electronics, I'd call the police.
I think you just answered your own question.
I can do things (like keep the doors locked) to reduce the chances that someone breaks into my place.
And you can do things like keep your phone in your pocket, and not let people play with it. If it gets stolen, then they aren't tracking you, they're tracking the thief.
And how is it Google's fault that people are dumb?
Let's start addressing the real problem -- that of dumb people -- and stop trying to protect them from themselves.
Take another example: Teenagers are always posting shit on Myspace, Livejournal, Facebook, and everything else they can find. They aren't always obscuring their own name, meaning future employers can look them up this way. So is the right solution to "protect" these kids by blocking all access to social networking, or to block employers from using search engines to evaluate potential hires?
I have a better idea -- let's educate the ones we can, and let the rest suffer from their own mistakes.
Assuming you can make the case for doing so -- and it looks as though you have plenty of help there -- it might be worth clarifying what an "open standard" means.
For something to be considered an open standard, it must meet the following criteria:
- A comprehensive formal specification. (This should be obvious.)
- At least one reference implementation for which source code is freely available. (It doesn't have to be freely re-usable, so long as it's there.) OR, many very different implementations which can communicate. (There probably isn't a reference HTML/CSS renderer, but there are enough implementations that one isn't needed.)
- No legal issues for either of the above points, or the use of the specification. (Obvious example: No patents allowed, unless they've been turned over to the public domain.)
It should also meet the following criteria:
- A well-written, accessible, comprehensive formal specification. Or, both a formal specification and easier-to-read documentation.
- Both an official open source reference implementation, and several competing implementations.
- Corporate backing -- especially a corporate stake in it. This implies that said corporation has had their lawyers verify that there are no legal issues.
- Simple, clean design, especially relative to other standards providing the same thing. For example, if the choice is between SOAP and XML-RPC, you probably want XML-RPC -- and you might prefer REST to either of those, especially if your data is not XML.
- Popularity. This really matters the least, so long as the others are met -- it's more important that I can hold the ideals of REST in my head, and implement it from scratch in a few lines of code, than that there are probably more SOAP and XML-RPC implementations. But it shouldn't be ignored -- it would be insane to try to replace HTML with something completely different, for instance. (Both HTML5 and XHTML are incremental improvements, and are sane. Trying to replace HTML with a YAML-based format would not be sane.)
I'm not suggesting that policy has to follow these to the letter, but that's what I personally consider an open standard, and especially, what I consider to be a good standard. In the past, when I've called Microsoft's "Open" XML various names -- "Neither open nor standard" comes to mind -- these are the guidelines I was using.
It doesn't seem to stop bugs from getting into Open Source software.It doesn't seem to stop bugs from getting into Open Source software.
Stop all? No. It would certainly stop some.
But this was also about fixing bugs once they're there.
This idea that with OSS if you find a bug you can patch it is also bullshit for the most part. It assumes that the user is both a competent programmer
If we're talking about a security hole, I'd say that's a fair bet.
Even if they're not, they have options -- they can hire someone else to fix it, if it's important to them. To an individual, that might be unrealistic -- to a large organization doing a security audit, it's essential.
As far as most people are concerned, if they find a bug in OSS, they will just report it to the development organisation in much the same way as they would for proprietary software. At that point, they are in pretty much the same boat. Almost.
Almost. Because at that point, there's still the part where the community generally has the same priorities and the same motivations that they do.
In the rare case where that is not true, there is nothing stopping you (or anyone else) from forking it and attempting to build a community who does have your best interests at heart. I don't mean to imply that this is easy, only that it is possible -- with a proprietary product, you're going to have all the same challenges of gathering a development team and convincing them to work on your fork, with the additional challenge (or impossibility) of getting permission from the original vendor, let alone source code.
With proprietary software, quite often, the company's interests are not at all aligned with the consumer's interests. Often, the consumer is not the same as the customer. And if that's the case, there's nothing you can do about it -- again, forking is much harder, if it's even possible.
The whole development process is open to scrutiny from the customers. This is almost never the case with proprietary software.
I'll agree with that.
Completely shoddy, backwards arguments, too:
any flaws on commercial applications tend to get patched a lot faster than on open source, as the vendors producing the software have a lot more to lose than an open source programmer
This ignores the "many eyes" factor, and the additional effect that anyone who finds a security vulnerability can also patch it, and can inform people of the patch at the same time as the vulnerability. Contrast this to proprietary software, where anyone who does find a breach will also find that the best they can do is report it to the vendor and hope for the best -- and when some of them take many months to be patched, it may be worthwhile for them to start exploiting it, if for no other reason than to get Microsoft to take them seriously.
All of those have been argued to death... Let's assume I'm completely wrong. There's still the fact that there are many corporations which support open source. If an IBM, or a RedHat, or a Canonical ships an insecure product, they have every bit as much to lose as a proprietary vendor -- often moreso, as they tend to have quite a lot more competition.
All of which has very little to do with the supposed counterargument:
We need to move in the direction of what are known as 'open standards' - in effect, creating a common language for government IT. This technical change is crucial because it allows different types of software and systems to work side by side in government.
Microsoft aside, there is plenty of proprietary software that not only supports open standards, but actually revels in them. Unless the argument about security implied that there's an inherent insecurity in ODF itself, I don't see what the relevance is.
However, this article unfortunately presents it as an argument of security against hot new stuff. I don't think anyone is urging the government to become less secure.
But on a modern system how fast most OS features act is the split between milliseconds and who really cares?
Depends how much you're using those features, and the split can be rather large.
For example: Even if you got a decent POSIX system on Windows, I doubt you'd get an efficient fork. On OS X, process creation time is relatively slow. On Linux, fork is very, very fast, uses copy-on-write, and actually makes Unix pipes a sane alternative to threads, under some circumstances.
Now, my information on that is all very old, so it's possible I'm completely wrong, and Windows and OS X had caught up.
The summary is highly misleading "Ubuntu as much as twice as fast!" At extremely short unnoticeable tasks which no human would care to measure except in a benchmark.
Yeah, that's probably true. I haven't actually looked at the benchmark, and I suspect that once you've got 5 gigs of RAM to throw at it, and a fast enough CPU, the bloat starts looking less important.
Bah. Reading the other post, I should stop using the word "throw".
It's got full SDL support, so it should work fine.
Does "full SDL support" include full OpenGL drivers? Can I play Doom 3 on it?
That's not really what Haiku is for. It's a desktop system. You should not use it for any headless tasks.
I consider that to be a detriment. It's nice to be able to deploy software to the same environment I use at home. It's one less thing to think about when I'm wondering why it worked on my development machine, but not the test server.
You're trapped in the X mindset. I can't think of any other windowing systems that behave like that
Which in no way means that it is a bad way to behave.
Or do you mean, you've never seen another windowing system get too screwed up to boot, either from a driver issue, configuration issue, or something's wrong with your desktop session?
Another example: If something has trapped all my input within X -- say it's a game, and it's capturing all keyboard and mouse actions, and has stopped responding -- I can still ctrl+alt+f1 and kill it. Or, worst case, the whole system's hosed, I can ctrl+alt+backspace -- without losing my network connections, or anything non-X I've left running -- and then re-enter my session, only this time, I have the benefit that the entire windowing system might still be in the filesystem's cache.
All of these are real, user-visible advantages to X. How is your system "better" by not supporting these features?
WTF does "boots cleaner" mean?
You really need to try it.
Translation: You can't explain it. Maybe it's a completely subjective feeling.
it won't be long before it quietly glides past Ubuntu in overall ease of use.
If that happens, it'll be worth thinking about, but don't forget the "worse is better" component of ease of use. Linux was here first. All my stuff works on Linux. Even if it is easier to boot, connect to a network, and get a browser, that doesn't matter at all if it doesn't have my favorite editor, or if I have to rewrite Amarok from scratch...
With a few more developers, I think Haiku has far more potential to beat Windows and Mac in desktop usability than linux does.
Linux has already broken both in several important ways. The rest is all about software support and commercial support.
linux is reaching certain limits in the usability and overall "smoothness" department that a nice ground-up redesign would do nicely to correct.
Maybe. However, I think it would be foolish to throw away the majority of open source drivers in order to accomplish that -- at least, not yet.
In commercial development, it's not strange to wipe out your code and rethink your work from the bottom up occasionally. The unix world has been incapable of doing this for decades...
Which "commercial development" are you talking about?
If it's OS X, you might have a point, if they didn't simply take NeXT and develop it into a usable product -- and remember, NeXT is a Unix, and so is OS X.
if it's Windows, well, Vista was a disaster, and that's the closest they've come since Windows 2000... which was taken from NT, and large chunks of the 98/ME line. And that's the closest since Windows 95... which just sucked in tons of stuff from 3.1, both of which ran on top of DOS.
If it's anything else, you're either talking about an entirely embedded system, which implies much less code and compatibility required, or you're talking about a piece of software, and if you truly believe the open source community has never started from scratch on any software, you clearly haven't been paying much attention. The obvious example would be KDE4.
I think I get what you've said here, though:
It's smaller and of a more sane design. It was designed by a professional ke
Haiku is not just a kernel, it's a whole operating system.
Even better! Port it! With the noise you've been making about POSIX, you'd think Haiku would be POSIX-compliant, right?
What is it that the Haiku kernel does that the Linux kernel doesn't? And if there is something, why couldn't the Linux kernel be adapted to do that as well?
All of the things you mention in linux are rather hackish and feel overall awkward.
Are they implemented at all in Haiku?
In Haiku, when I boot plug in a network cable, does the OS notice and automatically try to connect?
Can I connect to an open wireless network in two clicks, the way I can on OS X or Ubuntu? Is my wireless card even supported?
The speed, elegance, or efficiency of a nonworking or nonexistent program is irrelevant.
Linux always feels sort of heavy when compared to systems like Mac or Be that just do exactly what they want to do.
Oh, they do exactly what they want to, sure. And as long as you want to do exactly what Steve Jobs has thought of, they're great!
And as soon as you don't, you're out of luck. Something as simple as sloppy focus will likely never be implemented in OS X, never mind sane keyboard shortcuts.
I think the point is that a redesigned approach might provide similar functionality with reduced resources because there are fewer unecessary layers of abstraction.
Except it doesn't provide similar functionality. From what you've told me, I can no longer play games, use wireless, or watch HD videos, despite that Be was supposed to be a multimedia operating system.
Removing the networking aspect from the window manager and mode switching with the kernel is a simplification.
I'll agree with you there, only because the network design of X seems to have been broken to begin with and has only gotten worse over the years. I used to enjoy being able to run programs remotely using ssh's X-forwarding, but lately, the various X11 environments seem to crash when connecting to a remote system that isn't identical to their own.
Regardless, the networking aspect is a Good Idea, just poorly implemented. Why shouldn't I expect to be able to run any program, anywhere, as if it were local to my desktop?
The driver system is simplified
If by "simplified" you mean "missing major drivers", OK.
it's not designed to run headlessly so it runs natively with a GUI.
Ew. This is one of the things I always hated about Windows Server -- no debugging over a serial console, and you need a video card in the machine, even if it's never used. No thanks.
And on a desktop, you've now set yourself up such that, if the video driver ever has a problem, you're hosed. I can at least boot to text-mode and try to fix the problem -- depending on how serious it is, I could try various configurations, launching an X server, killing it, launching it again...
Serviceability is also good stuff for a desktop.
I'll grant you, mode-switching in the kernel is good. Video drivers really belong in the kernel, for performance reasons, as long as we have a monolithic kernel approach.
But what reason could you possibly have for requiring it? What do you gain by not being able to remove it, like any other kernel module? Like, I don't know, filesystems, sound drivers, RAID controller drivers? Easily most of what a Linux system needs at boot can be swapped out -- why is that a bad thing?
It boots faster
What are you comparing? There are versions of Linux which boot in under a second. Please be specific.
it boots cleaner
WTF does "boots cleaner" mean?
its overall simpler for the end user
Until they want to use wireless.
In theory- it's
I don't think you understand- the point is that Haiku is designed from the ground up for the desktop based on BeOS's example.
And Linux was designed from the ground up to be a terminal emulator. Neither are being used for what they were originally designed for.
The point is that it's not linux
If that's your only point, there are lots of things that aren't Linux.
it's designed for the desktop.
So is Linux, for that matter. At the time, desktops needed terminal emulators.
The companies supporting linux are doing it for the server and HPC--
And you somehow think they control the direction of kernel development.
What's with this Katamari Damacy attitude in open source? I thought F/OSS was supposed to be more flexible.
And Linux is. It can be in the server, and in HPC, and on the desktop. There are many things other than kernels which contribute to that.
A trivial example: There's no reason the same networking system, at the kernel level, can't work for both server and desktop. Certainly the same firewall will be useful on both. But the Debian networking system is more useful for servers (and the loopback interface), whereas NetworkManager is more useful for desktops (and especially laptops).
The kernel really isn't a great place to differentiate these. The kernel really should just be a giant device driver -- which is the point. Does Haiku support my wireless card? Can it provide accelerated OpenGL on nvidia hardware? Since Be was so good at video, once upon a time, how well does mplayer work on Haiku, especially with, say, hardware-accelerated h.264 playback?
It's becoming increasingly clear that X11 will never be hacked into a usable local display option.
Not as long as nvidia is hacking it, no.
The open source community badly needs something more desktop centric.
Show me an open option. Until then, well, certain Intel cards seem to do better with Compiz -- mostly, I would guess, because of more open drivers.
Or, you know, you can turn off desktop effects, and let it use technology that was boring in 1995 to give you a smooth desktop experience.
I wish more developers would pay attention to Haiku for the desktop so we can have a POSIX compliant free desktop operating system that is built with the desktop in mind.
Far too late for that -- the Linux kernel supports entirely too much stuff, most of it very well. And are there any good nvidia drivers, open or not, for Haiku?
If you really want to make that succeed, start trying to port the more interesting Haiku features to Linux.
Neither runs the other's software (without proper tools, and even then it's not nearly perfect),
Close enough, though. Aside from Wine running similar tools on both platforms -- and, when Wine works, it's often faster than running the same app on Windows -- there's plenty of cross-platform development.
Let me put it this way: Suppose I'm a Java developer with Eclipse. That'll run fine on any platform I throw at it. But, even before I get to my own software, Eclipse is such a hog that I'll want every ounce of performance I can throw at it.
For that matter, if I'm developing a Java program -- or Ruby, or Python, or anything else sufficiently cross-platform -- I may well care when it gets to deployment time which OS is faster. If developing a new app, I may choose to support one platform over another for performance reasons.
It's probably not as useful as benchmarks within an OS (between Linux filesystems, say) or between POSIX-compliant OSes (but these benchmarks don't test the Windows POSIX layer, I'm sure), but it's worth mentioning.
Oh, and I like being amaturish -- a big HA HA to everyone who tried to convince me that Vista is fast, when you give it enough RAM.
I have no problems with corporations using P2P to cut costs.
Where I have a problem is when they also want to revoke rights.
If they just used BitTorrent to deliver streams, no problem, I applaud that. Maybe multicast would be better for some streams, but whatever. If it's really a problem, I can limit my upload, or block them from uploading at all.
However, when their EULA prevents me from reading my own logs (much less any of the above shaping), that's when I have a problem.
Granted, they're going to have to change something or they'll _eventually_ fall behind.
Change what, though?
If I were Microsoft, I would look first to changing internal development practices.
I think the point is, rather, that there's plenty of old code they could release now, as id has done with old games like Doom (1, 2) and Quake (up to 3)
Only less useful, if it's actually a slot.
Take out, plug into slot, and possibly configure a bluetooth device...
Versus, be in range of a wireless access point, then point and shoot.
What makes it slightly scary is that it claims to be a parking violation.
However, I would likely make a very loud noise about being required to not only have Internet, but also a specific browser and a specific operating system, and having to download their software.
For unemployment, at least here, the entire thing is done over the Internet. However, the website pretty much works in any browser (though the layout was slightly off in Konqueror), and if you don't have Internet (or a computer), you walk to the unemployment office, they sit you down at one of their computers, and you do it there.
For a parking violation to be so unaccessible has got to be violating some regulation somewhere.
Just goes to show that no matter how much protection you have on the tech side, there's always a social engineering way around it.
Also goes to show how a little paranoia goes a long way.
This always confuses me. It's legal to download DDLs at random?
My assumption has always been that since I do have a valid Windows license, it would be splitting hairs to try to find the exact version that's in my Windows (when I could probably download a version known to work with that combination of app / wine version).
However, this seems to happen less and less as time goes by. Used to be, you'd need IE or some IE DLLs. Now, most things will work just fine with the Gecko shim.
Ok. I always use a "test" user for downloaded binaries (other than the OS updates.) For example, to avoid compiling google-earth, just downloaded the binary and always use it on the "test user", so malware infection danger is minimized to that dummy account.
I tend to create one per piece of software, when I bother. If it's one "test" user, that could start to get like nobody -- because you run so many things under the same user, any one of them could blow up more than it should.
As an example, I got tired of waiting for Ubuntu to update Rubygems, and the Rubygems update itself didn't work well. Ruby 1.9 was always a bit behind, too. So I downloaded Rubygems and Ruby 1.9, gave them each a user, and had them install to their home directory. Now, if I need to completely wipe it and start over, I won't be picking pieces of Rubygems out of every corner of my system -- I just nuke /home/rubygems, and start over.
Since Wine is implemented in user mode, I think that that prevents me from liberal access to my partitions.
I see -- but, this was a response to you being afraid of installing Vista again. I wouldn't be afraid to have Vista or XP on this laptop, if it's properly backed up. If it's an emergency, I can pull out the SATA disk and put it in the fileserver where the backups are. Your "two full days" don't seem realistic, with good backups.
What I'd be more afraid of is a deliberate, specialized attack -- something which finds my home partition and grabs my ssh keys, for example. Ok, those can be encrypted, but realistically, I should be doing full-disk encryption to protect against that...
That's also the reason I hate IBM when forces me to run their installers as root (I think in order to install some secret-shitty-license-key-file)
I wonder if those will run chroot'ed? Or in a VM? Debian's "fakeroot" might also help you here.
KVM isn't going to be as nice as the bare metal, and an ntfsclone image is going to be easier to handle than a raw (even sparse) image.
No, the nice part about something like KVM or Qemu is when you run it in snapshot mode -- all changes are written, copy-on-write fashion, to a temporary file. If you need to make a change permanent, you'd boot a fresh version, make the change, then commit.
But my Windows is mainly for playing games, so my solution is to simply pipe the image back over ssh, then bzcat, then ntfsclone. Or, if I want to try something fun, ntfsclone to lzop to the local Linux partition, restore after the next boot.
Cashier has no honor!
I always felt plain text and HTML suffices for any and all communication requirements.
I feel that way now, although I would either severely restrict the HTML (to facilitate WYSIWYG editing), or use ODF.
But I'm curious -- did you actually get that multi-national company to use any open standards, or are they still doing Exchange and Word? I'm going to guess that buying the Office licenses will be cheaper, for many organizations like this, at least in the short term.
Receipts for snow shovels from your Australia office. Never mind that it doesn't snow there.
So, what, they don't write software in China? Or you're arguing that since they probably won't be caught or punished, they should do whatever they want?