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

 


Forgot your password?
Close
typodupeerror
×
Hardware

First, WinModems. Now, WinWiFi. 386

Posted by michael from the wireless-cards-to-avoid dept.
zulux writes "Microsoft is actively encouraging WiFi (802.11b) hardware manufacturers to strip their devices of costly electronics, and use Microsoft software/drivers to make up the slack. And you thought WinModems were bad!"
This discussion has been archived. No new comments can be posted.

First, WinModems. Now, WinWiFi. More

First, WinModems. Now, WinWiFi.

Comments Filter:

  • Great... (Score:3, Funny)

    by NanoGator ( 522640 ) on Saturday April 13, 2002 @12:06AM (#3333500) Homepage Journal
    ... Now Tom's Hardware can benchmark AMD's running WiFi.

  • Feature bloat ahead (Score:4, Interesting)

    by MxTxL ( 307166 ) on Saturday April 13, 2002 @12:10AM (#3333512)
    Software is just too intensive to use for low level operations. It's SOOO much faster to have it in the hardware. Sure, software can offer a lot more flexibility, and it might keep some costs down, but that hardly makes up for the performance loss.

    Plus, with the flexibility comes the idea that it's ok to write in more and more features... software bloat is the result.

    • hardware vs software as a tactic (Score:4, Interesting)

      by Alien54 ( 180860 ) on Saturday April 13, 2002 @12:29AM (#3333569) Journal
      Software is just too intensive to use for low level operations. It's SOOO much faster to have it in the hardware.

      MS is depending on Moore's law to save them again. And this seems to be a long term strategy - to convert hardware to software, which ties things into the windows OS again.

      Another secret of bloatware is reveiled.

    • Using CPU cycles can only go so far (Score:5, Interesting)

      by MtViewGuy ( 197597 ) on Saturday April 13, 2002 @12:43AM (#3333603)
      I think it's a terrible idea!

      Despite the speed of today's CPU's, having to use CPU cycles to do WiFi networking is not a great idea, especially when you also have to take into account for CPU cycles being used for everything else in the system.

      I mean, consider the situation of playing DVD discs on a computer. Sure, you can do it completely in software if the CPU is fast enough, but the CPU cycles it requires to do this even on a very fast CPU can drag a system down pretty quickly. Now you know why ATI has Hardware Motion Compensation (HWMC) and Inverse Discrete Cosine Transform (IDCT) decoding assistance on their graphics chipsets starting with the Rage 128 series, and nVidia has pretty much done the same with the current GeForce4 MX/Ti chipset series.

      • Re:Using CPU cycles can only go so far (Score:5, Funny)

        by DunbarTheInept ( 764 ) on Saturday April 13, 2002 @01:12AM (#3333682) Homepage
        You're forgetting the One Microsoft Way of thinking - Nobody needs to have enough machine to run more than one app at a time. If you buy more memory or CPU speed clearly the only possible reason you could have for doing so is to run a bigger application, never to run multiple applications - running more than one thing at a time? What, are you some kind of Unix propeller-head?
      • Despite the speed of today's CPU's, having to use CPU cycles to do WiFi networking is not a great idea

        Unless your Intel, looking to give an excuse to Joe Consumer to upgrade.

    • It's SOOO much faster to have it in the hardware.

      Ok what is the exact definition of "hardware"?

      Reason I bring this up is because most "hardware" almost allways uses some type of software, but instead of being stored on a floppy or hard disk it's stored on a ROM.

      This is true for everything from routers to motherboards. It's all hardware running software. So why is this a bad thing to have WiFi? We've seen linux drivers surface for winmodems, and certainly they will surface for WiFi as well.

      When you're living off of ramen, to quote oddtodd, "You realize that money is important to do stuff" I'm for anything that can take a $200 dollar 802.11 basestation and reduce it's price to $15.

      • WinWiFi != WinModem (Score:5, Informative)

        by Pizza ( 87623 ) on Saturday April 13, 2002 @09:40AM (#3334656) Homepage Journal
        As someone who spends a considerable amount of time these days hacking on WiFi card drivers, Host-based MAC is actualla a VERY good thing.

        A good analogy of this is PPP. The current situation is similar to a modem manufacturer embedding PPP in the hardware, which is horribly complex and expensive to implement. It is much simpler and cheaper to let the OS provide the PPP services.

        WinModems come in two flavors; host-based controller and host-based signal processing. The latter is pure evil; the hardware is nothing more than a A/D/A converter, and the host CPU has to perform all DSP functions to make it into a modem. The host-based-controllers have real hardware DSPs and whatnot, but they just tell the DSP what to do, essentially replacing an on-board processor+firmware with the driver on the host machine.

        WinWiFi (which is really host-based-MAC) is neither. The WinWiFi card would become about as smart as the average ethernet card; ie it would be able to transmit and receive raw 802.11 frames, and then pass them off to the driver which then figures out what to do with them.

        A good portion of the wireless cards out there already do this, and nearly all of the new ones will do this. Why? complexity and cost.

        802.11 is rather complicated. The MAC must handle a complex state machine; with all sorts of little nuances. Handling transmits/receives, and their acks, association, channel hopping, and then the real doozey: encryption.

        WEP sucks. Not just because it's fundamentally broken, but because it takes a bit of oomph to work with, and it's a little complex. And if this is done in hardware, you can't update it to handle newer standards.

        Every single one of the 802.11 extensions to replace/augment WEP will require considerably more computation power in hardware; but in fact, most 802.11 (windows) drivers now do WEP on the host, because it has far more computational power to spare with zero additional hardware cost.

        This WinWiFi initiative is nothing more than "hey, all of you guys have already written this host-based-MAC stuff (or are going to have to write it anyway) so why not just use the stuff already part of the OS? It's already been extensively tested and that way, you don't need to reinvent the wheel."

        It's called shared code, and makes a lot of sense.

        I've been banging my head against the wall a lot lately because of buggy firmware in WiFi cards; If they let the host OS do the work instead, these bugs wouldn't exist, because the 802.11 spec is well-documented.

        And again, it's not WinWiFi, it's Host-based-MAC. It's a work-in-progress for Linux too. And it is a GoodThing(tm).

        - Pizza
      • Ok what is the exact definition of "hardware"?

        The part of the computer you can kick when the fucking thing just won't cooperate.

        --saint
    • I am not sure I agree. Any given piece of hardare (and WiFi is probably a _really_ good example of this) has two parts (1) the core functionality associated with the hardware itself and (2) the associated "computation" to make the hardware work.

      For example in the WiFi case, there is the hardware need to trancieive and then all the other stuff needed to make whatever is received into a useful stream of data (maybe even only analog if youhave A/D hardware) or vice verse.

      Now within certain limits, the marginal cost of adding CPU processing power is probably less than the cost of adding the first bit of processing power to the hardware solution. Horses for courses if you lke. CPUs are good at being CPUs and WiFi hardware is good at WiFiing. So let the CPU compute and offload the need for the addition of complexity at the hardware level.

      I think the use of the CPU in the machine to doo all the CPUing required for the system is an, in principle, good idea in some sense.

      Having said all that. Hardware is a good way to do a job that does not change much so maybe all this is kinda irrelevant.

  • Gah! (Score:2)

    by zephc ( 225327 )
    Great... wireless ethernet over COM1?

  • Aren't they a little late to the party? (Score:5, Interesting)

    by Cutriss ( 262920 ) on Saturday April 13, 2002 @12:10AM (#3333515) Homepage
    First, WiFi devices have been out for a few years now. Yeah, hardware modems had been too, but the markets are moving faster now than they had been 10/15 years ago. Furthermore, there's already a new big player in the WiFi market that won't stand and let Microsoft have exclusivity on WiFi drivers...Apple [apple.com].
    • Since when has Microsoft being late to a party ever been reason enough for them not to crash it anyway?

      Microsoft's business model, in case you haven't been reading for the past few years, is to have not only their finger in every single pie, but to cut off all the other fingers already there. This is they how and why of the windows monopoly.

      Pray they don't get into bio-engineering.

    • Not too late... Microsoft's WinWiFi is perfectly compatible with other WiFi devices so there's not really an inertia holding device manufacturers back from making these..And they will make them, just as they made WinModems...

      Personally I think software-based WinWiFi is a great idea...The vast majority of users buying new computers are just wasting a mammoth amount of CPU anyway that might get put to some use if they do wireless networking. Don't get me wrong, its not for everybody and I know there are benefits to a hardware solution, but for a certain segment, these will work just fine, as WinModems did. I wish the software were based on something more open than what Microsoft is offering, but since nobody else but Apple has stepped up to the plate on software WiFi (and their solution is also closed) then we can't really whine about it.

      • Apple isn't using any kind of "software Wi-Fi"; AirPort cards are just OEM WaveLAN (er, Orinoco) cards.
      • You are wrong for a number of reasons:
        1. WinWiFi would have the same problems as winmodems. Initial glitches, etc. and incompatibility with other operating systems.
        2. If other hw manufacturers do it good ones like Lucent will be forced to just to keep price competitive.
        3. Vendors make one chipset and stuff it into every wireless device they sell. Tear apart an access point or airport and see what they're using for a wireless card.
    • Apple, actually, sells a large amount of WiFi compatible chips. There's no way Lucent is about to give up hardware copatibility. Especially since the wireless chipsets they use in their PC cards are the exact same as those in their embedded devices (RG/APs).

  • WEP + MS - oh great. (Score:3, Insightful)

    by Anonymous Coward on Saturday April 13, 2002 @12:10AM (#3333516)
    As if WEP wasn't insecure enough as it is... Microsoft making it even more exploitable is just what we need... I can just immagine an 802.11x code-redish worm floating around... sounds like fun to me!

    Just like MS to try and steal the thunder of something popular after the fact (coughcoughnetscape).

  • Software WEP? (Score:2, Interesting)

    by Anonymous Coward
    I wonder how much of an overhead 128bit WEP will put onto the WiFi "software". It already slows down some hardware cards, so using the host CPU really doesn't seem to be a good idea.....although when has MSFT ever been worried about my privacy?
    • RC4 is very fast; I doubt software WEP would cause a noticeable slowdown on today's CPUs.
    • The WEP standard chose to use the fundamental crypto algorithms in badly broken ways, but if you don't do stupid things to break RC4's security, there are some things that are very nice about it. It's dirt simple, and the little complexity it has is in the initial key setup. Key setup time and code size don't change significantly for key lengths between 8 and 248 bits, and the data encryption phase is all byte-based so it doesn't depend on the key length at all.

      The WEP standards group was trying to avoid US Pretending-To-Be-Anti-Communist export laws (and also French and Chinese policies), so they took a conservative approach and designed a system that could be used with short known-to-be-easily-cracked keys, but could also be used with medium-but-still-inadequate keylengths or acceptable keylengths if you set the correct values in the ROMs. Of course, like Microsoft PPTP before them, they did so in an incompetent manner without adequate adult supervision, so their work was shredded by some of the same people at Berkeley who helped shred the initial RC4/40 "export-quality" code, and who also shredded the GSM Telephony incompetent encryption algorithms over lunch one day.

  • Wannabe Hardware Company (Score:3, Funny)

    by Skavookie ( 3659 ) on Saturday April 13, 2002 @12:12AM (#3333522)
    Perhaps Microsoft wants to be a hardware company and has decided the easiest way to do that is to turn all hardware into software.
  • remember what a pain winmodems are/were for linux? almost impossible to get them working, at least in the old days. if m$ is successful in getting wireless companies to use software instead of hardware, could that be the end of wireless for linux?

  • I dunno... (Score:2, Insightful)

    by mesozoic ( 134277 )
    I think any company today who makes a "Soft WiFi" card will recognize that they're cutting out a serious chunk of their potential customers. It's not like the peak days of WinModems, where Linux users were a negligible percentage to the consumer hardware industry.

  • The bad part is... (Score:2, Insightful)

    by Chayce ( 199487 )
    ...people will go for this, because they dont know any better and because it's cheap. I just hope this doesnt become industry standard, because it will mean a step backwards instead of forwards, and because this is an obvious ploy by microsoft to push their domination of the OS market, anyone who's ever tried to install a winmodem in linux knows that. Oh well, thats my 2c.
    • Comment removed based on user account deletion
      • How does offloading the work from dedicated hardware to a more expensive CPU help the consumer? Last I checked people paid more money for faster cpus because they wanted faster computers, not because they wanted periphial manufacturers to save a couple bucks on the components in their modem or wireless card.

        I just don't see how using 2% of a $120 cpu to do the work of a $1 chip is in the consumers favor. It seems like all this does is increase manufacturer profit while marginally reducing consumer cost.

        Personally I could care less if microsoft endorses or pushes winWiFi as long as my option as a consumer to pay the extra $5 for a hardware based WiFi card remains. If the winmodem is their case study, I say we have nothing to worry about since last I checked you could still get a hardware modem years after winmodems hit the shelves.
  • Look at the numbers... Most linux installs are for servers and high usage machines. You NEED linux for a reason. Most wireless installs are the for opposite reasons, light duty machines and the non-tech people that don't know the different between a 5 meg wireless link or a 100 meg switched network.
    • Eh? I know a lot of people with wi-fi setups. Not one of them is a non-tech person. Wi-fi is still so new that only the techies get into it. And believe me, they realize the limitations of the technology. And if you're accessing the Internet over a DSL or cable modem, that 5 meg wireless link is plenty good enough.
    • If you have an extra linux box, adding a wireless NIC to it basically converts it into a high-end access point. Something that would cost about $1000 if purchased seperately.
  • The only benefit to the consumer would be a slightly reduced cost. This proves that microsoft cares about us. Personally, I like my hardware to be OS specific, because it prevents me from doing foolish things such as installing so called "free" software (which is un-American, btw).

    Ever try using a winmodem in linux? LOL, or in windows for that matter?
  • Fine... Someone's going to step up and write drivers for Linux... and that's not going to be the issue... Just like I'm really annoyed by flying icons sending rubish to the trash can eating up system resources... Sure very little but still... Having to crunch radio input is just sending more good hardware to the scrap sooner by making our processors do ALL the work. I'd hate to have a winmodem, winsoundcard, winwifi all plugged in at once... This is also isolating the hardware that does it inclusively into a higher price bracket as well... which would get a lot cheaper if we just leave the standard alone! Look at what you can get a good 10/100 card for these days.

  • Open-source troubles again (Score:5, Insightful)

    by jquirke ( 473496 ) on Saturday April 13, 2002 @12:24AM (#3333559)
    Some of you have commented on the possible performance implications of "soft" WiFi, but there is an even bigger issue, the same reason we hated WinModems so much.

    If the software routines / hardware API is kept proprietary, which is likely the case, us Linux/FreeBSD/other open-source OS users will be left in the dark.

    Either [a] hardware vendor thinks they will look good and support Linux by releasing a binary-only driver that is only compatible with kernel version X, and needs to be hacked to work with anything else, FreeBSD users like myself are out of luck (and anyway I would _never_ use a binary-only driver in an open-source kernel - hence my gripes with NVIDIA).

    or [b] some of the brave of us attempt to reverse-engineer Windows drivers.

    Either way, consider the next wave of laptops coming with built in "soft" WiFi - a definite possibility considering the amount of money manufacturers could save, and offer WiFi standard even on their lowest-end models. This means chances are we have to fork out and buy a traditional PCMCIA hardware adapter. And a lot of us run Linux/FreeBSD/whatever on our notebooks, I know I won't be happy. I think I'll be paying the $US45 for an 802.11b card while I can!

    Which raises another interesting point - you may think "yeah there will always be hardware PCMCIA WiFi cards". But look what happened to 56k modems - try and find a 56k modem on a PCI card that isn't a soft-modem!

    Of course this is not bad for everybody - the new cheap WiFi will be more widely spread since 99% of computers run Windows NT/Windows anyway, and this good be a good thing for prices of WiFi cards,etc.

    --jquirke
    • Sorry for incorrect use of terminology - s/PCMCIA/PC Card/
    • I don't know if cheap soft-802.11 cards will be a good thing for the price of the real stuff. After all, in this age of winmodems a Courier V.Everything still costs almost $300. Moore's law? What Moore's law?
      • Moore's Law has to do with transistor count. Nothing else.

        And while a courier modem will cost a bunch of dough, $300 is excessive. You can usually get a nice extermal v.whatever modem for $100 or under.

        • $300 isn't excessive for commercial grade equipment. The Courier is a well-designed modem. I've seen people try to save a few bucks by buying a cheap modem and then waste thousands of dollars on trying to make the cheap modem work properly, only to end up tossing the cheap modems and replacing them with Couriers.
    • In softmodems the hardware is little more and A/D and D/A converters - the software driver needs to perform complex signal processing algorithms. The resources required for developing and testing these algorithms are probably beyond the capabilities of open source developers.

      I don't think a soft WiFi card will continously receive bits and let the software do everthinng else - the hardware should still be capable of decoding packets, matching the MAC address and detecting CRC errors. The software will need to do the encryption/decryption and the algorithms for network detection and handshake, transmission speed power control and perhaps some other housekeeping stuff. This doesn't sound so bad.

      Actually, in a card without firmware there may be less places for sneaky undocumented features than in a card that simply exposes the bare hardware capabilities to the host.

    • some of the brave of us attempt to reverse- Ah ah. (Score:4, Informative)

      by crovira ( 10242 ) on Saturday April 13, 2002 @08:03AM (#3334379) Homepage
      Naughty naughty...

      There's the DMCA (or whatever floated to the top of the "Alphabits"TM bowl of some congressman's breakast that morning,) to slap you down with if you even try that.

      Don't you know yet that YOU have no rights?

  • Does it matter? (Score:5, Insightful)

    by JabberWokky ( 19442 ) <slashdot.com@timewarp.org> on Saturday April 13, 2002 @12:27AM (#3333563) Homepage Journal
    Yes, you're looking at a situation where you're trading off CPU power for the operation of a peripheral. I don't really like that myself. But really - how does this affect Linux?

    Everybody seems to be making the assumption that there won't be drivers. Why not? Linux has a small but appreciable market share, and that market share is more apt to get WiFi than most other users. Unlike the situation when WinModems first came out, there is a viable base and thus economic incentive to release Linux drivers.

    Now, let's hope they come with source - too many chipsets require that the end manufacturer can't release open source drivers. mda_hal.o and the like are workable, but not optimal - to a certain extent, open source drivers for software driven accessories like the so called Win* hardware makes it *more* powerful for the open source realm, where talented hackers can alter and upgrade the drivers to drive the hardware beyond the original specifications, purposes and features that were originally designed for it.

    --
    Evan

      • Linux has a small but appreciable market share, and that market share is more apt to get WiFi than most other users

      Ooh, so wrong. Linux has a decent share of the server market, which is 100% wired. It has some desktops, mostly wired. But the 802.11* market is laptops. Sure, I'm writing this on my couch with a (wired) laptop running SuSE, and I once heard of a guy in Portland using a linux laptop as well, but he may be a myth. I mean really, what percentage of laptops are running linux? 5%? And what percentage of corporate laptops are running linux? 1%?

      We know that it costs nothing to release technical specs, but apparently manufacturers don't know that. Even manufacturers like Creative - who have a linux driver project - can be very slow and reluctant to release specs (no? I still can't get my Creative Webcam Go to work reliably). I'm guessing that we'll have to prod and poke and guess and hack WinWiFi like we had to do with WinModems for years. :(

    • What are you afraid of? (Score:5, Informative)

      by Gumber ( 17306 ) on Saturday April 13, 2002 @01:55PM (#3335594) Homepage
      What is the chipset used in most WiFi cards? The Prism 2/2.5.

      Who makes it?
      Intersil

      Are there linux drivers?
      Yes! With full source!

      And guess what, Intersil comissioned the drivers!

      Not only that, but the drivers offer support for advanced functions typically not offered on Windows based PCs (host based access point support).

      So, based on past history, there seems a good chance that there will be a path to Linux support for WinWi-Fi cards.

  • Hmmm, I wonder what (Score:3, Insightful)

    by nzhavok ( 254960 ) on Saturday April 13, 2002 @12:27AM (#3333564) Homepage
    the processor usage of this will be? Given that using a 56k winmodem can take a noticable amount of processor time, what will is it likely to take up in these high-bandwidth devices?

    Personally I haven't had any bad experiences with winmodems, I've only had one (Lucent chip) and it seems to do a fair job in my linux gateway for browsing, but forget games!

  • This isn't necessarily evil... (Score:5, Interesting)

    by Sivar ( 316343 ) <charlesnburns[@[ ]ail.com [']gm' in gap]> on Saturday April 13, 2002 @12:35AM (#3333583)
    One way to reduce hardware cost is to put hardware functions in software.
    You don't see anyone calling "monopoly" about software RAID cards, and those that do pay far more (andget only marginally better performance) from hardware RAID.
    Winmodems may be a PITA for us, but you can get then for $5, vs. $70ish for a hardware modem (the 3Com Performance Pro comes to mind)
    I can see that Microsoft may look at this as another opportunity to extend the duration of their doomed monopoly, but honestly I don't believe that they are morally obligated to keep hardware prices up by NOT integrating their functions into software. They are, after all, a software company.

    Does it not make sense to introduce new stolen ideas to make more use of software?
    Besides, these are Microsoft drivers. They'll probably be slow enough to help the ailing hardware industry sell a few more chips. That's aid that they could use now.
    Yes, I know it isn't kisher to say that not *everything* Microsoft does is evil. Mod me down if you like.
    • I still don't think it's a good idea.

      The problem is that even if the software WiFi takes advantage of AMD's 3DNow! Professional or Intel's SSE2 multimedia extensions to speed up operation, the CPU cycle usage would be considerable, which may drag down the rest of the system. No thanks!

      By the way, today's winmodems are way better than the original 3Com/US Robotics designs. I've played with a winmodem that uses the PCTel chipset and it does actually work quite well, thanks to the fact the PCTel chipset does use at least the MMX registers on newer CPU's to keep its performance reasonable.
      • Get the fuck out of here.

        I'm philosophically against this too, but claiming that offloading the processing of an 802.11b card onto a 1.xGHz processor is going to "drag down" the system is a steaming pile.

        It is actually a good idea from a the perspective of operating the cards. The less going on in the card the cleaner your signal will be.

        OTOH, from a hardware peripheral point of view it is plain stupid to tie your device's ability to operate to a particular runtime environment. One would be wiser to have a clean and simple interface to simplify writing the driver. From a peripheral manufacturers point of view a driver is an expense that doesn't generate revenue (generally).

        -Peter
    • I'm curious what the reaction here would be if a different company developed this? If some linux company pioneered this would the posts here be about how linux saves you even more money and the triumphs of open source? I think so. Of course that's depending on whether this product and the hypothetical one work well, are stable, aren't hogs, etc.

      Its really not a bad idea for most part. The people buying this are not going to be hardcore wifi enthusiasts, but home and business buyers with CPUs on almost eternal idle.


      • I'm curious what the reaction here would be if a different company developed this? If some linux company pioneered this would the posts here be about how linux saves you even more money and the triumphs of open source? I think so.

        I have to disagree. The grinding of teeth you hear is not a knee-jerk reaction to the word "Microsoft", as you imply. It is a reaction to the memories of "winmodems" this idea brings up. Remember, politics aside, this site IS frequented by the more technically minded. And software modems have a far less than sparkling reputation. Linux only adds to the pre-existing list of grievances.


        The people buying this are not going to be hardcore wifi enthusiasts, but home and business buyers with CPUs on almost eternal idle.

        Right. In other words, people who don't know any better.
    • Oh man please tell me where you get your dope from as I havent had anything that good in years. Lets see the difference between microsoft or linux soft raid vs. my dual channel 64bit DPT u160 raid card with a 5/0 array. Guess which RAID system wins?
      And this guy gets a 4?
      • Yes, obviously a high end RAID array is going to need a real RAID card, but we aren't talking about server class hardware here.
        Guess what, if a server has a modem it will probably be hardware based as well.
        Since we are talking about consumer products here, my RAID example was more referring to two and four channel IDE RAID.
        That said, if you want to be not-picky, Linux and FreeBSD andeven Windows 2000 have very good software RAID support--good enough that they would indeed compete with your hardware RAID array if the CPU wasn't completely pinned doing other things.
        In fact, software RAID has an advantage in that it has a large, powerful CPU to back it up, whereas most hardware RAID cards (yours not included) have relatively wimpy processors and can be a bottleneck on large RAID5 arrays. Software can also support mixed SCSI/IDE drives without complaining, thoguh this wouldn;t be optimal.

        Anyway, back on topic now, take a Winmodem and a hardware modem. Guess which wins? Neither. There is nothing about either that inherantly makes it faster, and the .5% of CPU clock time it uses isn't significant. Even if MS's hardware drivers eat 30% of a GHz Athlon, big deal, the ordinary consumer can't tell the difference between a K6-2/500 and a AthlonXP 2000+ in most apps, and CPUs will get faster anyway, as will the drivers.
        Remember the first winmodems ate quite a bit of clock time, but now it isn't even on the radar, the only concern is that Winmodems aren't multiplatform.
        We can't rationally tear down a technology just because it doesn't support Linux or just because it was made by Microsoft. Even Microsoft has come up with some really good technologies. I can't think of any at the moment... Anyway, you're probably just a troll trying to brag about your expensive hardware so I'll stop this long, boring, partly offtopic reply now.

  • Impressive irony, even for Redmond (Score:5, Interesting)

    by John Miles ( 108215 ) on Saturday April 13, 2002 @12:35AM (#3333584) Homepage Journal
    It wasn't that many years ago that Microsoft, along with just about everyone else in the PC business with an ounce of common sense, launched a jihad against Intel's NSP (Native Signal Processing) initiative.

    NSP was the logical response to Intel's realization that CPU cycles in the Pentium era were becoming less and less valuable to the end user. They considered it a task of strategic importance to soak up extra cycles wherever they could be found... never mind that game developers still needed every cycle they could find at the time. Had NSP succeeded, it would have had a wide array of effects on the PC hardware and software businesses, almost all of them too ugly to contemplate. The nascent market for high-performance 3D and environmental audio hardware would likely have been crushed under the treads of Intel's marketing machine, and WinModems would have taken over the scene years earlier than they did. The development of online gaming technology would have been pushed back indefinitely, pending the ubiquitous adoption of broadband (which, obviously, has yet to happen).

    Of course, MS's primary interest in killing NSP was to keep Windows from having to run as just another NSP client. Owning the boot process from BIOS to bluescreen was as important to them in 1994 as it is now. But now, it appears that they've taken leave of their technical senses as well as their ethics. If this is anything like Intel's earlier push to run modem data pumps on the CPU -- and to be fair to MS, the article is by no means clear on this point -- then 802.11 fans, and consumers in general, should fight it where they find it.

  • mini monopoly (Score:2, Insightful)

    by catwh0re ( 540371 )
    interesting way of making your 802.11b device only work on windows... imagine it if you don't activate your MS software they can not only disable your PC, but your entire network. fun fun fun
    • interesting way of making your 802.11b device only work on windows... imagine it if you don't activate your MS software they can not only disable your PC, but your entire network. fun fun fun

      *imitating Beach Boys*

      And we'll have fun fun fun til MS takes the network away!

      (yes, it is past 2 am here)

  • Not likely (Score:2, Insightful)

    by nmos ( 25822 )
    I don't think too many companies are going to fall for this. Remember most of them use the same cards in their base units too and these don't have anywhere enough extra power to run Windows let alone a winnic.
  • I wouldn't call this "all bad". Fundamentally, the idea seems pretty smart: move all the things that necessitate expensive chips over to the CPU, and lower the price of the finished product. Granted, when you make this proprietary to one OS, it sucks. But the kind of computing power available to the masses today is just ridiculous overkill. This was the case a year ago, and it's even moreso the case now. Why reinvent the wheel for every peripheral you have when most of the processing can be offloaded to the CPU? I wonder how much money you could save if you could buy a WinGeForce3 (granted, this is a stretch, even with today's computing power), WinRAID controller (which is actually what the HPT series of IDE RAID controllers are, as I understand it,) WinSoundblaster Audigy, etc. (When I say "Win" I don't mean "runs in Windows," but rather "runs in software.")
  • WinWiFi would definately slow down Linux/FreeBSD/FreeOS support in the short term, but think about the bright side. People now are using winmodems as a phone line interface; it gives us free OS users a tool we wouldn't otherwise have. We may have the same thing with WinWiFi: imagine if you could adjust the speed at which RTS/CTS/ACK/broadcast are sent, or send certain packets with a PIFS interframe spacing, or change aSlotTime... Maybe even make fundamental changes to the MAC, such as CEDAR.

    This could be really great! Can you imagine Linux DoS tools based on flooding frames without participating in the MAC?
  • I don't think this is going to be as bad as WinModems. With WinModems, everyone and their cousin had a different way of dealing with emulating hardware with their drivers, so it was almost impossible to replicate functionality on Linux. If Microsoft is offering some kind of standard library that emulates the hardware they'd be removing from their boards, that means you have a documented API you can write an implementation of for Linux, and it wont' be as hard to make drivers for these WinWiFi devices.

  • 802.11a already has this problem (Score:3, Informative)

    by billstewart ( 78916 ) on Saturday April 13, 2002 @12:51AM (#3333624) Journal
    I've been told by somebody working at an 802.11a manufacturer that the specs for that are designed to have significant parts of the system run on the host computer's CPU rather than the card, and that therefore there'd be issues with getting Linux drivers unless the manufacturers funded them.

    Some of the concerns are the amount of processing horsepower required for security and maybe also for some of the communications functions, since it's easier to add computational horsepower when you're not crammed into a small card competing for space and heat load with the radio circuitry, and also convenience in upgrading the system, especially if upgrades may require even more substantial increases in CPU crunching, such as bigger RSA modular multiply/exponentiations.

    The importance of convenient upgrades has been amply demonstrated by the repeated failures of WEP :-)

  • WiFi explained, with helpful comparisons (Score:4, Funny)

    by steveha ( 103154 ) on Saturday April 13, 2002 @12:55AM (#3333635) Homepage
    From the article:

    The radio, about the size of a can of beer, extends the wire line and connects with any mobile devices

    "...the size of a can of beer"? I love it! Let's keep this going.

    Mobile devices can use a PCMCIA card, slightly smaller than a Hershey chocolate bar. Pocket computers can use a CompactFlash WiFi modem, slightly bigger than a "Fun Size" Hershey bar.

    The WiFi base station connects to your computer, which of course is bigger than a bread box.

    The wire line is your telephone line, which is about the size of a really, really long strand of spaghetti. This connects to the telephone office, which is about the size of a telephone office. This in turn connects you to the Internet, which is sort of hard to measure the size of... let's just say it is the size of the whole world and be done with it.

    Hope this helps.

    P.S. I wonder what percent of Slashdot readers actually know how big a bread box is?

    steveha

  • Server market (Score:2, Insightful)

    by Tyreth ( 523822 )
    I think Linux/Unix is already established well enough in the network server market that this will be no problem. Companies will still have to produce network cards with full electronics - or produce Linux drivers - ensuring that Microsoft will fail if this is a strategy to help push away competitor operating systems.

    Winmodems were a barrier for linux, but network cards are cheap and people will probably consider them inferior in design if they won't work on Linux. Linux servers are becoming popular in the average house, very well respected. This won't be seen as a weakness of Linux/Unix, but instead a weakness of the card itself.

  • Why use WIN modems when you can use GNUmodem?

    Ok, its time to start the GNUModem project

  • How much cheaper can they really get? (Score:4, Insightful)

    by sterno ( 16320 ) on Saturday April 13, 2002 @01:25AM (#3333709) Homepage
    The logic behind winmodems was to reduce the hardware costs and drive down the prices. My question is how much cheaper can they really drive these prices down. Right now you can pick up a wireless ethernet card for $50. Modems are runing as low as $30 for comparison.

    So as demand increases, quantities of scale continue to increase, we can expect the cost for those same cards to come down. It's unlikely that WiFi cards will be able to press much further down in price even with using software drivers.

    Another thing to keep in mind is that most of the wireless manufacturers tend to save costs by reducing redundancy in their wireless products. If you use a Lucent AP-1000 access point, it runs on the same cards that you'd put in your laptop. I have yet to see a wireless adapater for a desktop that wasn't, in reality, a PCMCIA slot with a wireless card. It's a big cost savings to them to only have to manufacture one set of devices to fill their needs in laptops, PC's, and access points. Trying to do software drivers would totally screw up these possibilities.
    • The logic behind winmodems was to reduce the hardware costs and drive down the prices. The logic behind winmodems was to reduce the hardware costs and drive down the prices.

      That may be true... but of course, if prices stay the same but this reduces manufacturing costs, then profits go up.

  • Hacking possibilities (Score:2, Interesting)

    by Anm ( 18575 )
    Am I not the only thinking about the bountiful hacking possibilities of a DSP controlled radio transciever card? Of course that assumes the card APIs are reverse enginnered (or pigs fly and the specs are published).

    Anm

  • Big Winmodems Event! (Score:2, Informative)

    by puppetluva ( 46903 )
    Conexant (formerly Rockwell-and one of the biggest winmodem makers) just released a lot of their drivers for linux with half-source/half-binary drivers for Mandrake and Redhat. (thanks to the hard work of Marc Boucher)

    http://www.mbsi.ca/cnxtlindrv/index.html

    The whole Winmodem thing isn't all about Microsoft evil, by the way, its about patents (that should be your second guess for sources of evil after M$ by now). My understanding is that Winmodem drivers expose the code for V.92 and other compression/transmission implementations.Because of this, the makers aren't allowed to open-source the code for these patented implementations.Think about it this way, the regular hardware modem makers aren't exactly shipping you microcode and chip diagrams in the back of your manual either.

    For the first time, I'm using the modem that came with my 2 year old Vaio at 56K as I type this. (thank God modem/speed technology has-gone/is-going nowhere!)

  • == Apple's "Software Base Station" (Score:3, Informative)

    by rheiser ( 42454 ) on Saturday April 13, 2002 @01:46AM (#3333745)
    This sounds suspiciously like the "Software Base Station" available on Macintoshes for a number of years now (surprise, surprise!) It allows you to use a computer with an AirPort card to act as a Base Station for other computers with AirPort cards, instead of spending the $250 to buy a dedicated one.

    rheiser

  • Wrong Interpretation? (Score:5, Informative)

    by __aadidx2690 ( 313265 ) on Saturday April 13, 2002 @01:48AM (#3333748)
    Judging from what I've read elsewhere, the submitter may have interpreted the article a bit wrong. It's not so much that MS and Intel (also mentioned in the article) want to have the WinModem equivilent of 802.11, but that they want to make the access points cheaper by providing a software solution.
    Apple has had a similar product, the "Software Base Station," available for Mac OS 9 for quite some time!
    See this (much better) article [eet.com] for details.
  • Back when we had Pentium 100s with 16MB of ram and WIndows 95, the windows modem concept was a clear winner for the bean counters but dammit, it sucked the life away of the machine.

    Now we are in the land of the 1+ GHZ Celeron with 128MB of ram. The overhead of the winmodem should be tiny, unless the drivers are horribly written.

    Not that I give a crap, years ago I decided to bite the bullet and get a hardware modem that I eventually made work in Win95, Win98, WinME, NT4 Workstation, 2000 Pro, SuSe and RedHat.

    The average WiFi card for a laptop right now is around $100. For $100 you can buy an Apple Airport or a Linksys WPC11. If companies start pumping out soft cards with less electronics that rely on a fat driver then the windows user can expect to pay a fraction of that cost. I doubt we are going to see $20 Lucent WinWiFi cards any time soon, but there is going to be a sweet spot in the price chart that is going to help with increasing the popularity of WiFi.

    We have a bunch of early adopters at my office and so far people love being able to walk around the house with a laptop when they are telecommuting. I added a Netgear ME102 to my home network in December and use a WPC11 for my laptop and I like it so much sometimes I don't ever step into my home office when I telecommute. Had the WiFi card been $50-$60 instead of $100 I could have bought it a month or two earlier, plus it would make it easier to convince the IT folks at the office to shell out for a test base and a few cards to do a field test.

    Now, we are always wary of Big Bad Microsoft getting their hands on anything, but dammit, this standard is already open, and non Microsoft entities are huge players. Apple bases all their wireless networking on the same standard! Making a cheap, reliable windows-only wireless card does not affect Apple since they are a niche shop. It does not affect the open source folks since there will always be a full hardware solution, just like we have always had real modems sold alongside winmodems. And there is always an enterprising soul that wants to figure out how to make a winmodem work under Linux, so let's be honest, I know theres a few people out there eager as hell to give it a try :-)

  • The TI Win/4 printer (Score:2, Informative)

    by BernaMa ( 569516 )
    Did you remember the TI Win/4 printer? It works only with w95, eating the 5% of CPU time while idle.
    The genicom.com [genicom.com], who bought TI printers line, wrote:
    <CITE>
    There is not and will not be a Win/4 driver for Windows 2000, Windows NT, Windows 98, or any other operating system.
    Customers requiring operation on these platforms are advised to purchase a more versatile printer.
    </CITE>
    Alas, this stuff wasn't reported on the Win/4 box...
    Who has time enough to develope a linux driver for it can raise a lot of money buying them for 5$ each *before* to release the driver!
  • I think the whole 'soft' device argument is a red herring. DSPs and ASICs cost cents each in large lots, and their elimination means a CPU costing several hundred dollars has to make up the slack. Whereas these DSPs/ASICs are optimised to handle the task at hand, the CPU is not, meaning a slowdown for any other apps you're concurrently running. So whilst the manufacturer may save $0.50 per unit, each consumer suffers with decreased performance. Bah! Fight the insanity!

  • Future SoC Design (Score:3, Informative)

    by LoRdTAW ( 99712 ) on Saturday April 13, 2002 @02:56AM (#3333885)
    My first reaction to this topic was "Oh god here we go again" but then I sat back and thought a little.

    First think of the popularity of CNR and AMR sound, ethernet, dsl and modem cards you have seen for sale. Answer none. It just never took off because noone wants to sell cheap hardware when they can make more money with a real hardware solution. At one point they litterd every intel bord I ever came across and then only appeared on cheap bargin pc motherboards. I have been hearing about soft dsl for a long time and I have never seen a soft dsl card, also all soft dsl cards support the G.light standard which no provider cares about.

    Second face it, There is alot of money to be made with hardware. The 3D graphics market took the exact opposite approach to this problem. the first popular true 3d polygon game was quake. now the old dos quake ran in total software, everything from the AI to 3d graphics was done on your cpu (back then I ran it on a 486 with a p83 overdrive) and now look Nvidia took the whole graphics pipe and threw it on a chip which is totally opposite the software approach. Some people would sell there own mother for a geforce 4 if they could. Shure today the cost of CPU's have come down enough to justify the $400 tag on a GF4 Ti but take away that pipe on a chip and do it in soft. People would scream bloody murder. And sound cards are going this way too. Pretty soon we might be shelling out 200+ for a sound card with a APU (audio processing unit) that will imerse you in a whole new world of sound

    And last relax people with SoC (System on Chip) design coming along nice these days I wouldent worry. There are already all in one SoC's for DSL modems and cable modems. I imagine a cheap WiFi solution is in the works as we bitch.

  • Monopoly Leveraging (Score:3, Insightful)

    by Anonymous Coward on Saturday April 13, 2002 @02:58AM (#3333893)
    And let's not forget that Microsoft has a brand new way of leveraging its monopoly: Driver Signing.

    ..So a manufacturer announces that they're going to make a software WiFi card. Knowing that a significant portion of their market base are people running Linux, they publicize their intent to make both Windows and Linux drivers.

    Now they go to Microsoft to get their driver signed. Behind closed doors, Microsoft says "Nice product! We don't approve of the Linux driver though. Make your product Windows only, or we won't sign your driver." If they refuse, then publically, Microsoft claims that the driver didn't meet Microsoft's standards of quality for a kernel driver. They both have a defensible excuse, and can smear the uncooperative company.

    Now the company is faced with a business decision. Face the 95% of their customers who use Windows and tell them to "Just click okay" when Windows says "This driver isn't signed! It's really, really bad to install unsigned drivers," thus reducing their image in front of their customers. Or, don't release a Linux driver, and save face with their Windows clients.

    ..And it can only get worse. How long until Microsoft doesn't allow unsigned drivers to be installed in the name of reducing their tech support costs?
  • "A Microsoft representative was
    unaware of any additional details
    about what was to be presented."

    With the problems I have getting any support from their pay support for winmodem compatibility, this is such a telling statement.

  • The difference between WinModems and WinWiFi would be that Linux is much more popular now than it was when WinModems were relevant. If the good hardware manufacturers went WinWiFi, they could probably be convinced to allow LinWiFi drivers. I can't imagine Lucent (who actually offers LTwinmodem Linux drivers now) not doing that, for example.

    I think Lucent almost gets it. Other companies I'm not so sure of. Make sure to vote with your wallet, manufacturers have more incentive to listen now than they ever did.

    Oh, and for all that people whine about hardware not being supported in Linux, I actually have hardware that works under Linux but is completely unusable in Windows 2000. Eat it.

  • No, I didn't think winmodems were necessarily bad (Score:4, Interesting)

    by MadFarmAnimalz ( 460972 ) on Saturday April 13, 2002 @04:25AM (#3334048) Homepage
    Where I live, buying a new lower-end pc sets you back the equivalent of 2 months average salary. A hardware modem costs on average 5 to 6 times as much as a winmodem.

    Consequentially, it's winmodems that have people round here online in troves. I currently have both kinds of modems, having been forced out of linux's charm to buy a hardware modem, but most people are not like that; with most people, getting online and being a netizen is a priority overriding hardware design ethics and operating system chauvinism.

    If wifi ever takes hold in this country, it will only be if they're cheap; that can only be helped if there's a software layer somewhere in there saving you some moolah.

    Soaking up CPU cycles? C'mon. Even in a power-user thick forum such as this one, how many people utilise their cycles beyond 10 or 20% over time? Distributed.net and SETI@home don't count, mind you.

    Seems to me the only question here is whether we will go through the same heartache we did with winmodems, what with closed chipset specs and chipset makers digging their heels in not to release such information. This seems to me to mainly be an issue of profit margins: what makes more money, hardware solutions or their corresponding software emulations?

    Generally, a more expensive product is more likely to carry a larger profit margin for many reasons. The higher complexity of the product acts as a kind of barrier to entry into the market segment freeing up the supplier to play a bit with the price, and there are always economies of scale, even at this level.

    In other words, the per-unit profit is likely to be higher for hardware solutions. Now the question has become one of pure demand and supply; are the incremental profits from a hardware solution greater than the incremental volume generated from a software solution?

  • I've the worse (Score:4, Funny)

    by jsse ( 254124 ) on Saturday April 13, 2002 @04:27AM (#3334050) Homepage Journal
    WinBoss - cannot work on anything but Microsoft Windows.
  • One possible outcome of this is a fork in the wireless data protocols if the "real" protocols are too expensive (patents etc) or too difficult to implement. I do not find it hard to imagine a scenario where somebody says "screw it" and writes a lightweight packet radio implementation using just enough of the hardware to get by and inventing a protocol with real security (none of this WEP or 802.11x crap). Add an instant D.I.Y. gateway (mini PC with OS of your choice) and voila.

    Of course, that assumes that getting enough info to talk to the DSP etc is possible. I guess the far more likely outcome would just be more pain and hurt for non-M$ folks (but that's what M$'s objective is anyway!). Sigh.

    Windows XP already downloads new runtime firmware to my wavelan card... I discovered this because it broke my old base station that didn't support link layer fragmentation with WEP enabled. I had to update the firmware on the base station to get it to work again after installing XP.

  • Great opportunity (Score:2, Interesting)

    by mc6809e ( 214243 )

    This is actually a good thing for those wanting to experiment. Think about being able to alter the software driving these things. What might be done? Might alternate coding schemes be used? How about your own encryption method?

Slashdot Top Deals

UNIX was half a billion (500000000) seconds old on Tue Nov 5 00:53:20 1985 GMT (measuring since the time(2) epoch). -- Andy Tannenbaum

Close