Forgot your password?
typodupeerror
Linux Software

Open Source And Net Telephony 79

Posted by CmdrTaco
from the interesting-stuff-people dept.
Gark writes: "There's an interesting article at Upside Today that talks about the Bayonne project, which has the potential to change the telecommunications industry the same way that Free Software has changed computing. It's interesting that the project got its start when a proprietary software developer was going out of business, and decided to GPL their source code, thereby creating new business opportunitites." The article talks a lot about Open Source and the Net, compared to the Telco industry and its history of proprietary systems. It's a good read.
This discussion has been archived. No new comments can be posted.

Open Source and Net Telephony

Comments Filter:
  • by sung (147113) on Thursday July 20, 2000 @09:44AM (#917785) Homepage
    Most of the older software companies should GPL their old source like what id did. I mean, wouldn't it be cool to rewrite your own versions of Warcraft? I like this company already. And old hardware comps. should gpl their schematics and all the relative specs, too!
  • From my experience with telco industry is alot like Win9x, the large base that it's on is pretty much proprietary, noone cares how it works (except for people that build stuff for it or have too much free time). However there is a very opensource top level interface that everyone and thier brother has tinkered with. It's been that way since the Carterfone decision and the breakup of AT&T and ma Bell (even though bells like SBC are buying other bells like Pacific Bell now). The only thing that really needs to be changed to change the whole infrastructure is the lowest levels.
  • This would be nice, except for the fact that most companies, after releasing their new products, no longer support their older products and sometimes even refuse to acknowledge the fact that they ever existed.

    --
    Your friendly neighborhood mIRC scripter.
    if (ismoderator(reader)) hidemessage(this);
  • the amazing is that cisco didn't buy this out already?! in Santa Barbara there were quite a few companies working on telephony equipment and technologoes. Many of them were bought out by the larger companies looking to get a strangle hold on the telephony market. I think digital sound was aquired by Unisys, and cadence and others were aquired by Cisco. Ericson jumped into the ball game too.

    kick some CAD [cadfu.com]
  • by Chiasmus_ (171285) <ayatollah_hyperbole@@@yahoo...com> on Thursday July 20, 2000 @09:50AM (#917789) Journal
    That article was not easy to read. For a while, I kept thinking, "Wait a minute.. this is about open source software... not open source telephony." I mean, the focus of the article was some ten-year-old abandonware made by a company that was going out of business. That sort of grudging "Fine, we're out of business, we'll open the damn thing" mentality is good for the Open Source movement, but it isn't exactly a triumph of ideology.

    Anyway, I guess the article in question is more about "internet voice conferencing" than anything else - something that's been around for quite a while, but apparently with ugly time delays. I didn't actually read anything that claimed that this open-source conferencer solved the problem, but I guess the idea is that eventually, we'll solve the problem. And by "we", I don't mean "me", so don't bother asking me for it.

    In conclusion, this is a Good Thing, but not exactly earth-shattering.
  • So many companies consider their software to be valuable even when it is totally obselete and useless. If you can't make money from it, its not going to cost much to give it away so why don't more companies do this. Just how much software over 10 years old could actually be sold? I bet there's practically none over 20.
  • Sounds like an idea with great potential to me.

    This could (eventually) lead to some damn fine software. I think it's high time that, as the net becomes more ubiquitous, I shouldn't have to pay long distance charges. Telephony just makes sense in terms of maximizing bandwith potential, and OST might just make quality, vesatile telephony a reality.
  • Actually, probably the best thing to do would be to modify copyright law since software gets obsoleted so fast.

    I realize that software is a never-ending project, and that code written several years ago may still be critical to a company's well-being.

    Anyway, suppose a new law was put in place that put software into the public domain some time after it was copyrighted, I suggest 10 years. This would help a little to increase market competition, would allow for more efficient programming (it's still marginally useful) and would entice companies to innovate, otherwise their ideas they repeatedly use would just get into the hands of other companies (take MacOS 9 as an example, it's ESSENTIALLY the same as 10 years ago except for the colors textures and bugs, don't flame me, it's an example)
  • Isn't that what you call it when you know the phone is going to ring?

    http://www.ridiculopathy.com
  • Good... "the current telecom market is like a Malibu, Calif., hillside in late August: rich, beautiful and just waiting for the well-timed spark to set off the inevitable oxidation process."...We, as users and supporters of free software ideologies and technologies need to expand our view of what we can accomplish with organised effort.

    Why? Reliable, useful, scalable technology can be developed without the presence of a monopolistic profit-hungry entity...What other tech industries and protocols will be affected next? I am really encouraged by this article. Too bad they couldn't make Iridium a free technology!

  • .........They say every ending contains a new beginning. In the world of open source telephony, where revolutionary breakthroughs have been slow to come, it took the final act of an outgoing proprietary vendor to give the movement its much-needed jumpstart. .....

    Its sad to know that these companies with so much potential doesnt realise the potential in Opensource until they are going out of business. If only they had designed and built it in the first place and released the specs for peer review by the open source community and standardising it, they would have been a success than having to shut down.

    .....the telecom market remains a labyrinthine hive of black box technologies, a place where customers and value-added resellers (VARs) remain dependent on the proprietary hardware and software systems of a dominant telco or wireless company......

    Just reiterating what I said above. But the sad part is they also made their money out of it and now that they have to shut down, only then they realise the potential of a GPL.
  • yeah, but nothing prevents the buyer from giving
    it away for FREE..so not really feasible for
    someone to "sell" the software..unless u're
    charging for *support*
  • While this may help the oddball company with a legacy system, I don't think this will help average consumers all that much. Actually, I think this would contribute more to the "Buy the latest release" upgrade cycle treadmill that 90% of the computing world is already on. The pressure for companies to add feature bloat so that the company is actually selling something new will be overpowering. You think M$ Word is big now at about 70 Megs with all the goodies? It'll take up a good gig if M$ has to worry about its software deathgrip expiring.
    --
  • by 11223 (201561) on Thursday July 20, 2000 @10:04AM (#917798)
    Why does everybody think that "Open Everything" should be good? In the Telco industry, software==hardware. You ship not just software or hardware, but a "platform" for use in all situations. In general, they're inseperable. Open Telco!=Good. Open Telco==No business model. That's the problem of the marketplace in this instance.

    Can't people step back and realize that Open isn't always good? Would you write your novel Open Source? Would you make your washing machine Open Source? No, so why would you make your Telco solution Open Source?

    As much as I'm a fan of the open development model, I sometimes think that people around here are just communists - they wish all products to be developed in an Open fashion, and that just ain't right (at least without a major rewrite of our governmental structure!)

  • Reminds me of this [slashdot.org] article about high-availability (99.999% uptime) telco systems, specifically Motorola developing a flavor of Linux to meet those needs.

    Today's article isn't about the telephone system at all.. It really looks like a bunch of back-patting because someone did the nice thing and open-sourced their old code. But how often have you heard Linux developers say "Gee, I'd really like to write an IVR app, if only there were good tools available.." C'mon folks. All that happens on NT, and I don't see a great deal of interest in changing that. As the article stated, it's not exactly a glamorous headline-filled buzzword-compliant field like e-commerce or streaming web content.

    Sure, this code will prove useful to someone.. but I don't see it being the foundation of anything big any time soon. I hope to be proven wrong.
  • The coders don't want to have to be expected to keep a dress code at home :-)

    "It's amost ready, we're bug testing"
    Actually it sounds like a few Dilbert cartoons.
    Or these User Friendly's:
    http://ars.userfriendly.org/cartoons/?id=1998082 9&mode=
    http://ars.userfriendly.org/cartoons/?id=1998090 8&mode=
  • by wishus (174405) on Thursday July 20, 2000 @10:09AM (#917801) Journal
    A brief history of the telecom network:

    In the late '70s a great change happened to the telecom systems of the world - the separation of voice and signaling networks. No longer was signaling (call setup, etc.) done over the voice trunks. Old phreaks may remember when the blue boxes stopped working. This was because of "out of band signaling".. the call setup and takedown, etc. was taken out of the voice lines. Having a separate circuit-switched network for signaling also allowed us to do things like 800 numbers, 911, etc.

    Now, the standards for Signal Points (SP) on today's SS7 network are very rigorous. The US uses ANSI while most foreign countries use ITU. Both standards are very very rigid. SPs (depending on their flavor) are allowed somewhere between 3 seconds and 3 minutes of downtime a year. That's for the system itself, not the network links, which fail more often for various reasons. The system has to be there to reroute traffic when the links go down.

    Keep in mind this is very oversimplified.

    Now, the latest revelation in telecom is the SoftSwitch - a "Software Switch" basically.. These don't run on Windows NT, or *BSD, or anything you might run on a personal computer or internet server. A successful SoftSwitch demands a fast, tight, realtime OS, that can be fitted snugly with the hardware.

    I don't think the open source model will work with telecom becasue this is not the kind of thing you work on at home in your spare time. This is not the kind of thing you can release to the community in alpha, and wait for them to lend a hand. This is not the kind of product that would benefit from having the source shipped with it. The customer doesn't have time or know-how to hack on the code and recompile if there's a problem, or the equipment to test it properly before putting it on the network.

    Sorry.. Nice idea, but until corporate America is seriously restructured, it won't happen.

    wish
    ---
  • This isn't really an article about a brave new Open Source world. It's about yet another OS beachhead. The real question is: Is this beachhead going to hold? So far, so good.
  • This would be nice, except for the fact that most companies, after releasing their new products, no longer support their older products and sometimes even refuse to acknowledge the fact that they ever existed.

    s/except for/because of/


    ---
  • I'm sure that the telecoms giants will give this the same warm reception that the music industry has extened to Napster.

    But it is the future.

    In the same way that digital CDs replaced analogue vinal, and compressed computer formats (eg MP3) are replacing CDs, the same will happen in the telecoms industry.

    Analogue phone lines are being (/have been, depending on where in the world you live) replaced by IDSN/xDSL, and these will be replaced by dedicated internet connections, with no capability to carry uncompressed voice data.

    just my $0.02

  • On the surface, MacOS9 looks like a good example, because it behaves a lot like older versions of the MacOS, but the truth is that it contains precious little of the old "System 7" code from 10 years ago. Ever since the PowerPC Macs came out, the OS code has been gradually switched over to be optimized for the new architecture.

    The MacOS was never written to be easilly portable the way UNIX was. That's one of its main strengths (tight integration with the hardware) as well as one of its main weaknesses (over-reliance on a handful of hardware makers like Motorolla).

    By the time 8.6 came out, there was very little of the old '040-centric code left.

    IIRC, Apple gives away MacOS 7.5.3 and older from their web page. (Alas, free beer only... no source code.)

    Your example is not completely wrong though... there are a few core chucks that they simply can't replace without creating huge compatibility issues. (Hence the total change to Mach-based OSX, along with the whole "carbon" migration kludge to keep their old customers & developers from abandoning them while the software catches up.)

  • Several posting have sid that in effect computer/network/IP telephony is a mess of proprietary hardware, softwware, and protocols. This isn't quite true - MVIP, SCSA, and H.100 are hardware standards for telephony servers (mostly specing big PCD-TDM highways to carry all that voice). Most platforms are one of these plus PCI. There some standards on the API side, CSTA, TAPI, TSAPI, although they've been rather weak 'standards'. H.323 is the recent multimedia over an IP network standard.

    A lot of a CT application or product is in the rest of the software - switch functions, human interface, feature set. I suspect that this GPL'd package will need a fair amount of bashing at the interface level to get it into shape for platform portability, but it should have some real value.

  • Heh, that works

    Though what I meant to say is that the some of the old software companies aren't going to give a rat's ass about their old software and, consequently, won't feel like putting the (minimal) time and effort toward GPLing the sourcecode.

    They also probably don't want to have to put up with thousands people complaining that they can't get the code to compile with their compilers.

    --
    Your friendly neighborhood mIRC scripter.
    if (ismoderator(reader)) hidemessage(this);
  • we lived in libertasia...

    http://www.libertasia.com/ [libertasia.com]

    Electricity and Telco are free and powered by the people underground running in hamster wheels that didn't pass the citizenship test.
  • Yeah... just like Win9x has an opensource top level interface that everyone and their brother has tinkered with...

    Did I miss something here? Somehow I can't seem to make that metaphor fit.

  • If there isnt a "Cold call, scream FIRST POST!, hang up" module, it aint Open Source.
  • is a Linux client for dialpad. That would rock hard.
  • Still, the abandonware beachhead isn't a great moral victory for OS. It's kind of like fighting World War 2 by waiting for the Nazis to get fed up with the snootiness of the French and decide they didn't want the country anyway.
  • Of course not, but most probably you will only ever sell one copy of your program.
  • by Greyfox (87712)
    I just want to use my Linux box as an answering machine. Time to look into the voice modem stuff again...
  • by K8Fan (37875) on Thursday July 20, 2000 @10:25AM (#917815) Journal
    Can't people step back and realize that Open isn't always good? Would you write your novel Open Source? Would you make your washing machine Open Source? No, so why would you make your Telco solution Open Source?

    The idea of copyright being of a limited term (originally 14 years plus an option for another 14 - recently perverted into life of the author plus 70 years at the behest of Disney) was that every creative effort would eventually enter into the public domain. And, "Public Domain" can be read as "Open SOurce" if you like.

    The works of Shakespeare are in the public domain - i.e. you can publish them, put them on web sites, use them as the basis of new creative works, print them on tolet paper, whatever - and no one can stop you.

    So, yeah...eventually everything (barring further evil efforts of the Mouse) will be in the public domain, thus "open source".

    Arguably, the intention of the Framers of the US Constitution was to maximize the amount of material freely available to the public. A case could be made that if a creative copyrighted work has been abandoned by it's creator and if no further effort is going to be made to make it available to the public, then it should automatically revert to the public domain.

    Copyright is a good thing, but I don't believe that it should be used to keep creative effort away from the public. Case in point: GO's Pen Point. This was the last real innovation in GUI interface (you may disagree, but check it out and really consider the implications of it's design before you judge). But still, long after Bill FUDded it to death with the idiotic "Pen for Windows" it's unavailable. If GO had placed it into the public domain, or it if had reverted to it by no longer being available, we would all benefit.

  • Still, the abandonware beachhead isn't a great moral victory for OS. It's kind of like fighting World War 2 by waiting for the Nazis to get fed up with the snootiness of the French and decide they didn't want the country anyway.

    I think you're doing these people an injustice. It looks like a lot of people believed in this product; and even though they just couldn't make a traditional company based on it work, they wanted better for their efforts than to see them vanish into obscurity.

    It's more like the small farmer turning his money-losing farm into a self-supporting commune instead of looking for a big agribusiness conglomerate to buy him out. That is a victory for Open Source, at least as I see it.

  • It's more like the small farmer turning his money-losing farm into a self-supporting commune instead of looking for a big agribusiness conglomerate to buy him out.

    Except, in this case, the developers are abandoning their project. If you read the article, it claims that the reason for open-sourcing the product was that "it wasn't exactly the kind of thing you could unload in a going-out-of-business sale."

    It's more like a farmer who hasn't had any bids from big agribusiness conglomerates finally sticking up a sign that says "Farm here if you want", and moving to Vegas to become a Blackjack dealer.
  • Would you write your novel Open Source?
    Well, I'm about to begin work on a physics textbook that I intend to make Open Source ... just as soon as I figure out what that means, exactly. :)
  • I tend to agree that not everything should be open sourced, but telco could benefit. Linux is an ideal platform for telephony because it can be stripped down and made very fast, it can be embedded and it is very stable. There is a great deal of telephony hardware available with opensourced drivers for Linux. Quicknet's drivers are even in the kernel starting with v.2.2.14; this is what created the heading of telephony devices in the Linux Kernel. There are even opensource VoIP protocol stacks avaialable such as OpenH323.org. This, to me, constitutes a decent platform. Now developers only need to create tha applications to sit on top, and the entire package would only cost a fraction of what a commercial system would traditionally cost. There is the opportunity to create custom, flexible call control applications for a platform that wouldn't have have a starting price of $30,000 and $250 per port like some traditional PBS's. The Opensource community would benefit by stepping back and relizing that their effort must in someway result in a marketable business solution in order to perpetuate itself, but don't discount the whole movement.
  • I think you just proved his point! You are saying "I don't even know what I am doing, but I will figure out some way to open source my book and do it". If you don't know what it means, why do it? Is there a practical reason or do you just walk around all day saying things like "Hmm, how can I open source my breakfast?"

    sig:

  • Reliable, useful, scalable technology can be developed without the presence of a monopolistic profit-hungry entity...

    True, but without that monopolistic entity the infrastructure that makes telephony and even the net possible just wouldn't be there. It's one thing to order up a point to point circuit from your local telco or one of the big three, but what would things be like if you had to run a cable from A to B instead.

  • ...but most probably you will only ever sell one copy of your program.

    [ incredulous "hello, earth to Stary" look ]

    That ... just hasn't been happening, though...

    You can even buy, for example, RH CDs without any mote of support and -- get this -- people are still buying them. Lots of Free Software-oriented companies actually make some (or nearly all -- i.e. Cheapbytes) of their money from continuously selling software-on-media-with-no-support.

    Hell, RH even puts RPMs and even ISO images up on their site so people can download and/or burn their own, and people STILL buy CDs from them.

    People even buy DEBIAN CDs for crying out loud -- there's NEVER any support for those, and Debian's a distro where apt-get mostly removes the need for a local copy on CD anyway...

    Apparently, it's suprisingly difficult to saturate the market for copies of a particular piece of software.


  • Support is not the only way to make money off GPL'd software. Just look at Cygnus and what they did with gcc. Sure, they did offer support, but they also charged people to let them add features. I'm not sure how well it worked, but hey, they're still in business :D
  • Guess what d00d, you're wrong.
    In the real world of suits, idiots, and people whose time is valuble, EVERY piece of software needs support. It doesn't matter how good it is.
    Joe and Jane may not need support for their iMac, but you can bet that if you walked into a Fortune 500 boardroom and said, "I'm going to sell you the best piece of software that could possibly be written", and they believed you, their first question would be "Who's going to support it?"
  • Cisco is indeed burning up the back door with voice over IP. A lot of work has gone into that technology. It is far cheaper to send large amounts of data over IP than a phone. People don't want to have multiple wires. So on and so forth.

    The needs of voice and network traffic are not exactly the same. (With data you care about perfect transmission of bursts of information, with voice you care about reliably holding a connection when you get it.) However it looks to me like the technology for data transmission is getting so much better than the technology for voice transmission that it is a question of time until the market says that we only need one of these...

    Cheers,
    Ben
  • My SoftSwitch running on a Sun Netra begs to differ with you on the "real-time, tight-to-the-hardware" statement. Note that the ANSI standard for SS7 calls for 56 or 64K links. Not exactly hard to keep up with, mind you.
  • Many companies do voice over frame relay now for internal communications. Since the FR is involved, the IT department gets to deal with it. Often, they don't have to deal with the rigorous uptime requirements the public phone system does. Usually, they have the PSTN connections in place as a backup anyway.

    In that situation, open sourcing can help significantly.

  • Hell, RH even puts RPMs and even ISO images up on their site so people can download and/or burn their own, and people STILL buy CDs from them.

    The problem with this is that not everyone has broadband available to them. I live in a medium-sized town (~35000), but yet there is no cable or DSL!(*#@!#! I'm stuck on a fscking 56k, which isn't very conductive to being able to download a 650 MB ISO file.


    =================================
  • If the source comes out, then that would allow us to make a copy of StarWorks - ie StarOffice with only the stuff people actually use, in the same way we got AppleWorks and Microsoft-Works. Oh and better integration amongst the modules.
  • I love your sig. :)

    There are people who mess with SS7 at home in there spare time. That is my latest hobby even though I don't work in the telco industry. Also there have a number of books on the subject surfacing over the last few years, indicating it's growing in popularity (or telco engineers are getting bored). Not everyone wants to be a carrier and there are many other places of interest for hobbiest like cell phones, VOIP interfaces, etc.
  • Support is actually an important feature, even if it is not used. Most commercial companies won't touch a product if they don't have something to fall back onto. By being able to buy support they live in the knowledge that if something really does go wrong they can talk to someone - though it does not mean that they will need to use the support they have bought.
  • Copyright is a good thing, but I don't believe that it should be used to keep creative effort away from the public. Case in point: GO's Pen Point. This was the last real innovation in GUI interface (you may disagree, but check it out and really consider the implications of it's design before you judge). But still, long after Bill FUDded it to death with the idiotic "Pen for Windows" it's unavailable. If GO had placed it into the public domain, or it if had reverted to it by no longer being available, we would all benefit.

    Yes, but what about the creditors? Suppose GO had placed their intellectual property in the public domain when they realized they had to fold, wouldn't that have been a form of fraud from the creditor's point of view? I mean, they were taking the intellectual property they had created over the years, and had suddenly made it impossible to charge for, effectively destroying it for purposes of resale and recovery of debt.

    The reason I ask is because a company I know, that has made this kickass abstract product that is so new that VC's don't Get It in one sentence or less - like spreadsheets when they were introduced in 1981 - that they cannot secure funding. The crators want to Open SOurce their product so that they can at least keep developing it and use it in their new gigs, but they are worried that the institutions the company is in debt to will not look kindly upon having this piece of value taken away.

    Has anyone else dealt with this before?

  • by dyfet (154716) on Thursday July 20, 2000 @12:43PM (#917833) Homepage
    While I understand the sentiment that some feel sometimes things are being done "open" just for the sake of claiming to be open even when its felt not to be the appropriate thing to do (though IMHO there is no case to be made for an advantage to "closed" solutions in any major market, but I digress), I believe telephony has an exceptionally strong case specifically for "open" solutions and that the GPL is the best vehicle both to protect the freedom offered in open telephony solutions and to promote it commercially.

    One respect that makes telephony unique is that most telephony solutions (such as small office PBX's, voice mail systems, IVR's, etc) are primarly delivered thru a VAR channel, and there are some 5000 indipendent telephony dealers in this country alone. These dealerships are often forced to supply customers proprietary products that they are only allowed to modify in the expressed manner that has been permitted by the manufacturer. While each mfg talks about "forming partnerships" with their reseller channel, manufacturers use various means to keep control of the reseller and the end user (such as requiring "exclusivity" to carry brand X product, disclaiming of warrenties, etc) and, in that they are actually quite removed from what the end user actually wants and desires, they often produce less than "clueful" products.

    What the GPL would mean for this market is that the reseller would be free to adjust a given solution to meet real and actual customer needs, and that the means to do so can not be taken away from the reseller later on. In this sense, it means one has to form a "real" partnership with the reseller channel rather than marketing sound-bites.

    The telephony reseller channel is one where solutions are fit to customer needs and the majority of the profit often comes from service agreements and support, rather than the sale of tangible goods. This sounds like a classic case of the "open source" service business model to me, and it's practiced every day by a 50 billion dollar industry.

  • I was reading a wired interview of the director(?)
    of Xerox PARC, and when they mentioned linux, he
    said that "Open Source is Literacy". Maybe people
    won't want to hack their own telephone switch, but
    if it brings some people to a higher level of
    understanding, can't it shake up the monopolies in
    other ways? Maybe it will lead to better standards
    in the future for something as "trivial" as
    digital answering machines or something else
    that can benefit us hackers.
  • Yes, but what about the creditors? Suppose GO had placed their intellectual property in the public domain when they realized they had to fold, wouldn't that have been a form of fraud from the creditor's point of view? I mean, they were taking the intellectual property they had created over the years, and had suddenly made it impossible to charge for, effectively destroying it for purposes of resale and recovery of debt.

    An interesting question. Penpoint is currently locked up in limbo for exactly that reason. The creditors are going to want all their money back with interest if anyone wants to revive it, or at least a reasonable dollar value. But as nobody will give them their $75 million dollars back, they have nothing.

    Perhaps the investors already got their money back in the form of a tax write-off? And if an effort like this is allowed as a write-off, the copyright would be voided and the work would be allowed to enter the public domain?

  • Telecos don't make money from software. Bandwidth and other hardware things perhaps, but not really software.
  • by sjames (1099) on Thursday July 20, 2000 @01:03PM (#917837) Homepage

    Nobody said it would be good for the members of the Telco industry. What was said is that it was good for customers of the Telco industry.

  • While in fact neither ACS nor the Bayonne project actually used any of the Ingate code (I actually have never even looked at it), I do believe one or more interesting projects could certainly emerge from it. I recall being told there is an example of the Excel switch control protocol, and perhaps a free switch control daemon project could come from it. There is presumably some good stuff related to clustering shared file systems between multiple hosts (what I recall of the Ingate platform is that it used a bunch of QNX/2 servers that all accessed a common SCSI array and common file system from it). There may be some code related to Dialogic hardware and it could certainly help seed the eccs "free" Dialogic driver project.

    I am not sure if the Ingate platform could itself be ressurected as a free software project in whole, but one never knows. It might be easier for me to be somewhat skeptical on the matter as I have no emotional/sentimental attachment to that particular code base. However, in my own opinion, when you have a very large code base that is already x number of years old, and originally was built in a proprietary manner, it might be best to use it as a "reference library" while writing something new to replace it with.
  • These both make sense - the idea of an open source text book would be rather than many people putting effort into many textbooks, they would be able to combine their efforts into one textbook, hopefully creating a best of class textbook.

    And open sourcing your breakfast doesn't make any more sense than open sourcing an individual Redhat CD (the material, not the data). But open sourcing your breakfast recipes so that they may be worked on by many people - why not.

    tangent - art and creation are a higher purpose
  • I've installed a pair of Netra 1400t space heaters which currently support >200 A-links. The big linksets often run 12 or more links. Total traffic of 160 erlangs getting routed on 64k ITU links.

    But you are right, most SPs run on Sun Solaris or HP UX boxes. The underlying OS generally doesn't get in the way of software. The real-time means transit times measured in tens of milliseconds, no sweat for a dual 466 Mhz processor Netra.

    Compare the traffic through an STP with a modern packet router. There is no comparison. The complexity doesn't come from the basic routing function, it comes from the reliability function. Calculating every link status, queue, congestion, acknowledgement and timer adds a huge amount of overhead to the routing process. The state machine is enormous, and the few manufacturers of SS7 software go through a formal code review process to prove the state machine.

    Given a stable linux distribution, an open telephony project could achieve the same results as the extremely expensive proprietary software. This will become more important as the european telco market opens up to real competition in the next decade. The small competitors will have to withstand a punishing pressure on their margins by the old monopoly players. By using a 'free' SP softswitch instead of a US$500k/year license will help their bottom line.

    the AC
  • Hmmm... I *do* work for a company making the kind of software you're talking about here. FYI we recently had a chance to see a major switch supplier demo their softswitch app - running in *JAVA*. Ok, it sucked, but the point is that softswitch can be done on whatever the hell you like - if a Java app can only handle 10 calls a second it still doesn't matter if it means that you can have an array of 10, 100, 1000 cheapo pcs to handle whatever number of calls your after rather than spend multiple millions of dollars on expensive switching hardware. The only thing that's important is reliability.

  • I have in my hand a copy of the AT&T Standard Technical Training Course, job aid 9, special services telephone handbook. Dated September 1978, this book makes absolutely no mention whatsoever of out-of-band signaling. It's a great review of signalling units for the various carrier systems, trouble-clearing procedures on 2- and 4-wire circuits, so on and so forth.. but everything's still VF signaling. (Even right down to the little note about using 2604 Hz test tones: "Do not measure at this frequency if 2600 Hz signaling units are used in the layout. Instead, interpolate from the values measured at 2504 and 2704 Hz.")

    Yes, the switch to SS7 did happen, but guessing from the evidence in my hand, it wasn't until later.

    I'm not familiar with the SoftSwitch, but I'm guessing it's a best-route selector for data circuits? Given the intelligence required to run one, it sounds like it probably performs error correction too. Damn, I spend too much time with transport and not enough time with switching.

    Here's another question, since you might know: There seems to be so much publicly available info about the intricate workings of IP networks, tutorials on everyone's products, etc etc.. But there's absolutely nothing out there about the real deep telco stuff. Is it because /everything/ is bound by an NDA? Or is there just not enough interest to warrant writing a set of basic texts on the subject? Everything I can find dates from the early dawn of ESS, and was apparently written in 40-column mode on an Apple //. I'm just looking for something a bit broader and more juicy than the ss7 overview [webproforum.com] I've found.
  • I think the greatest opportunity for opening up the phone equipment will be for business PBX type equipment. Currently if you decide to install a Nortel PBX you are locked into using Nortel phone sets and Nortel modules running Nortel software. I don't know if anyone has looked into the prices of the better PBX systems lately, but I can tell you that they are very expensive.


    Now imagine if these systems were open. You can basically plug in a box which would be more like a traditional computer with a T1, PRI, or maybe just plain copper interface card. The software running on that computer you can use to setup new phone sets, assign speed dial numbers, configure call blocking, and manage voice mail boxes. Because it's open you can get plug in software and hardware. Just a plugin away from an automated attendant that will read back your current emails over the phone. It would also simplify the voice over IP situation because just because your office has one brand of PBX doesn't mean you cannot talk to the branch office with an entirely different PBX over your existing WAN or VPN links.


    On the hardware side of things you no longer need to buy your phone sets from the same provider as the PBX. Just buy whatever model has the features you need and plug it in. Say you want one of those new 2 mile range cordless phones instead of a traditional phone set, just plug it in.


    Who is going to build this equipment you might ask? The answer is every telephone and network equipment company that is looking for a new market to sqeeze some profit into their bottom line. This is a perfect opportunity for the computer industry to undercut the traditional manufacturers. Most network equipment manufacturers would kill for half the mark-up that PBX's are sold for today and they will be the ones to create the technology. Someday your phone may have a Cisco logo on it.

  • by voxman (3163)
    The article was about breaking through the walls of one of the last and most proprietary software industries, and the one with the most to gain by this happening. The history at the start just highlighted the difficulty in changing legacy thinking. In this industry, it's earth shattering.
  • You obviously have never looked at a phone system. Software does not = hardware. Telephony systems are being developed out of commodity PC hardware these days, just like everything else. Application cards that handle analog lines and voip functions are becoming just as much a commodity.
  • No one cares how it works? Where do you think hackers came from? Or UNIX? The phone system is interesting as all hell and getting better all the time. Why do you think Cisco includes VOIP software in every copy of IOS that ships? Because it's not just the pphone companies and the hackers that are interested, it's the new generation of communications companies as well.
  • Listen, a full implementation of an H.323 stack is a softswitch. For an open one, see openh323.org. Every cisco switch shipped within the last year is a softswitch, it's in IOS. A softswitch is a call manager the responds via a standard protocol interface and carries out switching logic. THE GOAL IS TO CREATE OPEN SOFTSWITCHES. Your PBX doesn't use SS7, but when you are getting 32 channels of voice and 1M of data from Sprint next year over an MMMD wireless connection, you will be happy to find that your linux box can split this out to USB phones on your LAN. Why? Because some people are out there working their asses off to create softswitches that are open-source. That's why.
  • every linux company sells GPL'd software. SUN makes over half their consulting revenue off of sevices.
  • Get the disclaimers & disclosures out of the way first: I work for a company getting into IP telephony.

    I have an uneasy feeling about open IP telephony for individuals. It's really simple -- if you want enough bandwidth to support IP telephony, you're limited mostly to DSL or cable modems (wireless isn't widespread enough yet). DSL comes from the traditional telcos, and cable companies are also hot to provide telephony services these days.

    A CMTS (Cable Modem Termination System, the headend equipment) can be configured to ignore a cable modem whose MAC address doesn't fall into a certain range (I'm not sure whether this is part of the standard MIBs [cablelabs.com] or not) -- it's quite likely that a cable company will be able to throttle telephony packets that don't come from "approved" equipment. In the same way, I doubt that telcos are going to let everyone run their own IP telephone calls down a DSL line.

    As others have pointed out, open IP telephony will likely be an option for commercial outfits who can lease a T1 or fiber connection. I could see a company with offices in several cities using IP telephony to connect their offices.

    The good news, of course, is that IP telephony will eventually compete with traditional carrier-grade service, and that should mean cheaper phone calls for everyone.

    -- Dirt Road

  • Note: I had explicetly stated "except people who build stuff for it or have too much free time".
  • Your hopes have been realized. See the Bayonne project [bayonne.cx] for a next-gen open source IVR system running on Linux or FreeBSD and OST's website [ostel.com] for commercial support.
  • "open source software" for telecommunications isn't so much for the network, but for the end user on the desktop.

    The existing telecommunications networks are defunct: the so called 'intelligent' network that employed SS7 and smart nation-wide routing systems is a technological legacy due to the increasing ubiquity of the Internet and the progress of technology in the guise of moore's law. It's only a manner of time before it all migrates to an IP based network, and even AT&T has acknowledged this by declining from buying anything other than IP based equipment - this is why Nortel, Lucent, Ericcson and the major provides of old-school equipment are forging into new markets.

    Using the net for voice/media calls is still problematic because QoS and billing issues have yet to be sorted out. The mistake that many people make is to think that there will be an "internet telephone company" - there will not be, because the whole concept of "a telephone company" is defunct in light of the Internet. What there is, is a collection of bandwidth providers, and on top of that, there are all sorts of services, of which telephone is one.

    What will happen is that internet QoS mechanisms will come into place, and you will be able to "buy" guaranteed information streams across the network, and on top of that, you can run your video/audio/multimedia one-to-one or group communications tool. That obviously assumes the availability of considerable bandwidth, and it will take a while before the networks are in place, but you can be sure that they are heading there.

    While a lot of other software has become popularised through open source, no integrated telephony software has - I accept that there are various bits of software (desktop phone, internet videophones, etc), but there is no real movement or interest group in the way that there is for other software technologies and no real common understanding of what it means to make voice calls across the Internet - not enough people understand this concept yet. Even systems like IRC had a large interest community of developors that evolved the code. Where has that been happening for telephony ?

    This is summary: the concept of a telephone company is dead, and so is the idea of SS7 and 'old school' switching technologies. The internet is rapidly replacing all of this as a common transport medium upon which voice calls will be just "one type of service". Although various internet telephone companies are around, there is no real unification or movement to develop an internet based phone/media software platform that is open, extensable and integrated on the desktop for the use of the masses. And the masses do not yet understand the concept, nor have an incentive make the change - there needs to be some kind of infrastructure in place (i.e. once many people start to use DSL services, then perhaps if voice/video phone services became available, people would make the voice/video call over DSL/internet rather than making a standard POTS call, and from then onwards, would just not go back to making POTS calls ever again).

    Too many people still think that the internet is something at the end of their phone line, that do not yet realise that their phone line is the internet, and a voice call is just some service offered.

    The sorts of concept video phones that have been thrown around are also in the wrong paradigm, they work on the idea of an 'enhanced telephone' - the whole idea of a telephone is dead, as is an enhanced one. What is alive, and always will be, is the idea of 'communication' - "call mom" is something people want to do irrespective of whether they use a phone or an internet multimedia service -- all the more better by using the latter and being able to see her face as well.

    Through all of this, the biggest problem is the paradigm shift - people still thinking in old ways until a sort of critical mass and juncture hits them in the face, and then they see the new world, and by then, the old world is long on the way out.
  • Bzzzt. Wrong.

    Traditionally the big traditional telco manufacturers - AT&T, Nortel, Ericsson, Alcatel, ... - manufactured their own hardware and wrote their own software to run the stuff. This includes proprietary real time OS's and hardware drivers. Newer players in the router market, eg Cisco, Xedia, etc. do as well.

    These days the old manufacturers are integrating commodity hardware with their proprietry hardware.

    There are some very low end PABX applications running on commodity PCs, but the main game in big back-end switching and routing is still commodity, and probably will always be so, because you just can't switch or route calls or packets fast enough on a general purpose computer, in all but trivial cases.
  • The value of the software in question is largely in it's reference designs for a variety of protocols heretofore without open source stacks. In telecom, a large amount of equipment is shipped with a proprietary protocol of some kind, so for those who have such equipment (most businesses), a reverse engineered stack is a must if you are to gainfully interface to it, (i.e. with a linux-based voicemail system). The second area where the code is valuable is in ten yers of hardcore experience with messaging systems. The data structures that are outlined by the code are not obvious and did not exist prior to the code release in question, basically saving open-source programmers ten years of grief.
  • Why don't you look at running ACS on a quicknet phonejack?
  • Would you write your novel Open Source?

    I might, if my novel was really good, so that I could get a reputation as a good writer. Then people might even buy my books (I haven't even tried to publish anything so far). When I write, it's out of personal need to "spill my guts out" (metaphorically), not making money.

    Would you make your washing machine Open Source?

    Yes I would, if somebody would design an upgrade to make my laundry cleaner with less water or power used.

    why would you make your Telco solution Open Source?

    I think it would enable me to get "more bang for the buck", so that the phone service would be more useful and would have more features (I have no new features in mind, but someone else could).

    I sometimes think that people around here are just communists

    Granted, the Open Source movements' "things to everybody equally, for free" might sound to an uneducated reader like Communism, that is not the case. Why? Open Source doesn't prohibit anyone from charging money for services generated with OS software. Actually, the name "Open Source" itself explains it everything: Yes, the Source of a program might be available, but isn't it much easier to shell out a few bucks and buy the whole program (or service), packaged, with manuals and some nice add-ons too? That is Open Source at its best.
  • You might want to check this article

    http://www.pcquest.com/content/linux/howto/10007 0701.asp
  • Eeerm, not exactly sure I agree with you on this. In my view Open Source is exactly what telephony needs. You quite rightly pointed out the telephony needs obscene amounts of reliability and uptime, but this is exactly what Open Source is good at.

    You get many experts from all around the world to actually look at the code and you get less bugs. Even the thought that your code may be looked at by lots of people much cleverer than you makes you think twice when writing it.

    Before you scoff, I've been there and done that. I've spent ten years writing telephony code and managing telephony projects, I've done the SS7 thing, and believe me, the code I've written as Open Source is better quality and has had more thought put into it than the closed source code I have which has millions of calls running through it. For example a lot of my code handles millions of calls a month on British Telecoms internal networks, it ain't pretty bit it works, but the Open Source code I've written not only works, it's pretty too.

    If you want some examples of extremely reliable, well thought out and well tested Open Source code check out OpenH323 [openh323.org] and OpenGatekeeper [opengatekeeper.org]. (Okay the last bit is a blantant plug!)

  • by Anonymous Coward
    MODERATE THIS UP PLEASE

    In a former job, I used to design small PABX systems for a living. The code is a real bear and way too complicated, but it has to achieve a very high level of reliability. Telephone switches are basically simple robots that have to operate unattended for very long periods of time.

    Many on this list have mentioned 5 9's (99.999) but that is for providers. For an office PABX or small system, a bit more down time would be (and is) tolerated.

    Linux fits very nicely into this middle to low-end market, e.g., PABX for hotel/motel, PABX for small office, VPIP solution for home office. Linux can and does achieve very long uptimes on commodity hardware, more than adequate for an office system. Many vendors are using NT (e.g., Mitel), so if NT is acceptable, then Linux should be more than acceptable!

    With projects like that at Motorola, Linux would even fit nicely in a 5 9's environment with full redundancy, hot swappable compact PCI CPUs, cards, etc. Linux's standardization, network capabilities, and cross platform operation simply the development of application code by allowing for the use of COTS PC platforms for much development (call processing code, applications, WWW interfaces, databases, etc.).

    The marginal cost of a Linux-based system is also attractive. If you want to deploy 100 boxes, it is 30K-50K cheaper than the same boxes with NT/W2K, as each copy of Linux is essentially free. For someone manufacturing systems, this is a fantastic cost savings that directly affects the bottom line.

    In summary, Linux makes an attractive platform for TELCO products when deployed on COTS PC platforms.

  • Moderate the parent AC post of this UP, it's interesting and relevant.

    The only question I have about this is the following: how many open source developers do you think there are out there who have a clue how to develop (and SUPPORT, BIG question) telecom apps?
  • Anyway, suppose a new law was put in place that put software into the public domain some time after it was copyrighted, I suggest 10 years.

    Are you volunteering to be the U.S. representative to the treaty negotiations?

    Under the Universal Copyright Convention (Berne), no party may have a copyright term of less than life+50/75 years, and no party may reduce their term of copyright. There are roughly 180 nations other than the U.S. who are parties to the UCC; IIRC you need 2/3 to go along with changes before they take effect for the change-ratifying parties.

    So, before such a law could take effect (unless you wanted to make U.S. copyrights invalid across the world), you need to get 120 nations to sign off on it.

    Yes, it's a good idea. No, there's not a chance in hell it will ever be acted upon.

    Steven E. Ehrbar
  • It seems to be a common misconception that real time means "really fast" -- it doesn't.

    What it actually means is that operations will occur within a specified period of time -- no matter what.

    So just because you have a very fast computer assigned to the job, it doesn't mean it's hard real time. A very slow system could in fact be a real-time system.

    (Yes, there are also "soft" real time systems. But that's different.)

A failure will not appear until a unit has passed final inspection.

Working...