Submitted by Using CoLinux, the Xming X server, and PulseAudio for Windows, the boys and girls behind the Portable Ubuntu for Windows project have turned the entire Ubuntu distribution into a executable that you can run within Windows. You can see some screenshots, and download it from their web page.
Windows already has a perfectly good audio API and phonon (and SDL) has a backend for that. Why would anyone want to port PulseAudio to Windows? Who is such a masochist?
Perhaps its easier to port one known smaller library to run under windows than rewrite every single app that has ever used pulseaudio to use windows?
Maybe, just maybe because all other ubuntu apps use pulseaudio and not any windows api? These are linux apps remember?
The word you are looking for is “sadist”, not “masochist”. A masochist enjoys having pain inflicted upon himself. A sadist enjoys inflicting pain upon others.
That said… when has anything relating to PulseAudio ever made any logical sense? Writing it in the first place was every bit as illogical as a port to any platform.
Edited 2009-04-04 17:21 UTC
That is for freaking sure. I hate PulseAudio with the fire of a thousand suns. What a piece of crap. Ubuntu Gutsy and above has been hellish for me on my laptop because of that piece of crap PulseAudio.
I am not sure what I hate more, not having audio in Flash, or the fact that occasionally after coming back from sleep, the Gnome Power Manager decides to eat all my RAM because of some PulseAudio bug.
What is weird is that I have no such problems on my Dell desktop.
People think that they are stuck with it. But fortunately, it’s not like some things that so intertwine themselves into your system that they are not surgically operable:
# yum remove pulseaudio
or
$ sudo apt-get remove pulseaudio
work like a charm.
End your suffering now! Free yourself from the malignant pulseaudio and enjoy life again!
Edited 2009-04-04 19:00 UTC
Or, in Ubuntu 9.04 and above, if you want to keep all your ubuntu-desktop packages in tact, create the empty file:
.pulse_a11y_nostart
in your home folder to disable it.
I think I am still on 8.10LTS, so this may not work.
But the fact that they added it is nice in that they have noticed how many people have had trouble with PulseAudio.
On a completely unrelated note, for those with a dual boot Windows/Linux machine, I cannot recommend enough the Windows ext2 driver, which allows you to read/write a ext2/ext3 partition from Windows (though just to be safe I put it on read-only mode.) It is here:
http://www.fs-driver.org/
I have to reboot to Windows to watch any videos on my TV through my laptop because the X in Gutsy will no longer do the dual desktop setup that worked great in Feisty. Also the current video drivers for my ATI on-board suck and playing movies uses a ton of CPU. Of course all this worked fine on Feisty. Grumble, grumble…..
Of course running Windows lets me run my favorite browser Chrome, so it isn’t all bad.
You, sir, have to look up what a11y stands for and stop spreading BS.
Well, I know what a11y is (in fact I use various a11y systems) and I don’t see how he was spreading BS. Perhaps you, sir, need to get control of your outburst?
You sir just may have made my day. I will definitely try this. I had always thought it might be an option, but figured there was some giant dependency to all the audio apps.
Actually, darknexus’ response from a minute or two ago might be a better alternative. I did not know about that solution.
Don’t feel bad. I only know about it because it was originally added in as a fix for some accessibility-related packages that were recompiled and don’t work with Pulseaudio out of the box on Ubuntu anymore. Ridiculous, you’d think they could at least keep consistency, at least before it worked either completely with Pulse or without it. Anyway, I had to look at the /usr/bin/pulse-session script to even find out about that file, because I wanted to find out how it had been disabled so I could re-enable it. Naturally, this wasn’t documented.
Personally, methinks they should add this in as a checkbox in the sound preferences to enable or disable Pulseaudio according to the user’s preference–easy enough to do, now that they’ve added this quick way to disable and enable it. Other Ubuntu packages do not depend on it running, after all, except for the Pulseaudio panel applets which are easy enough to flip with non-Pulse ones if you wish–hell the setting could even do that too.
Well, I know you hate Pulseaudio, but for me it’s the only thing in Linux that allows my Logitech USB wireless headset to work properly. Straight ALSA doesn’t resample properly, rendering anything below 44.1khz as barely distinct static, and my wired USB headset won’t even give me that without Pulseaudio. If nothing else, it’s useful for that, as the kernel devs don’t want ALSA itself to have a proper resampler in kernel space for some reason, even though it’s necessary for a good number of audio interfaces. If nothing else, Pulseaudio resamples better than ALSA itself, though granted at the price of a few CPU cycles.
As I have indicated before, cancer patients benefit from chemotherapy and radiation treatments. But that does not mean that we should all be subjected to prophylactic chemo and radiation therapies.
Edited 2009-04-04 19:08 UTC
Ok, but to take your analogy further, that does not mean these things should not exist for those who need them.
That is completely reasonable.
Interesting then that you insist that Pulseaudio needs to be removed from the Linux audio ecosystem, if it’s reasonable that some of us need it. I don’t think Pulse is the optimal solution, unfortunately it’s the best we’ve got and, as long as the ALSA team remains strict on the issue of a full software resampler, probably the best we’re going to get at least, barring a major restructure/rethinking of ALSA or moving to OSS 4 (even less likely). Best, then, that Pulse be developed further and mature, at least imho. The optimal solution, from my point of view, would be for ALSA to include a full-featured software resampler that would be loaded for drivers that need it, this is what OSS 4 does and does it quite well. That stuff belongs in the sound driver space, not in user-space, and shouldn’t need a user-space sound server to wrap around the problem instead of fixing it.
I don’t think I have ever said that. Put it in Universe, or whatever repo your distro configures by default. Or, better yet, check for bad hardware during OS intallation and install and configure it right then and there if it is needed. (Along with a warning to the user that their hardware is broken and that their audio may be unreliable due to instability in the software that tries to correct for it.) But don’t inflict it upon everyone. That is what I am saying. And that is the message that you misrepresent when you claim that I am insisting that “Pulseaudio needs to be removed from the Linux audio ecosystem”.
Edited 2009-04-04 21:10 UTC
Uh, Pulseaudio has already been ported to Windows. They didn’t need to do anything for that. As for why… well, it’s default in Ubuntu, so it makes sense if they’re porting Ubuntu doesn’t it? Remember, Linux apps couldn’t access the Windows audio API, and any type of ALSA to Windows bridge would’ve taken more effort than using the Windows version of Pulseaudio. And as for Fonon… how many Ubuntu (not Kubuntu) apps would be using that? Or in other words… Fonon is KDE, Ubuntu is GNOME, so it wouldn’t make sense. SDL makes a bit of sense, but it’s not used by a good number of apps at least not as default.
If there are so many well know problems with PulseAudio why aren’t there more efforts being made to improve how it works .. better yet why aren’t the ALSA people working with the PulseAudio people to make things better? That is one of the things I still ask about the OSS vs ALSA situation. But please respond to my PulseAudio question .. it seems to me to have alot of potential .. same goes for Phonon is there any effort being made by these competing technologies to interoperate? I mean they all do something good but right next to that is the fact that they all seem bent on muscling each out.
Because it’s a Fedora project. Now that RHEL6 and RHEL6 Desktop are on the horizon, we should finally see some work put into stability issues. Red Hat has probably already issued the decree through whatever private channels they have to the PA devs.
Edited 2009-04-05 15:50 UTC
Bzzt. That’s a bad sign, since either the developer doesn’t intend to post it (I don’t think anyone would be that silly, but who knows), or the developer doesn’t have it in a releasable state. In the latter case, that means no automated build, no testing, no nothing, which implies that the project is nowhewre near usable.
There’s an awful lot of assumptions based on one comment there.
This sounds like the worse of two worlds. You get all of Windows’ security problems and then get to run some crappy Ubuntu applications on it (especially if they need that piece of crap PulseAudio.)
By the way, I was a big Ubuntu fan until Gutsy, when all my problems started. PulseAudio is a big part of the problem, but I have also had tons of problems with X and my video setup. I now hate Ubuntu and only continue to use because of the trouble required to put another Linux, and of course the fact that there are 10^999 distro options these days. Though if anyone here has recommendations I am all ears.
Hopefully soon Haiku can become my primary OS.
You should never, ever, ever, willingly run an Operating System you hate. Your hair will turn grey, fall out, you’ll get the measles and snivels, your children will grow up trying to become dictators, and your wife will leave for the man who’s boots have been under your bed.
Okay, seriously, it’s not hard to try another distribution though. Just download and burn some CD’s, and shoot em for a spin. I would try Mandriva, OpenSuse, Fedora, Debian, CentOS, PCLinuxOS, or just randomly search and see what tickles your fancy.
I can’t believe that this has not been the talk of OSNews:
http://unetbootin.sourceforge.net/
Burning CDs is sooooo 1990s.
As much as USB flash drives have dropped in price, CDs are still much cheaper.
CD-RWs yield about 900MB per dollar. USB sticks yield about 500MB per dollar. But CD-RWs are so slow, low capacity, and so unreliable that I went back to CD-Rs because trying to actually reuse CD-RWs was just not worth the hassle.
And USB flash is compact. Is that a 25 pack of CD-RWs in your pocket? Or are you just happy to see me?
USB flash rules! 🙂
Edited 2009-04-05 17:35 UTC
But USB flash doesn’t boot on everyone’s system. I use Linux on an older desktop that does not have support for USB flash booting that is necessary for such installs. I might could flash the BIOS to support it but burning CD’s as rarely as I do works fine.
Also, the suitability for someone downloading a large number of OSes / distros for testing and trials, as suggested. The cost of burning twenty CDs compared to the cost of twenty USB drives…
One huge advantage of a USB-based distro, the ability to save user files and hardware configuration settings. That’s not necessary for testing and trials though.
Of course, booting from CD doesn’t work on everyone’s system either, if you have a netbook or other small laptop that doesn’t have a CD drive. Boot from flash is much easier than finding someone to borrow an external drive from.
Which would be why both USB Flash and CD media remain good choices for installing operating systems. One will usually work where the other won’t.
if they used their creativity to improve Linux/BSD or Solaris. Windows is dead all ready. Have you also heard about Debian kFreeBSD or Belenix?
Edited 2009-04-04 19:34 UTC
I have been using andLinux on an XP system at work for about a year. It is basically the same idea as portableubuntu:
andLinux uses coLinux as its core which is confusing for many people. coLinux is a port of the Linux kernel to Windows. Although this technology is a bit like running Linux in a virtual machine, coLinux differs itself by being more of a merger of Windows and the Linux kernel and not an emulated PC, making it more efficient. Xming is used as X server and PulseAudio as sound server.
They offer Ubuntu with XFce or KDE, I was running the lighter XFce version. They also warn that you might have problems with openGL but it worked fine for me with Ghemical which is an openGL based app and the prime reason I was running andLinux.
There is absolutely nothing wrong with Pulseaudio. Ubuntu developers didn’t implement it properly but there is a HOWTO on the Ubuntu Forums of how to fix Ubuntu’s implementation of it.
Since following this HOWTO, I’ve had absolutely no problems with Pulseaudio, and more importantly, absolutely no problems with OSS and ALSA programs. That’s more than I can say for the pre-PA days when you had to turn ESD off in order to run Audacity.
Pulseaudio itself is good – the volume control for individual programs is really handy. Maybe people should stop whinging and just try using it?
There is absolutely nothing wrong with Pulseaudio.
On a good distro it works as intended, atleast for me. Haven’t had any issues with it so far. And atleast on Mandriva you can just enable/disable Pulseaudio from the configuration utility, or per-user by choosing “Pulseaudio Preferences” from the menu.
As for the usability of Pulseaudio..I sure can understand the reasoning behind it, not all sound hardware can do hardware mixing of streams, and Pulseaudio allows for unlimited number of streams to be mixed whereas hardware always have some limitation.
Personally though, I have absolutely no benefit from using Pulseaudio, but I have no issues stemming from it either. I don’t need or even want to control the volume on a per-application basis and my hardware can handle just fine mixing the few audio streams I might have.
This pulseaudio stuff is exactly the problem with linux! Look at all the arguments/debate it’s caused in this one comments area… What on earth is all this needed for anyway? Alsa, OSS, Phonon (or whatever it’s called, I don’t follow linux very much for this very reason), and Pulseaudio????
It’s stupid, look how easy this is in Windows – install your sound driver, and oh hang on a minute, is that it? Hell in Vista or Windows 7, most people won’t even have to do that anymore it’d just work out of the box… No weird levels of audio stack, just a working sound card with a speaker icon next to a clock that lets you adjust volumes… As long as there’s no driver issues, everything else which requires audio simply plays sound. Everything is standardised so all higher levels apps no matter what you want to use will just play audio.
How many different api’s and sound modules do you need? Why should you uninstall or reconfigure various parts of the audio stack? I agree with a previous comment about just getting them all interacting and playing nice together – all this choice/competition/a million different ways to do the same damn thing is not what people want when they want to do something as simple as play an mp3 or listen to audio on youtube etc etc…..
I don’t think it is _quite_ as complicated as it sounds. Really, the “official” sound api is ALSA, as this is supported in the kernel. Older apps may use OSS, so ALSA also provides a “backward-compatability” mode for OSS, so apps shouldn’t have to worry. As far as I can tell, PulseAudio, etc are frameworks on top of lower level frameworks like ALSA. If I am wrong, someone can correct me. This reminds me of graphics frameworks in Windows. There is GDI, DirectX, WPF, etc. Sure, your graphics card may work with your driver, but your driver must support GDI and DirectX. I think WPF is built on top of GDI, but I am not sure. People run into this problem when a game requires DirectX 10 but the card only has a DirectX 9.0c driver. Even sound has layers under Windows. Is the app using Win32 calls, MFC calls, DirectX calls, .NET calls, etc. There are more frameworks in Windows than you the end user may realize.
Actually installing xubuntu was easier than windows.
Things have improved considerably. My videocard and audiocard work out of the box. For windows I have to hunt the MFG site to find a driver.
(It’s not perfect yet there are still some issues with OpenGL and X11 should have been abandoned in the 1960’s but in the end it works)