Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
Linux Software

IBM Releases GPLd WinModem Support For Linux 144

horst writes: "Subject says it all -- IBM has released first GPL winmodem driver. Link found at LWN" I'll be even more excited when they release the code that works with my T20 ... I've never even dialed my modem *sniff*, but if you've got an MWave (600, 600E, 770) then you should be golden. But props to IBM for making a cool move. Hopefully it's not an isolated one.
This discussion has been archived. No new comments can be posted.

IBM Releases GPLd WinModem support for Linux

Comments Filter:
  • That some of the more rabid linux-ers(?) will soon be using a 'win'-modem....
  • I hope for the sake of anyone using Linux with an MWave card that the drivers are of a superior quality to the buggy/bloated/disgraceful drivers produced for Windows.

    One would hope Linux developers were generally more switched on to writing decent code, although it may just be the underlying hardware to blame!.

    In saying that, top marks to Big Blue for doing the right thing.

  • by hafnium ( 79240 ) on Tuesday February 13, 2001 @02:00AM (#435676)
    So does this mean that we have to rename all 'win' modems to unidimensional polychromated multiplexing software based super-modems? I mean - if Billy Gates wanted to keep modems for himself -- he's gotta try better than just to name something "win". Some smart soul or two is just gonna try to get it to run on linux! muhahaha
  • Does this mean that every modem will become Linux-compatible? No.

    It does, however, give a nice shred of hope to those of us who can't really afford an external or expensive internal modem.

    One thought though. Since this IS GPLed, could we use it to make more modems work? I don't know if it is, but if the MWave uses the AC '97 standard, we could get things moving quickly. Intel may actually have helped Linux again. (Note: If you are an Intel PR person, don't say that this was your original intention; admit that you made a mistake. We would still love you...)

    Just make the MDP3880-W(U) a priority, k? ;-)
  • Intel and IBM are different companies :)

    +++++
  • ... and I thought, IBM developed this thing for their ThinkPad. Hmm.
  • by gattaca ( 27954 ) on Tuesday February 13, 2001 @02:19AM (#435680)
    According [theregister.co.uk] to the register [theregister.co.uk], Alcatel will be releasing Linux USB drivers for its ASDL modem in the next month or so...Open source etc...
    It will be here [alcatel.com].
  • whats it with companies refusing to document spec? i would love to know why lucent is hiding info on thier winmodems. do they honestly think some competitor will clone them and sell them for less? by that time lucent has gone on to the next version...

    of course, i still dont know why ibm just doesnt just use real modems. they worked great in the earlier models. i somehow doubt the cost difference would be more than a few more dollars at the factory...
  • ...that this actually works really well. I love Thinkpads. I've been using an old one (600, no E) for a long time.
    Linux has been wiped off my drive twice due to interconnect failures.
    I've also flown > 20K miles with it (I know some readers will laugh, but I'm a tech guy), and it still works. Two years of a laptop doing well for me, and two separate Linux attempts makes me hopefull. Ill at ease, but hopefull. Hell, I like BSD and Solaris, so I know I'm going to be shot down, but laptops are different. People need easy, fast, interoperable [whatever] before they'll even comment. No cares about which layer it acts on, or whether it is proprietary, _people need email_. The rest is cool, too. I'm done.

    Sure, modems are dead, or something.
    Tell me that when I need to get email through via a random hotel.

    -j

  • Can't comment on the stability of the driver as I haven't tried it -- yet.

    The driver though is developed from the NT driver which IMHO is a good sign. NT is way more linux like than Win9X/Me.

    Don't know if thats enough to get the MWave back on it's feet. I have heard of problems with IRQs and other things which led me to suspect that the MWave while a great design in theory was a poorly executed design in the real world :/

    See for example

    /usr/src/linux/Documentation/sound/mwave

    http://www.ibmmwavesettlement.com/

    on the matter
  • OK, I see the ploy, they wait for me to buy a winmodem, then switch to Linux, then be nice and give away the modem, THEN THEY RELEASE THE DRIVERS!
    Well, atleast I have RoadRunner now
  • by Jamie Lokier ( 104820 ) on Tuesday February 13, 2001 @02:53AM (#435685) Homepage
    Take a look at the very large number of ".dsp" files. These are binary only files. They contain all of the difficult modem algorithms. The GPL specifies that the code must be in "the preferred form for editing", which is long for "the source code". I assure you that binary .dsp files aren't the source code. The tar file includes the GPL in the usual "COPYING" file, but none of the driver source files refer to it. Neither does the documentation. In other words, it's a tar file which happens to contain a copy of the GPL text. In summary, none of the source code is GPLed, and all the difficult modem algorithms are binary only. Thanks for the helpful driver IBM, but don't pretend this is free software. BTW, there have been half-binary, half-source drivers for Lucent modems for a while now, and several other manufacturers too. -- Jamie
  • IBM have said in the past that unless a manager can come up with a very good reason not to release something under the GPL (such as proprietry code huge technological advantage etc) then everything must be released under the gpl. I for one am pleased that they are holding to this.
    Hopefully they can now sell there linux certified laptops with more of a clean consciusness.
    I do not know if this will help the other winmodem users (see www.linmodems.org) there is a lucent driver that works with my saterlite 4030cdt but hope fully this will help to getting an open source driver (or a second sound card)
    The other modem releated question is those amr (audio modem riser cards)
    does anyone know howto / if these work under linux
  • by zensonic ( 82242 ) on Tuesday February 13, 2001 @02:55AM (#435687) Homepage
    After having downloaded the source and unpacked it, I noticed some things:

    • Only support for the highend 3780 Mwave chip, not for the older 2780 chip. As the 2780 was found in a lot of aptivas I suspect that there is a lot of 2780 but not that many 3780. Could be wrong though.
    • No documentation on the MWave chip itself. I know: The code provides!! ... but actually no it doesn't. I can't enhance the driver to support my 2780 board. Can't tweak anything without going through the painfull compile-run-crash procedure :/
    • With no documentation on the Mwave chip we can't make our own litle .dsp files which does what we want it to do.
    • With no documentation ... we can't easily fix the sound problems that people are having using the sound resources of the mwave boards.

    Some problems can be overcome simply by the experimenting programmer (compile-run-crash type), but it will be a real pain. Why not open up for the documentation so that the MWave board can show it's potential: A bunch of resources (dsp, soundcard, telco interface, midi interface) tightly knit together with the dsp chip in control.

    Call me fanatic :) I think it's great what IBM has done, but they could have done more on the matter :/
  • Wow, this is a nice move. Hopefully some other companies will take a hint and do something similar....

    Even so, I've never really liked the premise by which winmodems work. They use the sound card and clock cycles to generate the modem-y sound, and then send it out on an RJ-45 (or whatever phone cable is). You'd think in the day of cheap hardware, it wouldn't cost that much to just adda nother processor (small) onto the modem itself and free up the other hardware a bit.

    Also, wouldn't it be nice if other generic winmodem manufacturers see this open standard of sorts and comply to it?

    For those who know, can this winmodem driver be ported to BSD and other OS's pretty easily?

  • I mean, how 'new' is this modem for which IBM is releasing the driver? I remember having a mwave-based modem/audio combo three computers ago.
    I can see a trend, here : when the technology of a piece of hardware is not hot anymore, they open the specs/ release the driver, so that they found a new market in Linux users.

    Well, not that I like to be a second-class user. But OTHA, I'v never been keen to hot hardware, either. And a driver is better than no driver. And an open-source driver is better than a half-hacked binary one (Lucent, please, follow IBM example! ).

  • of course, i still dont know why ibm just doesnt just use real modems. they worked great in the earlier models. i somehow doubt the cost difference would be more than a few more dollars at the factory...

    Disclaimer: I know nothing of intentions. The following is drawn from observations

    MWave isn't just a modem. It's a generalised DSP solution; effectively a DSP co-processor. Originally, way-back-when on the 750/760 even series Thinkpads, the MWave chip provided soundcard support (SoundBlaster Pro emulation I believe) and Modem support; the MWave chips ran a mini-multitasking OS that would time-slice these tasks so you could play Doom online thru the modem and still hear sound. There were some obvious drawbacks to this approach - IIRC if you wanted 33.6K modem dialups you had to fallback from stereo to mono 22KHz soundcard.

    I believe the design point was that soundcard+modemcard cost more than MWave and thus the total system cost was cheaper; the side-effects of a software upgradeable soundcard and modem were secondary.

    Some of the Aptiva model desktop PCs also got the full MWave for sound + modem (or was it sound only? I forget).

    What I don't understand is why in newer Stinkpads (such as the models referred to by the article), the MWave hardware (presumably in cut-down form) is only used for modem function; the sound card function is provided by a standalone chipset (uur... CS4236+ I believe). This Makes No Sense to me.

    Does anyone else who can organise thoughts more coherently have any better insights into the history here?

  • The Dutch PTT [mxstream.nl] uses these as well. But if you have the choice, go for the Ethernet version instead (costs EUR 50 extra over here). USB generally draws more CPU cycles than Ethernet. The USB version doesn't even work with Windows ME if the multiple IP addresses version of PPTP over ADSL is used! And as a plus, the Ethernet actually looks like a modem, not like a stupid green horseshoe crab ;-).

    You'll also need the Linux PPTP driver [mit.edu]. Hopefully it works with these USB drivers.

    Jacco
    ---
    # cd /var/log

  • Bill Mair setup a pettion over 2 years ago, and we have be goading IBM ever since, particularly over the "IBM supports linux 100%" even though it part binary it is nice to see, this IBM driver, not that i think anyone will use it, we all bought a pcmcia modem along time ago. Just its nice to see ibm supporting legacy machines, yup mwave only ships with disconnued machines. However this is good I guess cos LINUX brings old machines back to like. My 770X has has another 3 years in it.

    James
  • I know someone who's gotten the T20 modem to work.
    Check out http://www.cs.utexas.edu/users/kharker/linux-lapto p/
    for more info.
    Of course, this was with redhat 7.
    I dunno about others.
  • by Jamie Lokier ( 104820 ) on Tuesday February 13, 2001 @03:10AM (#435694) Homepage
    Oops! The driver source is GPLed, it just doesn't refer to the "COPYING" file like I'd incorrectly remembered in the traditional GPL copyright template. Ahem. The .dsp files, containing the modem algorithms, are still binary only though. The whole driver is still not free software. You don't have the freedom to port it to other hardware, run it over a sound card or a ham radio, modify the signalling methods, or study the DSP code to learn how a modem works for example, but parts of it are free and GPLed. (apologies for not reading the source properly), -- Jamie
  • IBM *did* develop the modem for their Thinkpad. BUT, Intel made a standard for cheap internal modems. It's part of at least the i8#0 chipsets (ie, i810, etc.).

    If the thinkpad internal modem is AC '97 compatible (which would make sense, since laptops do try to save space and using Ac '97 would help), we can reuse the AC '97 code for other AC '97 modems.
  • I know that! If the IBM modem conforms to INTEL's AC '97, INTEL's modem standard may be useful since we can reuse their code.

    Sheesh!
  • I didn't read your comment properly.

    I retract all allegations of muppethood.

    +++++
  • by snellac ( 314920 ) on Tuesday February 13, 2001 @03:21AM (#435698)
    OK, so most Linux people don't like the fact that winmodems are closed devices that are not supported on Linux. The conventional wisdom complains that existing winmodems give poor performance and kill the CPU. However in a recent /. post no less a personage than John Carmack suggested [slashdot.org] that winmodems could be implemented in a way that is better than conventional modems for the needs of interactive games.

    In the process of doing a web search I then turned up Stuart Cheshire's old home page [stanford.edu]. For those who don't know who he is, well before the web was popular he wrote a classic networked Mac game called bolo [stanford.edu]. (In fact when the web became popular the bolo players used to curse that the web was dragging the internet down too much...) Most links to it are dead, but the official home page is still up [lgm.com] although there has not been a release since 1995 [lgm.com]. (This was apparently done as research into the needs of interactive networked programs. Gee, all of those hours that I spent as a test subject without knowing it...)

    With Stuart's credentials established, it is well worth looking at his rants [stanford.edu]. In particular his latency [stanford.edu] rant, which was expanded out into a white paper [bolo.net].

    Once you are through reading those you will see that for anything interactive, particularly games, what really matters is latency, not bandwidth. And modems are a major source of this latency. In addition he and John Carmack agree that software modems (AKA winmodems) can be (though they are not currently) programmed to operate in a mode that reduces latency, and the result would be better for interactive games than conventional modems.

    So, are winmodems just a bad idea, or are they just poorly implemented? Conventional wisdom says that they are bad no matter what. But the people who should know best suggest otherwise.

    -snellac
  • Accentially what a win modem is a Digital to Analog Converter. And the OS itself has to basically send the sounds to the modem kina like a wav file. The atvantage of this is the fact that they are cheap to build. But will put a strain on the CPU of your computer and whathappends if you are doing some real Intensive CPU utilities, your modem can disconnect or hang on you. (Which WIN Modems are known for) Win Modems are a dumb Idea in begin with. Now with linux supporting Win Modems how will the real modem survive?
  • by mindstrm ( 20013 ) on Tuesday February 13, 2001 @03:31AM (#435700)
    Those 'dsp' files are firmware to be loaded into the modem card itself, and processed onboard. There is no reason we need source for these, and the same files would be used no matter what OS it is. The trick is how to get the contents of those file sinto the modem so the DSP on the modem can use them.

    They aren't even technically part of the 'driver'.

    You have the freedom to make their modem work on *any* hardware platform now; just not to steal their DSP code.

  • Although this may be a nice day for some thinkpad owners, these drivers doesn't help the linux

    community in general.

    What we really would need is an opensourced/GPL'd implementation of all the dsp-like algorithms used by software modems. They are currently high valued property by some manufacturers and therefore only given out as binaries for linux drivers if given out at all.

    IBM does it that way, and AVM (famous manufacturer of ISDN cards here in germany) did the same when releasing software modem / fax emulation.

    I would really love to see one of the main linux players sponsoring an open source development of such an implementation. This would allow the support of many winmodems after all (just replace the modem specific interfacing stuff), fax / modem emulation with all isdn cards, usage with ham radios and much more nice possibilities.

    There _are_ some opensource efforts (see www.linmodems.org for some pointers) but they are stuck in some main points and could probably need some financial support as well as development help by DSP specialists.

    So Redhat, VA or whoever feels responsible: Put some of your money here. Many people of the linux community will be grateful
  • I have got the lucent one on my thinkpad here. They only have a binary driver which only works for 2.2.1[234].

    Now for me to be able to use both internet and 2.4.1, I actually have to ipmasq via a windoze box at home...
  • I hope they'll avoid closed hardware in the future. I made sure to not get a Thinkpad last time I was shopping due to the MWave problems. (Yes, it was for Linux use. I bought a new Toshiba laptop and reformatted the disk for Red Hat.)
  • But if the DSP sources would have been open-sourced as well, it would have been possible to port them to other Winmodem (and ISDN4Linux) hardware as well. That would have made a huge difference.

    This is a good beginning, though.

  • Win Modems are a dumb Idea in begin with.
    Sure of that? I happen to have a Lucent winmodem on both my Linux boxes, bought when I still dual-booted. I often do other things, while downloading, and since I have flat rate I often leave connection open. Never got disconnected because of lack of CPU.

    Sure, a 'real' modem would be better. But so would be two CPUs ...
    It's all matter of trade-offs. The real concern is that in this case is not the user which makes the trade-off, since winmodem are often built-in in mass-marketed computers.

  • There is an Mwave developer's kit that was sold to OEM's and other people to make their own DSP files. I had it for a while, but there wasn't really any good docs in it. Check around on ebay for a copy.

  • by karmawarrior ( 311177 ) on Tuesday February 13, 2001 @03:51AM (#435707) Journal
    So, are winmodems just a bad idea, or are they just poorly implemented? Conventional wisdom says that they are bad no matter what. But the people who should know best suggest otherwise.
    There are two types of Winmodem, modems that have most of the logic in the "PC space" of the modem, and those that have most of the logic in the "Modem" space of the modem. Usually, the logic is in the "PC space", that's why they're cheap - instead of having an expensive DSP do the work of turning raw PCM into V.32/34/90/22/etc/*, the Pentium does it. This is cheap. It also means that the performance of the PC itself suffers, but adds as small compensation the fact that the data doesn't have to flow from the modem to the PC post-decoding, ie it's ready for the PC to use right away, reducing latency.

    In IBM's case, this model is not what's being used. The MWave is a DSP. So the "cheap" argument goes out of the window, except in that IBM recognised that a DSP could be used for multiple applications, and originally the MWave chipset was implemented by them exactly that way. My TP 760XD for instance uses the MWave chipset to provide both modem functionality and 16 bit soundcard support. Latency isn't likely to be a problem as the communications between the DSP and the "PC space" is much tighter than it is with a conventional modem, which usually goes via a real or imagined serial link controlled by a conventional UART chipset. So latency is going to be better than it would be with a real modem, but not as good as it would be with a conventional Winmodem.

    Essentially you could say there are three types of modem: Conventional, open, serial modems, which will work with everything at a minor latency tradeoff, Winmodems, which will only work with the operating systems (or rather system, support for non DOS Windows based operating systems is rare, and that includes other Microsoft operating systems such as NT) supported by the manufacturer, and will slow down your computer's performance with a small advantage in the latency stakes, and Other Proprietry Modems, such as the IBM MWave set-up, where you still have the problem that the OS has to be supported by the manufacturer, but neither reduced performance or latency are real issues.

    On the face of it, if someone could invent a generic device driver mechanism, or even just force, somehow, manufacturers to produce open source drivers, IBM's approach would probably be quite good. As it is, a year or so after IBM started this project we have a driver that only addresses the modem side of the MWave and only works with the later, less popular, Thinkpads. I'd have rather they worked on the soundcard functionality, a good PCMCIA Modem costs less than $30 on eBay these days. Grumble.
    --
    Keep attacking good things as "communist"

  • Actually, it means people will start calling modems by their proper names again instead of "winmodem". If a modem does all the modulation/demodulation in the hardware, it is a hard-modem, and if it does that in software you must install, it is a soft-modem.

    I have always called them that. I *hate* the word "winmodem".
  • I'm delighted that IBM have put together a driver for previously unsupported hardware, and I really don't want to sound ungrateful, but...
    • The MWave DSP is used on Thinkpads to provide Soundcard and modem functionality. We can get cheap PCMCIA modems for $30 on eBay. However getting decent sound card support on a MWave laptop isn't as easy - we can use the infamous Linux MWave hack, which gives us fixed rate 22.1kHz 8 bit sound which doesn't work with the majority of sound playing applications out there in my experience and which resets itself and ceases to work if the laptop is ever suspended, or get a PCMCIA Soundblaster card which is usually pricy and, to the best of my knowledge will have exactly the same problem with low sound quality => poor compatability. In short, why concentrate on the largely redundant modem functionality?
    • I'm glad IBM are supporting the last set of Thinkpads to come with MWave chipsets, but given there are almost certainly many, many, more Thinkpads and Aptivas out there with the older chipsets, is support coming for those machines? I couldn't see anything to suggest so.
    I guess it's a matter of priorities, but... I wonder whether enough documentation exists that would allow a reasonable programmer to implement the missing functionality using the IBM driver as a base?
    --
    Keep attacking good things as "communist"
  • by stripes ( 3681 ) on Tuesday February 13, 2001 @04:14AM (#435710) Homepage Journal
    Those 'dsp' files are firmware to be loaded into the modem card itself, and processed onboard. There is no reason we need source for these, and the same files would be used no matter what OS it is.

    Sure there is. If you want to fix or improve the DSP part, or even understand it (or build sonar with it). Which are very big parts of what opensource is about, not just "it can run everywhere", but "it runs good".

    I expect with the DSP part you could make a "voice modem" and build your own voice mail system.

    IBM did a good thing making the kernel part opensource, but the DSP part is still closed source, and to get full advantage of this hardware you need that part too.

  • I believe in business terms this is free software. I mean, they didn't charge anyone for the port. You can't expect IBM just to open source all their IP. Afterall, they are trying to make money too and it doesn't make sense to just give it all away at once.
  • When you have a short look at the source you will very fast discover that there is a lot of binary only stuff in the mwave directory called something.dsp. I guess these are the DSP based algorythms for implementing the different Modem protocols as the names let guess. IIRC the different protocols like V.34 and V.90 contain patented stuff so we will NEVER (Next 18 Years) see really open Winmodem drivers. I dont think the low number of saved transistors are worth this hassle - So please stop using winmodems. Glue he patented stuff into Hardware.

    Flo
    PS: Patents are evil but better stick them into hardware as HW is not really interesting to copy anyways.

  • DSP???
    I thought the point with theese "win" modems was that they did not have in dsp, but insted relayed on the main cpu as dsp.
  • by Anonymous Coward on Tuesday February 13, 2001 @04:29AM (#435714)
    The problem is that if you modify the dsp sources your modem ceases to be legally usable on the public telephone network. So what would happen is, since most people don't have access to telephone line emulators the drivers would have to be tested illegally. As far as I'm aware in every country it is illegal to attach unapproved equipment to the telephone network. So using the drivers would be illegal, and thus no distribution would risk the wrath of the FCC, and other governments. You wouldn't have any drivers at all if the dsp sources could be modified.
  • Another sometimes-used name is 'softmodem', which sounds correctly to me.
  • Their IP is hardware, A GPL'd driver will allow them to sell more hardware, as the various OS geniuses will port this driver to every possible platform. The other ~10% of the market may seem trivial to Microsofties but it's a huge market and as such it makes sense to try and sell to 100% of the market. After all, it's not as if IBM will have to pay someone to do the porting work, whereas they did have to for Windows and Linux. Having said that, this is yet another great move from IBM, more power to you.
  • Mwave is a completely different kind of beast, in fact it was developed way before Intel released their standard (my thinkpad has a mwave, and it was built in 1994). The mwave has a dsp running its own firmware, and emulates both a modem and a sound card (but not the two at the same time). Not a bad idea, but, as it has been said before, the implementation was terrible. Won't help you a bit for other modems.
  • That makes a lot of sense though as Linux appeals to people with older hardware that can't afford to keep up with the ever-hungry Windows upgrade cycle. IBM have little to lose here (Winmodem competition is pretty fierce already) and a great deal to gain.
  • I have got the lucent one on my thinkpad here. They only have a binary driver which only works for 2.2.1[234]. Now for me to be able to use both internet and 2.4.1, I actually have to ipmasq via a windoze box at home...

    That's not true any more, although I'm somewhat unsure of the legality of it, as it doesn't appear that the original ltmodem source code was officially released by Lucent but rather leaked by a third party with source access...

    Anyway, if you have a Lucent winmodem, check out http://walbran.org/sean/linux/stodolsk/ [walbran.org] where a fully working open-source driver based on the original Lucent driver is available. It has numerous bugfixes compared to the original Lucent binary-only release, and compiles cleanly for both 2.2 and 2.4 kernels.
  • Having some OS driver released by IBM for Linux is more important NEWS FOR NERDS than ... IBM's complicity with the Nazi's
    Correct.
  • If you design or change a modem algorithm, you are required to have it approved by the FCC before you are legally allowed to connect it to the telephone network.

    Most of you wouldn't know what to do with MWAVE DSP source code anyways... You don't need it.
  • I may be a young pup, but damnit, has anyone actually gotten this to insmod this driver. I only ask due to the fact that my old company, (God Bless you New Economy), was always spouting off, "and our new drivers...", which I was usually in the back ground doing a "No, No, NO!!!" and living the IBM new marketing plan(you know, "We can do it three months"). Talking with tech support is fine, but if noone has used it yet, I'll stick to my good old USR33.6, and you can tell me how to write non-runon sentances!!!


  • so.... anyone started reverse engineering the DSPs yet? what you describe is interoperability, in which case reverse engineering is very specifically allowed. Although I admit I am not bound by any DMCA..

    //rdj
  • BTW, there have been half-binary, half-source drivers for Lucent modems
    Unless I'm mistaken, the lucent source package consists of some c files and a binary library.
    Not open source, but good enough for me to get my lucent modem working under kernel 2.4 in a compaq armada e500...
  • GNU Image Manipulation Program Toolkit

    GNU==GNU's Not Unix...

    Damn, you've got me. :-)
  • Unfortunately not; the driver isn't really a winmodem driver :(

    The mwave modem appears to be a DSP-based solution (coltrollerless modem?) and a lot of the source is some very non-open-source binary DSP files that get uploaded to the onboard DSP. Seeing as how v34 (etc) are all pretty tied up in patents & licencing, this explains the release. They've released a soft-ish modem driver that will work with some machines. It's not a generic AC'97 "let's do DSP stuff on the actual PCM audio data" modem, and as such isn't a huge amount of use for many people :(

    It's great if you have an IBM 600E laptop, though. I've got a Vaio...

    Hugo
  • WIN == Windows Isn't Needed
    --
  • Yes, i agree with your parent to an extent.

    I had (and still have, it's in there somewhere :-)) a PCtel modem and found a nice binary kernel module for it.

    I too suffered from the problems of shitty connection, and frequent hangups when the CPU is loaded, there is probably a way to give realtime priority to the module / pppd but I didn't find it.

    On top of that, an interface must have changed because I couldn't get that mother to work in any kernel > 2.2.14 which is a bit of a drag in the fast moving releases of today.

    In the end I went out and bought a cheap RealModem(tm) and things are much better than they were, but i agree that linux drivers /SHOULD/ be available, the choice should be there, and as you said, a lot of people get them without realising it, like this poor sucker.

  • This was easy to predict, and expected.
    IBM has been releasing drivers and support for all their laptops with linux.

    If Taco had posted which problems he has with his T20, I could point him to IBM's Linux Technology center where they have the answers.

    I will come back and reply to this with the correct link where t20 support can be found.

    (I thought Rob liked Vaio's! good on him for using a thinkpad!)

    A host is a host from coast to coast, but no one uses a host that's close
  • But props to IBM for making a cool move. Hopefully it's not an isolated one.

    Didn't you mom teach you manners? Just say thanks man!

  • I wish IBM would spend some time to get my Lexmark 3200 colour printer to work with Linux. Since I've been on the @home cable Internet service I havn't used my modem at all except for the occasional fax in the last three years. I think that win-printers should be a bigger priority.

    Anyone know how to get one of these printers working?

  • I can agree with you that the complete driver probably does not validate as open source...

    But still I think this is a very good thing IBM has done... and I doubt that you'll get more than this out of them... at least at this stage...


    --
  • The point you make is fascinating, but not all of us use our modems to play games. Would the "give poor performance and kill the CPU" argument still hold for those of us who use modems primarily for file transfer, and never run anything more interaction-intensive than IRC?
  • How much improvement in latency are we talking about? If I dial in to the terminal server at work, I get a latency of around 120ms -- long distance dialing may increase this to 200ms or more. My DSL connection (which shares the same ISP as my employer) has an average latency of 20ms between home and work. A user currently dialed in from NYC has a latency of 220ms and a frame-relay connection to NYC has a latency of 60ms.

    The point is, how much latency is the serial connection actually adding? 10-20 ms? Any significant improvement will likely require least a 50% latency improvement and it's hard to see a perfect Winmodem driver cutting 120ms into 60ms.

    I'd love to know what the actual improvement really is and at what point the gaming community considers latency problems "cured". My online gaming experience with Half-Life has been that anything much over 250-300ms latency (measured in HL terms, which isn't really ping latency) starts to get really laggy.
  • Many winmodems do have a DSP chip onboard. The main difference between them and a "real" modem is that they control the modem and DSP directly over the PCI bus instead of through a faux UART interface, they download the DSP code on the fly instead of storing it on the board and they sometimes use system memory instead of onboard memory. All this uses little or no more CPU than a normal modem at high data speeds. The Mwave modem IBM has released drivers for appears to be of this type, as are all of the "winmodems" using Lucent chips.

    The real culprits for the bad rap that "winmodems" get are the cheapest of the cheap modems which do everything in software, suck up lots of CPU power and aren't worth the circuit boards they're printed on.

  • Costs a hell of a lot more extra here - for EXACTLY the same bandwidth you pay £250 instead of £150 installation and £100 instaed of £40 per month. Not even slightly close to a good deal as for the extra £100 installation you can buy network cards, hub and cabling and using free (as in beer) firewall and proxy software share out to all the computers you have, just like the more expensive ethernet version, but you have an internal LAN as well... BT are so stupid, but many of the public are even worse so they might just get away with it! Nice to hear the Dutch are being sensible about the pricing though.
  • They are uniformly afraid that their modems will not pass certification if they are seen to cooperate in any way with circumventing the certification process. And if a company helps people write non-certified software, that's what they're doing. Yeah, it sucks.
    -russ
  • It is my opinion that the preferred format for GPL hackers is the obfuscated binary format. They love to reverse engineer things so that "their OS can have the same features as any other", but the real intent is to pirate all the music and films they want.

    --

  • by Russ Nelson ( 33911 ) <slashdot@russnelson.com> on Tuesday February 13, 2001 @06:01AM (#435739) Homepage
    Every modem manufacturer I've talked to refuses to help with an Open Source driver because it's too easily modified. It's illegal to connect non-certified equipment to the public telephone network. These manufacturers don't want to participate in any way with something that's illegal. There's just no benefit to them.
    -russ
  • Whatever they are, and under whichever operating system they run, they will still be inferior to REAL MODEM!!!!! ("hardmodem")
  • It will never happen as long as equipment must be certified before it can legally be connected to the public telephone network. The barrier is not economic or technical. It is legal.
    -russ
  • Perhaps there are good ones but winmodems are notorious for being trouble makers among ISP techs. I know, I was one.

    The MWave in particular is hated. Newer versions may suck less but one of the original problems was that since it doubled as a sound card you couldn't be online and listening to sounds/gaming at the same time. It also had nasty problems connecting.

    IBM had extremely poort support for the product and indeed had a lawsuit brought up against them. No idea what happened to that.

    On the other hand there was a tech next to me that played Jedi Knight with his Supra winmodem and never had a disconnection.
  • But wait, I thought the conspiracy theories all say that Jews run the big companies? Could it be that the conspiracy theories are wrong? But there is no such thing as a false conspiracy theory (anyone proving it false is obviously a part of the conspiracy).
    -russ
  • Nonsense. The electrical part of the modem is certified, and software updates / drivers for winmodems are not certified seperate in most countries. So every update from your vendor would violate the certification too.

    Of course some countries still have such strange laws written, but they aren't followed that close anymore. Noone will sue you, if you use your winmodem with your own software. And btw: The IBM driver isn't certified as well. So what?

    You shouldn't have used your +2 Bonus for posting this, does not give a good light on you.

  • ...really,send them an email telling them how much you appreciate this move...theyll feel good and IBM will know how much we appreciate this,even if you dont have a winmodem.
    Paul Schroeder [mailto]
    Mike Sullivan [mailto]
    thankyouthankyou
  • Actually, this story is not by a Jewish conspiracy to extort monies from IBM...

    And anyone who can seriously combine the words occupied and palestine in the same sentence reveals that they know nothing about history or current events.

    A host is a host from coast to coast, but no one uses a host that's close
  • by BitKat ( 14840 ) on Tuesday February 13, 2001 @06:12AM (#435749) Homepage
    One would think that the LGPL would be a more appropriate license for driver code. Isn't it hard or practically impossible to integrate this in non-GPL open source systems?

    Not that I'm waiting for Winmodem drivers (I'll use a real modem thank you) but there may come a time that there is practically no choice (think of laptop-integrated winmodems).
  • If I'm being stupid, let me know... But shouldn't there be a way to get ALL WinModems to work under Linux using something like Wine?

    That question has just haunted me since I tried to install linux on my HP - I've a got single card that does everything - at home...

  • Expect others in the industry to foloow suit? I have a Sony Vaio and although i can still use a pcmcia modem, having the winmodem work would be really nice. Then I can actually keep my wireless card installed also. Does anyone know of any tricks to get the PCG series Vaio's winmodem to work?


    --DrMyke
  • They are the only big company that totally gets it. They have a deep commitment to Linux. On another note, they had enough guts to admit their ties to Nazi Germany. Do you think Bush will?
  • Read the following statement:
    When the Nazis took over, they siezed control of all German business. Either businesses played ball with the Nazis, or failed and were destroyed. I don't defend Nazis.

    If you're going to accuse IBM, at least go to NYU and get the facts rather then rely on this questionable new report floating around and the book on which it is based.

    A recently published book, as well as a recently filed lawsuit against the company, speculate on the uses of Hollerith equipment by the Nazi government and IBM's role.

    IBM finds the atrocities committed by the Nazi regime abhorrent and condemns any actions which aided their unspeakable acts. It has been known for decades that the Nazis used Hollerith equipment and that IBM's German subsidiary during the 1930s -- Deutsche Hollerith Maschinen GmbH (Dehomag) -- supplied Hollerith equipment. As with hundreds of foreign-owned companies that did business in Germany at that time, Dehomag came under the control of Nazi authorities prior to and during World War II. It is also widely known that Thomas J. Watson, Sr., received and subsequently repudiated and returned a medal presented to him by the German government for his role in global economic relations. These well-known facts appear to be the primary underpinning for these recent allegations.

    IBM does not have much information or records about this period or the operations of Dehomag. Most documents were destroyed or lost during the war. The documents that did exist were placed in the public domain some time ago to assist research and historical scholarship. The records were transferred from the company's New York and German operations to New York University and Hohenheim University in Stuttgart, Germany -- two respected institutions with academic credentials in this area. Independent academic experts at these universities are now the custodians of these records and supervise access to the documents by researchers and historians.

    IBM remains interested in any new information that advances understanding of this tragic era, and looks to the appropriate scholars and historians to verify it.



    A host is a host from coast to coast, but no one uses a host that's close
  • Unlike you some nerds have a sense of morality.
    Like me, some nerds know that we aren't always in full posession of the facts, and therefore choose not to come down on one side or the other based on some artice that someone wrote, that got referred to on slashdot. Do you boycott Bertelsmann? Nestle? Shell? Glaxo? Monsanto? All of these companies have been the subject of serious accusations regarding their ethics, but I simply don't have the time to investigate them all in order to make a fair and informed decision. Sometimes, if I see a story that sounds credible, and it comes from a reputable source, then I might stop buying their product. That isn't the sort of news that I, or I would guess most of the readers, come to slashdot for. There are other places to get that kind of news.
  • Call me a pedant and mod this to "Offtopic" if you want - I guess it's what I expect. But...

    The proper initialism is "ADSL" (not "ASDL"), which stands for "Asynchronous Digital Subscriber Line". It's common for people to want to put the S before the D, as in "ISDN", but it's not correct. If anything, use that as a mnemonic device - it's the OPPOSITE order of "ISDN", it's "ADSL".

    Not trying to sound condescending, but rather an attempt at being informative.



    Quidquid latine dictum sit, altum viditur.
  • MWave is not winmodem in the traditional sense.

    Mwave worked with os/2 and windows, only because these were the operating systems IBM provided drivers for originally.

    MWave relies on a programmable DSP which can be made to do anything.

    there are mWave sound cards, modems, ISDN cards (WaveRunner, my dad designed it), and ROM drives.

    The Rom drive is interesting, it's a mwave controlled cd-rom drive that can be reprogrammed into a dvd-rom drive. Very flexible.



    A host is a host from coast to coast, but no one uses a host that's close
  • If they had open specs some WinModems wouldn't be that bad. Some of them implement a lot of useful routines to help off-load a lot of work (like a dial-up accelerator card). On the other hand, some are basically sound cards.

    Brian Macy
  • http://reswat5.research.ibm.com/projects/linux/dev driver.nsf

    Linux Hardware Configuration and Compatibility Database.

    A host is a host from coast to coast, but no one uses a host that's close
  • by Gleef ( 86 )
    Marias suggests:

    But if the DSP sources would have been open-sourced as well, it would have been possible to port them to other Winmodem (and ISDN4Linux) hardware as well.

    Unlikely, since as I understand it, the ACP (MWave) modem design is radically different from most (eg. Rockwell or Lucent) "winmodems". For one thing, the MWave actually does have an onboard DSP, but it's a more general purpose one than in more traditional serial modems.

    Your suggestion is similar to trying to port a program to Windows by looking at its Macintosh Assembly Code. Technically possible, but more work than anyone cares to do.

    ----
  • All I need now is for BT to get round to actually implementing ADSL in my area. Judging by their current disgusting behaviour I'll get ADSL sometime in 2023.
  • Yup. I can't type - I'm going to join the BDA, the British Association of Dyslexics.

  • Even with a "Linmodem" driver, what have you got? You have saved maybe $1.50 in the manufacturing process and created a modem that will become unreliable under heavy processing loads. Mind you, I understand that some users of laptops are locked in to a soft modem, so this is great for them, but as somebody who has worked on Winmodem code, I really think that soft modems are not a trend to encourage at all!!!

    Also, some of the FAX state transitions must occur within certain time windows. This is very hard to ensure with a soft modem. Doubly hard when the softmodem is a USB softmodem.

    No friends, kudos to IBM for this, but please avoid soft modems if you can!
  • From one former ISP tech to another.:) Welcome brother and you are dead on right. The nice thing about the Supra's is that they had pretty good drivers. This made them usable. When I was doing ISP support we had many users with the Lucent things. We found a good set of drivers that worked with our stuff and advised users to use those they worked. The Packard Hell modems did the same thing as the MWaves and they where hellish. Some of the other ones we could not get to work. Drivers are key. My Supra worked very well. Having said that I still prefer a good old external modem over just about everything *very* easy to troubleshoot. And of course my Cisco 675 is *very* sweet. :)
  • I used to have one of those 1994-era MWave ThinkPads, and the MWave implementation was problematic, but it wasn't totally horrible. (They took a little tweaking in DOS to get set up, but so did everything in those days.)

    The cool thing was that your machine was probably originally advertised with a 14.4Kbps modem, and that was later software upgraded to 33.6K. Every competing laptop with a built-in modem was still at 14.4 and is still at 14.4K.

    The other good thing was that you didn't have to deal with the kludgy DOS/Windows PCMCIA drivers, where maybe after an hour of tweaking you'd have a 50/50 chance of getting a PCMCIA modem working. maybe. The MWave just worked and had relatively straight forward AT scripts.

    As far as the sound support went, the SoundBlaster emulation was always much better than the various ESS laptop chips of the era, and the MIDI support vastly superior to anything you could find a in a laptop. And yes, you could play sounds and use the modem at the same time (although you probably wouldn't want to use it for online gaming or streaming audio! But those weren't exactly popular applications back in '94)
  • Is that one of those wierd thingies that are kind of like a NIC? I heard that they make strange noises, but are really slow.

    Friends don't let friends operate at speeds less than 256K ...

  • Note that back in the early 90s, the MWave was a cost saving solution for IBM -- PCMCIA modems were a couple hundred bucks at retail and were pretty flakey to boot. (I don't know how much onboard laptop modems cost, but I imagine it was still substantial.)

    Modem standards were also changing rapidly, and IBM was able to go from 14.4K to 28.8K to 33.6K just using firmware upgrades - which was great for those of us who got free modem upgrades that would have cost a few hundred bucks at each step. (Not to mention that power management actually worked with the MWave, too.)

    Furthermore, IBM had to build a machine that could work internationally, and MWave allowed them to do it in software and not ship different versions of the same hardware.

    They were probably a little slow to move off of MWave when 56K solidified and winmodem type hardware got real cheap. But back in the day, I can see how MWave made sense as engineering solution for the time.

    (The MWave originally shipped with DOS and OS/2 support. Later NT and Win95 support was added. So it never really was a "Winmodem".)
  • This is an very good point. When a modem is certified, both the actual hardware and the software that drives it (whether it be uploaded software in this case, or software on a ROM chip) must be certified. If something changes in the firmware, you generally need a new certification.

    At the last company that I worked for, we had a modem driver written by another company (it was a software modem) for our game console, and we were trying to write a new driver for it (since we wanted bugfixes, etc.), so we had to check out the legal documentation. It seems that the FCC here in the US had certified the existing combination and didn't allow for changes, but the Canadian government didn't quite care as much (I don't know why).

    I believe (though I am not 100% sure) that much of the reasons for these requirements are the regulations for line load and interference. If you take too much of the line load, then you'll render all phones in the house inoperable, and if you allow too much interference, well, that pretty much speaks for itself.

    Joe
  • Quoting the AC out of obscurity:
    That's a terrible argument!
    Damn straight that's a terrible argument. Quite frankly, I have no idea of what to do with the linux kernel source [besides compile it that is], but that doesn't mean that I don't want to be able to have it.

    The better argument for releasing the binary-only DSP software is that this software doesn't actually run on the host computer so it could be described as part of the hardware. For those of you saying that you need to be able to modify the DSP code so that you can use a modem as a video card, get a grip. If you really want to do that kind of thing get a soldering iron and buy parts better suited to the job.
    _____________

  • No silly, now people will have to resort to a more proper term- "hardware deficient modems"
  • I assure you that binary .dsp files aren't the source code.

    Let me begin by saying I am one of the most die-hard free software advocates out there. I have yet to read any article by RMS that I disagree with.

    That said, your complaint that the dsp modem algorithms, which would typically be present in hardware, should be given to the world in source form, is enough to give even me pause.

    So-called hardware algorithms, written in microcode and etched in silicon, are pervasive in the computing world, not just in modems, but also CPUs, hard drives, network cards, video cards, and a whole lot more. The free software community has not demanded release of such silicon code in the past. Whether or not we should is a different question. I suspect that if you put the question to RMS directly, he would have to advocate free-ness of the code. However, even the FSF purchases and uses hardware containing such (closed-source) silicon code without compunction.

    I myself think (and I may even find myself disagreeing with RMS on this one, although to be fair I haven't asked him for his view) that the need for free software stops as soon as you start talking about software that is so integrally tied up with the hardware, that you would need whole new hardware to even contemplate making use of changes in the software. Processor microcode, hard drive error correction algorithms, and yes, modem dsp code, all fall into this category. I do not require the availability of Intel Pentium microcode, Seagate hard drive error correction code, or IBM modem dsp code when purchasing hardware, and neither should you.

  • mcc writes: Would the "give poor performance and kill the CPU" argument still hold for those of us who use modems primarily for file transfer, and never run anything more interaction-intensive than IRC

    Well, assuming that the sources know what they are talking about, then no. The winmodem can't suck up enough cpu/memory so that games would take a hit, therefore, we must assume that the overhead a winmodem imposes on a system is rather small. (Which makes some sense, since cheap winmodems work on a p133 or so.) Therefore, if you are running something less intensive then a 3d online game, presumably, you have the resources to spare for a winmodem. Thus, if a winmodem cost you less then a hardware modem, you come out ahead.

    Just my $.02
  • Not only do you need the source, but you also need a compiler for the particular DSP used by the modem.

    Yeah, ok for anything other then "understanding" the code you also need to compile it. Of corse a big old datasheet for the DSP would let someone hack up a version of gcc, or a perl script that may be able to compile things. It is much harder to turn a big binary glob into useful source.

    Driver source with binary firmware glob -- nice start. Driver source with firmware source (and bin) even better, but not perfect. Driver source with firmware source, and firmware compiler source way better (let's assume the firmware compile is written in a language we already have a compiler/interpreter for).

  • jesus fucking christ.

    why don't people read the fucking moderator rules.

    whomever rated this "overrated" is obviously a fucking moron.

    there is a "post at 2" threshold you stupid fuck. if you are above a certain level your posts default to 2. get a fucking clue

    [/rant]

"For the love of phlegm...a stupid wall of death rays. How tacky can ya get?" - Post Brothers comics

Working...