Skype For Linux To Be Open-Sourced "In the Nearest Future" 175
rysiek writes "Seems like there might be a revolution in the works, as far as VoIP software for Linux is concerned. After mailing Skype support about Skype providing Mandriva RPM packages, Olivier Faurax got an answer which suggests that the Linux Skype client will be open-sourced. After asking for verification of whether that was the case, the tech support answer claimed it is going to happen, and that it's supposed to happen 'in the nearest future.' Now, this probably only means the client (the underlying protocol will probably be handled by a binary-only library), but even if that's the case, it seems like there is still reason to celebrate."
Yay (Score:4, Funny)
Re:Yay (Score:4, Informative)
You can do that without Skype's source code. There are plenty of SIP clients out there and in fact an entire PBX system for Linux that includes the ability to war dial and use text to speech scripts on calls already.
I'd almost wager someone has already written an asterisk script to order for them.
Re: (Score:2)
Re: (Score:3, Interesting)
Asterisk supports Skype [digium.com]. As does FreeSWITCH [freeswitch.org].
Re: (Score:2)
From your link: "The SFA product will be the only solution that integrates Asterisk directly with Skype." It costs something like $80, and is closed-source. With an open-source Skype client, I doubt Digium's closed-source solution will remain the only solution for long.
Re: (Score:3, Informative)
Re: (Score:2)
It is not opensource, until... (Score:3, Insightful)
The source is not open, until I can build and use it on FreeBSD/amd64 or some other "exotic" platform like that...
Interestingly, the oft-criticized Java has always been more "open-sourced" (even before going GPL), than what the excited write-up is preparing to "celebrate"... Must all be about managing expectations
GUI Code Only (Score:5, Insightful)
Not quite open source then, but I guess it's better than the situation right now. Still no way of ensuring there are no backdoors in the encryption though.
Re:GUI Code Only (Score:5, Insightful)
Re: (Score:2, Insightful)
You're assuming you can feed anything into their comms layer. I suspect there'll be a heavy duty validation / sanity checking at that point already.
Re: (Score:2)
Re:GUI Code Only (Score:4, Insightful)
If you have the client code, you can pre-encrypt before the communications layer if you need the added security.
*Could* that be possible, you would lose interoperability with windows clients, so why not relying on one of the truly foss voip projects availble?
Re:GUI Code Only (Score:4, Insightful)
Because a truly foss voip project requires a server or open ports on at least 1 side.
Skype requires only 2 clients that speak the same protocol, the skype network handles the rest.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
"but you need to enter the address of your mail server into a mail client and end users seem to be able to manage that..."
Really? That is news to me. Last I checked end users need a disc from the ISP to do that, or they call their tech, or in a real pinch their teenager.
Re: (Score:2)
some penniless foss project isn't going to provide servers which can cope with 18,989,413 clients at the same time (number plucked from Skype right now, the peak is probably higher).
It is unlikely that a single project would need to support that many clients at the same time since SIP is a standard protocol which allows servers from many different service providers to interoperate, so the load is spread across many different providers. Additionally, you seem to be limiting yourself to "peniless FOSS projects" but there are a large number of commercial SIP servers out there who are supported by money made from gatewaying calls to/from the PSTN, etc.
By your argument, none of the existin
Re: (Score:3, Insightful)
Because a truly foss voip project requires a server or open ports on at least 1 side.
Plenty of free public SIP servers on the internet...
Skype requires only 2 clients that speak the same protocol, the skype network handles the rest.
I'm unclear on why you think that relying on the existence of a single proprietary network is better than relying on a SIP server (which may or may not be operated by yourself and you can switch to a different independent server if you want).
Re:GUI Code Only (Score:5, Interesting)
*Could* that be possible, you would lose interoperability with windows clients, so why not relying on one of the truly foss voip projects availble?
I recently took my Warcraft guild down a voice-comm path from Ventrilo to Mumble. Mumble had a lot of things going for it - open source, penny-for-a-year server deal, and a much better experience overall. Everyone could hear everyone and the software worked great without a hitch. That being said, as of today we are back on Vent.
See, while it worked great for us, no one else had ever heard of it, and no one would switch to it just for the benefit of talking to us. In the end we found Mumble made our lives more complex and in sum-total was not a better choice than Ventrilo.
Vis-à-vis Skype - yes you would need a special client to handle encryption, but a simple checkbox could re-enable traffic with those Windows clients. What truly foss voip project inter-operates with those?
Re: (Score:2)
My experience was a bit different.
We were using Ventrilo, but one person had a really crackly mic, so I googled for alternatives and suggested Mumble. We tried it, and the guy was instantly clearer. Nobody had any problems hearing him.
Too bad the crackly mic syndrome was passed on to me. On two computers, Mumble is completely useless - it's so bad they can barely hear me.
We're back in Vent, because even with his crackly mic, he's more understandable than I was. :P
Re: (Score:2)
What would this do to their codecs, which are presumably optimised for regular voice/video? Would it increase the bandwidth requirements?
Re:GUI Code Only (Score:5, Insightful)
It wouldn't work at all. Nearly all voip, and I'm sure Skype is no exception, uses lossy compression for the audio. If you stuff encrypted data in, you'll just get garbage out.
Re: (Score:2)
Sounds like a job for
!opensource
Re: (Score:2)
I doubt that. If the binary part is the codec+protocol, good luck trying to stuff anything except audio (and preferably voice then) through it. Lossy compression works poorly for encrypted data.
Re: (Score:3, Interesting)
Re: (Score:2)
Pretty much, as long as you could change the front-end to use some other VOIP protocol other than Skype. In fact, what I'd like to see is something like Pidgin for VOIP. Transparent support for many protocols. Then you can just have your contact list of friends and call them, regardless of what they use.
Re: (Score:3, Interesting)
sounds a bit like how the N900's phone and messaging system will work.
Re: (Score:2)
Afaik Pidgin uses plugins, so if the blackbox file was compiled properly (i.e. of 64 bit with -fPIC) one could create a plugin that is loaded at run-time. Important is what you need to compile and link it, libpurple is LGPL, so no problem, but pidgin.h in pidgin-dev states its GPL, so if that's needed, then it can not be distributed.
Re: (Score:2)
Okay, it's not really open source. But it does allow a whole range of new applications. For instance: sipskype bridges (for asterisk or any other pbx).
Re: (Score:2)
There is already a Skype API for (gui) programs to use. It operates on text commands.
For example, you can already integrate Skype into Pidgin: http://code.google.com/p/skype4pidgin/ [google.com]
And with some pidgin plugin for encryption (both on the Windows and Linux side, e.g. OTR or PGP), you can have private (text) conversations.
Re: (Score:2)
No, a worse situation. More people will become locked into a proprietary protocol, and ...
WTF is "the nearest future"? (Score:3, Interesting)
I'm trying to grasp what could possibly be the "nearest future"? A picosecond from now? But of course, you could have half a picosecond, and half that, and half that, etc.
Re:WTF is "the nearest future"? (Score:5, Funny)
Re: (Score:2)
Keep halving it and eventually you'll be in the present.
Re:WTF is "the nearest future"? (Score:4, Insightful)
Yes, but it'll take infinitely long to get there...
Re:WTF is "the nearest future"? (Score:4, Funny)
Actually, it's impossible to be anywhere other than the present.
Re: (Score:3, Funny)
Free candy tomorrow.
Re: (Score:2)
Would you mind telling my head that? It keeps leaping back and forth and sometimes warps it altogether.
Re: (Score:2)
Yeah, but *whose* present?
Re: (Score:2)
I suspect YOU know that *I* knew that, but still, I'm disappointed that you got modded insightful and I didn't get a "funny" :(
Re: (Score:2)
Yes; well, the mods appear to be hard to please today. I didn't get anything for this comment [slashdot.org], either.
Re: (Score:2)
lim future(t)
t->now+
This could be incredible... (Score:5, Interesting)
I use a lot of voice software on my laptop, and Skype is one of the few that is fine with my not using a headset. I'm not certain how it does it, but I assume they're filtering the sound coming out of the speakers against the mic input. I've always wanted to take Skype's client and plug it into, say, Ventrilo.
Could this open up that possibility?
Re: (Score:2)
I'm not certain how it does it, but I assume they're filtering the sound coming out of the speakers against the mic input.
Shifting windows. The software waits for a unique and easy to identify sound or frequency to be played, then tries to detect it. It checks how long it took for it to be picked up, then establishes a rough time for when the filtering has to be done, which is usually accurate within 1ms.
The filtering is beyond me - but I'm surprised programs can get the first bit so wrong. Measuring audio "ping" is easy - and if you don't immediately discard your played back audio from the buffer, surely some smart coder or m
Seems largely pointless. (Score:5, Insightful)
Obviously, this is exactly why Skype would be OSSing the GUI and not the protocol binary blob; but it is also why the news isn't of much interest. As long as basically all the program's important functions depend on a binary blob you can't see what it is doing, you can't port it to other architectures, you are really no better off than if the whole thing were binary.
IOW (Score:3, Informative)
Re: (Score:3, Informative)
Yes it is OSS. It's not GPL, but an open source frontend with the right license would still be OSS.
Re: (Score:2)
So then the Nvidia drivers are OSS right? I mean they ship with some open source parts.
FAIL
Re:IOW (Score:4, Informative)
The open source parts are open source (shocking!), just like an open source frontend on top of closed source libraries still is an open source frontend.
Is the full driver then open source, no. Is full Skype open source, no. But is the driver glue open source, yes. Is the frontend open source, yes.
Re: (Score:2)
Re: (Score:3, Informative)
Yes it is OSS. It's not GPL, but an open source frontend with the right license would still be OSS.
If the underlying driver isn't also GPL'd, then it's not open-source.
And as long as we don't have access to that underlying driver, we have no way to guarantee that there's no backdoor into our communications.
Of course, we already know that the Austrian interior ministry has confirmed it has no problem [h-online.com] listening to Skype conversations [theregister.co.uk].
If Austria can do it, it seems likely that other governments have that capability (even if they claim otherwise despite documentation to the contrary [wikileaks.org]).
Re: (Score:2)
Re: (Score:2, Interesting)
I agree that this isn't news, but I don't think it's pointless.
With the binary blob being available as library (well, it is speculated anyway), one can VoIP-enable one's own applications (whether it's an IM cilent or tech support tool) with a piece of code that is tested and is known to work well under all kinds of different configurations. A lot of the free VoIP out there isn't quite up to snuff, and requires a lot of end-user mucking around to get to work.
Re: (Score:2)
one can VoIP-enable one's own applications
You can already do this by using an industry standard VoIP protocol. There doesn't really seem to be much benefit in using an undocumented proprietary protocol instead.
A lot of the free VoIP out there isn't quite up to snuff, and requires a lot of end-user mucking around to get to work.
If by "low end mucking about" you mean typing in your chosen service provider's address, like everyone's done with email clients since the dawn of time. Hell, if you enjoy being locked into a single vendor there are a number of service providers that ship locked down SIP clients so you don't even need to do this.
Re: (Score:2, Insightful)
Re: (Score:2)
Re: (Score:2)
Yes you are. If you think like me: Wrap the whole thing in a server. Don't even think about writing a GUI for it. Write a SIP wrapper. Then put that gateway on one single server, and tell Skype to go fuck themselves. ^^
Re: (Score:2)
It's neither completely perfect nor completely worthless.
If they're actually putting the protocol implementation out as a binary-only library, and encouraging open source development on top of that, this enables some freedoms without enabling all freedoms.
For example, this makes it possible for me to write a program that monitors a twitter feed, looks for certain keywords in it, and when it finds them, calls me via Skype and uses my own text-to-speech code to read the message containing the keyword out loud
Re: (Score:2)
If they're actually putting the protocol implementation out as a binary-only library, and encouraging open source development on top of that, this enables some freedoms without enabling all freedoms.
Sounds quite dangerous to me - you're needlessly giving a third party the option to pull the plug on your project if they desire whilst also integrating all their security holes, backdoors and downright bad programming into your software.
We could build an IVR Wikipedia tool for the blind with this.
But why would you want to use a proprietary protocol with limited software and hardware support to do this? Seems more sensible to use an industry standard protocol.
So: some freedoms, yes, but certainly not all the freedoms folks might want.
I don't see this as giving people much freedom. I see it as giving people just enough rope to accidentally
Re: (Score:2)
But why would you want to use a proprietary protocol with limited software and hardware support to do this? Seems more sensible to use an industry standard protocol.
I don't have enough information at my fingertips to respond to this. Are there any industry standard protocols for this that are already as widely deployed today as Skype? That are already being used by as many customers in as many settings? If so, sure. If not, that answer might provide the answer to your own question above.
I do know that whenever I hear about people doing general-purpose IP telephony, I hear about exactly two things: either Skype, or some service where the fact that it's IP-based is i
Re: (Score:2)
I would love a UI that makes sense, on Win32 and Mac OS X.
The audio support is ghastly (Score:2)
Skype is a huge source of pain when people discuss Linux audio. It's possible to get it working (mainly by pretending pulseaudio doesn't exist), but so many people have so many problems with it that there's still a widespread belief that it's completely broken. If only the client-facing parts get open sourced, the audio interface ought to be part of that. The protocol, as far as I can tell, works pretty damn good (despite it occasionally telling me someone's offline when I know they're not). I'm fine wi
Good riddance, crappy ugly Skype client (Score:2)
Hopefully this means that libpurple, telepathy et al will be able to make Skype calls.
Once you get users out of a proprietary client, it's that much easier to transition them to a more open equivalent.
"You mean I just get a SIP account and calls cost less than with Skype?" Sold!
Re: (Score:2)
Hopefully this means that libpurple, telepathy et al will be able to make Skype calls.
That's pretty much it - this is the only real advantage to be derived from this. Not Skype somehow becoming "open", but that Pidgin, Ekiga etc can all support it as yet another protocol.
I wonder also... even if they release it as a Linux libskype.so, I can't imagine it being very strongly tied to Linux. I mean, how many system APIs would such a thing really need? Mostly just networking... and otherwise it's just x86 (and hopefully also amd64) code. In that case, it could probably be wrapped into a loader an
Re: (Score:3, Informative)
The problem with transitioning of course is when they ask "how do I call my Skype buddies with SIP?"
It works for the calling-landlines case. For everyone else, there are SIP->Skype gateways like Gizmo5's OpenSky.
protocol will probably be ... binary-only (Score:5, Insightful)
Re: (Score:2, Interesting)
So, assuming the OP is right, they're basically open-sourcing a telephone where the only thing you can change is where the numbers are placed and what the handset looks like. Maybe I'm missing the point, but how does this benefit anyone?
Well, that's not a bad analogy at all... to answer your question by continuing it a bit, imagine "what the handset looks like" is "covered in spikes" and "where the numbers are placed" is "at random" on their default handset.
Re: (Score:2)
I'd say that having a bit more control of the implementation of sound would lead to less problems in the future (the constantly-moving linux sound architecture has until recently had Skype on a constant catch-up). Also could lead to integration with Pidgin or similar IM-aggregators, which wouldn't be a bad thing.
Re: (Score:2)
How about integrating the binary-only Skype support into Ekiga or Pidgin or whatever other open-source video/conference/chat programs? Gets close to using one program for all of your real-time communcation needs, regardless of protocol...Skype was the big missing protocol in most of those programs, IMHO. Hopefully it wouldn't be too hard to write a wrapper layer around the binary Skype library and make it appear as just another protocol library/plugin to those programs.
Also as FlyingBishop brought up, it
Re: (Score:2)
where the only thing you can change is where the numbers are placed and what the handset looks like. Maybe I'm missing the point, but how does this benefit anyone?
Are you old enough to remember when we had to rent phones from the telephone company? When you couldn't
go to an electronics store and by the phone you want? Suffice it to say, the Bell phones stunk, if you wanted to do anything but what they expected you to (the quality was great for sitting at a desk and talking). My childhood friends' mothers
Did you RTFM ? (Score:2)
Nope (Score:5, Insightful)
I call bullshit. I think it's just a tech support guy misunderstanding (and it seems a bi-lingual conversation so the chances of that are even higher).
Open-sourcing Skype is very different to allowing Mandriva to add a non-trademarked icon to the Skype software (a bit like bundling Firefox - fine so long as you respect the trademark on the name and/or the logo and their requirements), or put a Mandriva icon onto the package etc. The two are discussed interchangeably and I don't see how they are related.
I think it's more likely a massive misunderstanding on the basis of zero evidence / poor translation. At best, I reckon that Skype for Linux will allow itself to be packaged more easily.
How slashdot works (Score:5, Funny)
1) In a bilingual conversation, Skype support employee says "Skype will from now on be part of the open source community."
2) Blogger posts saying that Skype will be open sourced in nearest future
3) get reposted on various blogs
4) ???
5) Verified "news" on slashdot
Re: (Score:2)
5) Verified "news" on slashdot
6) Other more or less reputable news media "lifts" it from slashdot
7) Somebody references it on wikipedia
8) Other more or less reputable news media "lifts" it from wikipedia
9) Somebody adds more sources to wikipedia
10) Welcome to 1984 - and then some
Re: (Score:2)
11) Increased attention causes Skype PR person to officially deny rumors, and everyone forgets this whole non-event ever happened.
Re:Nope (Score:4, Funny)
Re: (Score:2, Informative)
Story has been confirmed at the Skype Linux blog:
http://share.skype.com/sites/linux/2009/11/skype_open_source.html
They are making an open source UI to allow better integration with distros.
Re: (Score:3, Informative)
Cautiously Optimistic? (Score:2)
From a practical perpsective this is good news and a step forward.
However, if part of this "open source" announcement means a binary-blob needs to be included on an open-source OS (e.g. Linux) should we still be worried?
Off the top of my head I can think of graphics cards, wireless network adapters, software and scanner-type devices that need binary "blobs" to be usable.
I am worried because this could be a growing trend of reliance on companies policy of releasing binary only software onto a open-source OS.
Re: (Score:2)
Re: (Score:3, Insightful)
Perhaps you jumped to Linux without considering that it wasn't the end all be all solutions that you were told it was?
When switching from Windows to Linux you give certain things up, when switching from Windows to MacOS you give certain things up, and indeed when switching from Linux to * you give certain things up. If you didn't, they would all be equal/the exact same and you'd have no reason to switch at all.
Abandonware in 3....2....1.... (Score:5, Insightful)
This is the same old story. The business doesn't want to support a Linux client so they open the code they have and abandon it.
I didn't bother reading TFA so maybe someone else can inform us how would one go about acquiring the binary blob in the future? What distros will the blob track? What about an ARM build? Ebay wants to limit their dev hours but abandoning the gui doesn't help them much.
Which is why I think they'll just abandon the OS altogether sooner rather than later and put a happy face on it with this stunt.
having visual hallucinations (Score:2)
How do you deduce this from a single blog post
"the Linux Skype version will become open source [ofaurax.free.fr] in the nearest future"
Re: (Score:2)
Re: (Score:2)
What distros will the blob track?
What makes you think they will 'track' any distro? If they're OSing the GUI, the only thing left in the blob is some networking and crypto code. That can easily be statically linked, and then you just let the distros handle the UI dependencies on their own.
update from linuxcrunch (Score:2, Informative)
This .. (Score:2)
Re: (Score:2)
Official statement from Skype (Score:3, Informative)
does skype even run on 64-bit machines? (Score:2)
Last I checked, a few months ago, there didn't seem to be any simple options for running skype on a 64-bit GNU/Linux machine.
Re: (Score:2)
Last I checked, a few months ago, there didn't seem to be any simple options for running skype on a 64-bit GNU/Linux machine.
http://wiki.debian.org/skype [debian.org]
"Video did not work for me, but otherwise it was all fine."
Works for me (Score:2)
About time... (Score:2)
I remember getting reamed-out by a support tech on the Skype forums for even hinting that an open source client might make solving problems alot easier - I hope he enjoys eating his words now.
Ever since moving to Linux, Skype has been the only thing other than MS Office that doesn't have an equal in terms of both quality and user base (at least, where user base is a factor). The main problems with Skype on Linux has been its ability to keep the client compatible with the constantly changing landscape in so
Less than ideal, but a huge improvement (Score:2)
Re: (Score:2)
For example Empathy will have built in support for skype, or even Google Talk!
Are you trying to imply that Empathy doesn't already support Google Talk?
'cause if you are, you're wrong -- it has supported Jabber for years (install package telepathy-gabble if it doesn't work for you), and the latest couple of versions have even supported Jingle-based voice and video chat.
Corporate "open source" for... (Score:2)
The corporate meaning of "open source" seems to be, as often as not, "we will not support it any longer, but we want a client on your platform and you can use it". Their support and implementation of the Linux client thus far has been, if anything, sub-par and fairly static in development changes.
Re: (Score:2)
Why bother copying Skype's protocol? Just do your own and make it opensource. And of you run your own free server (just make money from advertising or something, not charge for the service itself) you'll take customers away from Skype immediately.
The first thing I'll download. (Score:2)
"I'll open-source that too, and will make huge enterprisey pitches for the PHBs, until the closed-source virus that is Skype dies out for all eternity, until the end of all time."
Much like Apple will die out. Here's a clue, Skype did for VOIP what Apple did for computers. Made it easier for the average person to be a participant.
Right now getting Skype running is simple download, install, run, create account, enjoy. SIP is a little less Plug and Play.
Re: (Score:2)
Skype did for VOIP what Apple did for computers. Made it easier for the average person to be a participant.
Not really. Skype made VoIP more _visible_ because they happened to be pretty good at marketing, but the only thing "easier" about it is the lack of customer choice. Compare:
Skype: Install it and it works.
SIP: Choose a service provider, install a client, enter your service provider's address and it works.
So the only real difference here is that with Skype you don't have to pick a service provider, which means that you don't get to choose the one that is going to be most suitable for you. I guess if peopl