Become a fan of Slashdot on Facebook


Forgot your password?

Compaq Announces Thin Client Running Linux 88

ansible writes " Saw this story on Compaq has announced thin client hardware, including one that runs Linux. " The most interesting thing is the fact that the thin clients have a PCMCIA port and 2 USB ports. USB? Wouldn't it be swell if Compaq had some code for us? (he says eyeing the unusable USB port on his desktop and laptop)
This discussion has been archived. No new comments can be posted.

Compaq Announces Thin Client Running Linux

Comments Filter:
  • by Johannes ( 33283 ) on Friday September 10, 1999 @07:00AM (#1689991)
    Take a visit to the somewhat out of date Linux USB homepage: []

    Both USB Host Controllers (UHCI and OHCI), Keyboards, Mice, Printers, some bulk devices (Hard Drive's, Floppies, etc) and some webcams have drivers off the top of my head.

    With many more to come and it's all in 2.3!
  • by Dr_LHA ( 30754 )
    Given that USB is not really a stable thing in Linux yet - what USB stack are Compaq going to be using on this device? Other than that it sounds pretty cool.
  • by sterwill ( 972 ) on Friday September 10, 1999 @07:03AM (#1689993) Homepage
    I have a laptop, and I use a USB mouse just fine. It's even a PowerPC laptop, so I'm using a Linux kernel configuration Linus largely ignores... and it still works, with kernel 2.2.12.

  • Hate to reply to my own message - but I meant to add that Caldera Openlinux has a built in USB stack based on the old USB stack (pre-Linus version) - and wondered how well that worked... i.e. is it good enough for this thin client business....
  • by pb ( 1020 )
    Look, it's a taint-free thin-client machine!

    It's interesting to note that they couldn't manage to cram a decent web browser onto Windows CE. It's a good thing we have Linux instead.

    Compaq is okay in my book, provided they don't mess up the Alpha too much. It's nice to see them not completely locked into the Microsoft vendor path.
  • I would defninitly like to see some of that USB code. Does anyone know of anything like the Aero 8000 that runs linux well. I could realy use a thin client laptop that I can use around campus and run porgrams remotely.
  • by Wee ( 17189 ) on Friday September 10, 1999 @07:08AM (#1689997)
    ...if Compaq decided to give back instead of cash in? It's plain that they (like many others) are looking at Linux as a fad they can make money from rather than a better way of doing things. In a way, I don't mind so much, because the more they cash in on Linux, the more Linux there is and I think that's a Good Thing.

    But the USB deal is just one of many things that is keeping Linux from taking greater hold in the desktop market. Once Linux gets things like more games, USB, cutting edge hardware/driver support, etc. then people will consider Linux as a viable alternative to Windows or Mac. I think this is an important step in the evolution of the OS. If Comapq could see their way clear to releasing some code for their USB implementation, then it would certainly help, and others might follow their lead.

    I'd really like to see a big name company like Compaq seriously embrace Linux and give back what it develops.



  • The Linux model will also be designed for users who desire a product without the "taint" of Microsoft.

    You should know that this is the "Texan" usage of the word taint.

    Example: "Bubba, can you come here and fix my Windows terminal? 'Taint workin' no more."

    disclaimer: I'm from Texas. I can make these jokes.
  • I didn't see any mention of USB modems (like my USB ModemBlaster)--anybody know what the status is there? I can't determine if it's just a serial modem hanging off a USB port or a WinModem working through a USB connection.

    (FWIW, the only reason I bought it is because every time I open a computer at home disaster strikes. No case-cracking to install a USB modem, and everybody's happy.)

    --- Chris
  • Now that compaq has announced its thin clients, I begin to wonder what people will do with them. Are thin clients that remote productivity solution of the future, or just a junk solution that will technically obsolete just after they hit the market?

    -- Moondog
  • by skroz ( 7870 )
    What's the estimated cost for one of these?
  • Thin client ? Isn't history repeating it self ? I thought thin client proved to be a failure 2 years ago. Why should it work now ?. Linux is ment to be used on a workstation not a thin client. IMHO. Stable USB code would be nice though.
  • by Johannes ( 33283 ) on Friday September 10, 1999 @07:35AM (#1690005)
    Yes, USB modems are also supported. Forgot about that.

    They aren't a serial to USB device, nor a WinModem, but they do understand AT commands. They show up as serial devices.

    Printers, OTOH, are parallel to USB devices. But, I digress
  • by josecanuc ( 91 ) on Friday September 10, 1999 @07:39AM (#1690006) Homepage Journal
    If I recall correctly, there was a discussion on the LinuxPPC mailing list about this. Apparently, Linus is ignoring the USB patches being sent to him by the PPC folks. Now, I'm not bashing Linus, apparently he doesn't want to add it to the tree until a stable version exists and there is another group working on it. Since Linus isn't too fond of non-Intel architectures, he will probably go with that other group (sorry, can't remember it now, but the link was posted earlier on this board)
  • by Anonymous Coward
    I wouldn't get too excited. I work at Compaq's ZKO facility in Nashua, NH, where most of the work is done on Tru64 Unix. I have a close friend who was using a USB keyboard and mouse on his Tru64 Unix 5.0 system, but stopped because he found it to be... uh, a bit lacking. Doubled keystrokes and such. The USB driver gal is just down the hall; I hope she's busy coding...

    That said, I have no idea about USB in Linux.
  • Ummm.....a working USB mouse doesn't mean the USB support in Linux works right. Some of the chipseeets have some of the USB functionality, but I don't believe there is a completely-implemented interface for any USB configuration.
  • Then it's not a sexy thin client, but an "old stodgy X terminal" and we can't have old buzzords, now can we? :)
  • I'm from Texas too, and I know that a lot of Linux users and /.'s have had bad experiences with Compaq; me too.

    I do think Compaq is making efforts to support Linux better and I know that they will take a lead in developing Linux on their Alpha and Intel platforms, but yesterday's post about their new Alpha compiler has me worried.

    I see them falling right back into their previous trap with SCO and other Unices. The funny thing is that the developers and hackers in Compaq are strong advocates of OSS. They work for a company that simply doesn't understand the model.

    I've written them many times about drivers for SCSI cards, tech. spec's for video, IDE, etc. and come up with nothing. I garuntee you they'll support Linux well, but I don't see them contributing to the community in a generous fashion. Thank goodness for GPL.

    Are any /.'s also Compaq kin? I'd appreciate knowing a little more about the internals of Compaq's Linux gear-up. Please help us help Compaq! Their servers really do scream.

    Spoze that's all I got swurth sayin fer now.

  • They're probably just using the standard USB support already in the later 2.2 kernels... I can't imagine them developing something proprietary on the side and not even releasing it.
  • by Anonymous Coward on Friday September 10, 1999 @08:05AM (#1690013)
    Wyse has had a thin client running linux for a couple of months now. I have been testing it for about a month. It looks pretty good, but, and I swear this is true, I have had to reboot it when setting it up as many times as thin clients running WinCE. It also doesn't remember the time unless you point it to an internet time server.
  • by um... Lucas ( 13147 ) on Friday September 10, 1999 @08:08AM (#1690014) Journal
    It's plain that they (like many others) are looking at Linux as a fad they can make money from rather than a better way of doing things.

    They're a business. Their job is to make money. If they make money by doing things better, that's great. Right now, they like everyone else that's hopping on to the Linux bandwagon (Oracle, Lotus, IBM, etc...) the interest is twofold.

    1 - it lessens their dependence on MSFT
    2 - it's because customers are asking/demanding it

    That's good enough for me. A company that decides that it's going to screw the notion of profitablility in order to only do things in a "better" manner is the one that won't be around next year.

    I'll again reiterate my argument that while Windows NT is extremely vulnerable to the threat of Linux, Windows 9x is less so, and the MacOS is even less threatened... Comparing NT Server to Linux showcases NT's bloat. Comparing Linux to Win9x shows how much further Linux must go. I really don't think that Linux/KDE/GNOME/etc will ever approach the current MacOS in terms of ease of use. We're not talking stability or anything else. That's not a fault of Linux, per se. Just it's heritage from Unix, plus its' development by programmers for programmers aren't exactly "user-friendly" in the way the Mac is for the computing neophyte.
  • When you hear "Compaq and Linux", it's natural to think "oh, they are just cashing in".

    But you've got to remember that there's a little division of Compaq from Maynard, Massachusetts. Ever hear of Digital Equipment Corp.? Ever hear of a The Man They Called maddog ?

    Of course, one only need to look at Compaq's stock performance [] to see that DEC has, ahem, infected Compaq's culture. I would hope that Linux, seemingly the last refuge of the desperate in corporate circles, might be able to help them recover a little glory.

    And yes, they are expected to continue to give back to the community as well.

    Bravery, Kindness, Clarity, Honesty, Compassion, Generosity

  • by trims ( 10010 ) on Friday September 10, 1999 @08:11AM (#1690016) Homepage

    I sincerely believe that Thin Clients are the way of the future, especially for businesses. That said, it seems that people are forgetting a couple of things when they impliment them these days...

    • You really should do rendering on the client. It used to make sense to do it on the server, since CPU was expensive, but now days, with really cheap high performance CPU and Video, it's stupid NOT to do it on the client.
    • Bandwidth. Alot of this is related to the above (drawing/rendering on the client cuts down seriously on rendering requirements), but please, Mr Thin Client maker, a dedicated 100Mbps to the desktop and a Gigbit ethernet for the server is ludicrous (hello, Sun???). Do some work, and at least have a protocol compressor on both ends (as there are for X).
    • Memory and swap. OK, I know you shouldn't need swap space for a thin client (everything runs on the server, right?) but it sure would be nice to have local caching for some stuff. Barring that, please have enough local memory to do proper caching? I see alot of Thin Clients with 8-12MB, which isn't anywhere enough to do decent caching. Again, bandwidth this the limiting factor here...

    WinCE, while in and of itself isn't a horrible thing for Thin Clients, depends on Window NT TS, which is one incredibly nasty hack. Personally, I would shoot anyone suggesting we use a NT TS solution.

    Linux is great for Thin Clients. Hopefully, the Compaq clients above are well-designed (basically, you should have a sophisticated X-Terminal). The thing here that Compaq and other thin-client makers need to pound on hard is: APPS, APPS, APPS. Being able to run StarOffice is a godsend, but they need to provide alot more to make a Thin Client truly useful.

    With the backing of Compaq, I see this as a good thing for thin-clients. Hopefully, we will see alot more software developers produce X-based apps for the community at large.


  • by Anonymous Coward on Friday September 10, 1999 @08:13AM (#1690017)
    he works where I work and he is more than willing to write USB drivers for any devices he is provided with. He has done a lot of work getting the USB code stable and in good maintainable shape. He has written drivers for USB printers, Zip drives, Web CAMs, Speakers and Modems. Look in the kernel source for 2.3.x for his e-mail and send suggestions.
  • by Anonymous Coward on Friday September 10, 1999 @08:18AM (#1690018)
    Compaq has now introduced an exciting new USB peripheral for the Linux version of their thin client computer.

    Known as "APOBET" (A Piece Of Black Electrical Tape) it will be applied to both USB ports on their new thin client system, to enhance the appearance of the machine and enhance usability (by preventing user confusion).

    The APOBET will also be available as an Upgrade Kit for Compaq customers moving to Linux from the Windows platform. A company spokesperson has been quoted as saying that the APOBET is also being tested for use as a retrofit for machines running Linux across the whole line of Comapq computers.

    A Linux spokesperson was quoted as responding that "This fulfills the need for a robust, timely USB solution on the Linux operating system. In addition, it yet again proves that with Linux stability always comes first."
  • If you want stable USB code, stop whining and code it yourself. The specs for OHCI are available; even UHCI specs are available for a price. If you're too lazy to do it yourself, just use NetBSD which has had working and stable USB code for months before Linux did.
  • ...its' development by programmers for programmers aren't exactly "user-friendly" in the way the Mac is for the computing neophyte.

    and who wrote the MacOS.... elves?

  • I worked for Compaq for a while, and while I wasn't working on anything Linux related, I need meet and talk to some of the people on the team... They all seemed like really great guys, and most of them wanted to be able to realease some GPLed code... here's the basic idea:

    Compaq is a company, and anything they do has to be justified in terms of $.

    In order to "help satisfy the customers needs" Linux is being "supported" but not "encouraged".

    I'm not sure how much of what I know is confidential, so I'll leave it at that...
  • by _Sprocket_ ( 42527 ) on Friday September 10, 1999 @08:35AM (#1690024)
    and who wrote the MacOS.... elves?

    Internal documents outline the use of Elves at Apple in an unpublished commercial. They were going to dance around in shiny, colorfull suits. Then someone scuttled the idea with the "tanks and supercomputers" angle.


  • PCMCIA is a really nice technology for laptops, but it doesn't make much sense to put it in a desktop machine. PCMCIA peripherals are more expensive, and you get less selection. SCSI and USB can give you any peripherals that PCMCIA can, and then some.

    Why did they do this? Is the PCMCIA socket stuff cheaper? It's hard to imagine that it would be, since there are all those commodity SCSI cards and intregrated-SCSI motherboards on the market. Does anyone have any ideas? Or is this just an unqualified mistake?

    Beer recipe: free! #Source
    Cold pints: $2 #Product

  • by Anonymous Coward
    Something important to consider here is that even if a company were to heavily modify the Linux kernel (such as adding USB support) it would behoove them to release it to the public because in time, the linux kernel most likely will get that support anyway, and the code bases would conflict. Why maintain your own copy of existing, conflicting, device drivers? Every new kernel patch becomes a maintenance nightmare and it gets only worse with time. Initially there may be reluctance to release custom code, but in time it becomes the only practical alternative.
  • ...on my iMac running NetBSD and LinuxPPC (depends on the day and what I'm a-doing as to which gets booted)

    Get with the program, x86 folks ;-)
  • >But you've got to remember that there's a little >division of Compaq from Maynard, Massachusetts. >Ever hear of Digital Equipment Corp.?
    >Ever hear of a The Man They Called maddog?

    Well yes and no. maddog is leaving Compaq, but I think he's going to VA Research so he can do even more Linux evangelism.
  • generally speaking compaq hates intel, not ms altho obviously there's no loyalty in this biz

  • He did say "by programmers for programmers". MacOS was presumably developed by programmers for users. If you're going to quote somebody, it might help to make the attempt to understand what they're saying first.

  • fyi it uses slackware
  • by Anonymous Coward
    Since there probably aren't any regular "slots" in such a machine (ISA/PCI) it probably answers the concerns with regard to 'expandability' (always an issue in the PC market, even when less than 1 in 10 machines actually use it) to put in the PCMCIA. For a thin client machine, it doesn't really make that much practical sense, though. The real pipeline in and out of the machine is going to be it's ethernet.

    Maybe PCMCIA was really cheap to enable, which makes sense if this hardware is some sort of an adaptation of laptop hardware.
  • I think he was refering to unix as created "by programmers for programmers" this is obviously not true of the MacOS.

  • by BJH ( 11355 ) on Friday September 10, 1999 @08:47AM (#1690034)

    No, no, no, no. Linus didn't reject the USB patches because they came from a non-Intel platform. He stated quite clearly that he didn't like the way that those patches tried to do everything all at once for USB, and wanted to take a more gradual and simpler approach. That's why the USB support in the current kernels is the one that Alan Cox and others are working on.

    Since when is Linus "not too fond of non-Intel architectures"?

  • And how is it different to the old TS?
  • Well Microsoft had the Halloween Douments leak. I imagine Macintosh will soon be publicly denouncing claims that elves write their software as claimed by the Christmas Docs. Santa was certainly good to Steve Jobs that year.
  • Obviosuly you have never used the new TS.

    How "new"? We use Citrix MetaFrame / TS pretty extensively around here, and it has some ... issues. Nothing too bad, but I can't imagine it being so good that someone would post in it's defense.

    Here are some of my proposed Windows Terminal Server / Citrix slogans. (none of which are as cool as 'X Window - You'll Envy the Dead'):

    • Terminal Server - Easier to install and configure than X!
    • TS - Mostly compatible with most common applications!
    • TS - Fewer patches released means fewer to install than regular NT!
    • TS - Do you need to spend another pile of money this year for internal political reasons?

    Please note that it's not all that bad, and it is better and easier to manage than a whole mess of desktops - but it's not in any way wonderful.

  • I agree that KDE/GNOME etc have a long way to go in terms of matching the Mac OS GUI (and Mac OS X goes a long ways towards matching the utility and versitility of a command shell - because it IS one).

    But weren't we talking about GNUStep the other day?
    Mac OS X GUI - somehow related to NeXT Step GUI - somehow derived to GNUStep?

    What I'm trying to get at here is, couldn't GNUStep be used to hobble together a GUI and tools as nice on newbies (and NHW, or Not Hacker Wannabes) as Mac's?

    "The number of suckers born each minute doubles every 18 months."
  • If you want working USB drivers you might want to try NetBSD.
  • With PC prices below $500.00, why you just set up a Linux box that boots off a server. My old school, West Chester University [] did this to their old, slow, Sparc Classics. They all run off of either a Linux box or bigger Sun Solaris box. They run great now. Its even better than an X Terminal; it gives you the best of both worlds.
  • Kind of doubt that will happen. I did some of the hardware design on the Aero 8000, and it contains a companion chip from Eclipse (my old company) for which too much IP is involved to release any useful specs (like register addresses, functions). All drivers were given to CPQ as binaries. The companion chip handles PS2, serial, PCMCIA, CF, and I think IrDA. If you want to ask them about this, go to Eclipse's website [] Don't blame me, I'm just a hardware guy. Now I'm working at Hitachi, and I believe someone has ported Linux to our SH4 development system, so just wait for a product based around that to come out. ps: The Aero 8000 has no USB
  • I think linus isn't fond of infecting the Linux tree with a "jump in the fire" attitude.

    And ahem.. if he wasn't fond of other platforms.. then why are there axp, arm, sun, and yes.. even ppc ports? Without the compromises, and platform independant code he wrote.. there'd probably never be the quick of migration.

    I remember that the 68k port was a fork from linux for a LONG time. PPC will need to cleanly layer into the linux source tree.
  • Or maybe the device was designed by (HP, NCD, etc) and they just left it in. Wyse's devices are designed by HP.. really. And they're pretty open about it.

  • Before everybody shoots me on sight, I better warn that I'm very much an OSS freak. I've pounding away like mad at my company for the last 3 years trying to get them to go Linux.
    Having said this:
    TS is actualy quite cool. If you know what a kludge WinNT is behind the scenes, you have to respect how well they managed to separate the GUI from the rest. It uses very litle bandwith : +- 4 clients run reasonably well on a 64 kbit line.
    With metaframe It gets better: I ocasionally have to fix a (extremely) bloated in-house access app running on a remote TS server through a 19.200 line and theres little diference from running it on localy (yeah, I *know* that's not saying much)
    Yes it crashes, at least as much as NT, but it actually works as advertised.
    For M$-products, that *has* to be a first.
    If one has to admin NT machines, then I'd rather they're TS: At least I don't have get up as much to fix the damn things ......
    No, I can't spell!
    -"Run to that wall until I tell you to stop"
    (tagadum,tagadum,tagadum .... *CRUNCH*)
  • I would shoot anyone suggesting we use a NT TS solution.

    I suggest you use a NT TS solution.

    *standing back*
    (just checking)
  • I have been surprised how little coverage it has gotten. It seems to be a VNC type box. It virtualizes the screen, sound, microphone and 4 USB ports (only supports keyboards and mice at the moment) back to a centralized server. It looks very slick with the smart card based access (you can just use a user id/password if you want.) It really makes all these other 'thin' clients look really chunky.

    The server side software is priced between 250 and 2,500. It is of course a Solaris only thing at the moment. You can get to your Win applications thru a Citrix client running on the Solaris server.

    If they can expand the USB support beyond the keyboard mice you would have a very slick system. I would assume that they are planning on this as the device has 4 USB ports.
  • Plus it would be illegal for them not to, under the GPL.

  • Bang!!!!!


  • Compaq recently released a GPL Linux driver for their SMART2 array cards - which in my book is huge because older Proliants with CPQArrays are laying all over places I've worked.

    The worst part of Compaq is the corporate culture that refuses to admit that (except for the high end server stuff) they just outsource from the same PC parts bin as everyone else and no longer makes their own NICs, SCSI cards and so on. Thus it is absolutely impossible to determine actually what hardware is in a Compaq from the documentation, which is laden with terms such as "Compaq Business Audio" (actually an ESS chip), and "Compaq NetFlex 3+" (actually an Intel NIC) and so on.

  • Sounds like you are advocating a traditional X Terminal instead of these "framebuffer" thin clients.

    Aside from the Does-it-work-with-WTS? question -- I'm not quite sure why X Terminals seem to be dying off (even at Sun). Perhaps X is viewed as too complicated to set up and admin? Not robust enough? Too many 486s to install Linux on for those who need real X?

    (As for huge bandwidth requirements - in most cases LAN bandwidth [switched 100 or 1000] will be cheaper than a bunch of PCs to manage.)
  • I remember the post saying Linus didn't accept the patches because he couldn't understand their stack so he wrote his own. Just what I remember, so don't quote me on it.
  • This sounds like a game console with usb. Anyway, i think that the only thin clients in homes will be consoles with a few more features that hook to the tv or to a monitor.
  • >They're a business. Their job is to make money.

    Um, no. Their job is to increase shareholder value, which is not necessarily the same thing. It could be that becoming part of the Linux/Open Source/GNU/GmOne [] movement will mean increased long term profits, even at the expense of short term profits.

    Companies are supposed to be fairly long-sighted entities. If they're not, it's bad news.

  • by Anonymous Coward
    Hey, I work for Compaq. I've found that Compaq does listen to suggestions. It's a pretty open environment from where I sit.

    What would people recommend that Compaq could do to support these fine people at in their efforts? I'll suggest it!

    In a way, just building a cheap thin-client machine with the 2 USB ports and offering a Linux solution creates synergies. If you build it, they will come...

  •, why bother with a Linux client if you are dependent on an NT server... hardly the way to avoid the "taint of microsoft!" (or to provide reasonable stability...)

    Maybe the second generation will be a little more thoroughly though out...
  • They're already having a hard enough time digesting digital. Right now, and for the forseable future, windows is the premiere for them platform to support.

    There is no concrete proof that Linux is more than a passing craze by the ABM crowd (Anyone But Micrsoft)... Windows 2000 isn't out the door yet, so, according to the many reports I've read, it's opened a window of opportunity for Linux. What if Win2000 ships with all the features Microsoft has promised, at the top of which it being their most stable OS ever? There'll be a lot of companies who've embraced Linux backing up (even just a little bit) in order t say that Win2000 is their favored platform. I don't mean to be a downer or anything...

    With all that said... I hope Linux thrives. I just don't think that a company with years of foundations laid developing and supporting other platforms, should jump head over heels for Linux. Wait it out and see what happens in the next few years.

    Lastly, I can't see how supporting Linux/etc... will actually increase long-term profits. Buyers are wise... They see a WinNT server selling for $3500... Even if $100 of that is the profit margin from them preinstalling NT and $500 is the actual NT cost, they're going to ask why they should spend any extra money above the bare-bones cost of the hardware. After all, most likely they'll just recieve the computer and then re-FDisk it in order to set the partitions the way that they see fit... A few years from now I may realize I just put my foot in my mouth, but today I feel comfortable with that statement.

    Shareholder value goes up in accordance with a companeis ability to make money, unless you're a .com company... But even those will be fewer and more far between.

    Am I off topic here? Hope not!

    Have a good night, everybody...
  • "by programmers for programmers"

    I was criticizing this comment. I meant that programmers wrote MacOS, just like Unix. What makes you think that the same type of people(programmers) have two different styles for different people.
    You see you make it sound like users wrote the MacOS for themselves and programmers wrote Unix for themselves.
    Programmers like to write elegant solutions. Unix is just a little bit older than MacOS and was also designed as a solution to different problems. But I think you would get alot of feedback from the people at Gnome or KDE about programmers writing for programmers.
  • I have been surprised how little coverage it has gotten.

    /me too.

    It virtualizes the screen, sound, microphone and 4 USB ports (only supports keyboards and mice at the moment) back to a centralized server. It looks very slick with the smart card based access (you can just use a user id/password if you want.)

    Especially the smart cart feature is interesting: You can pull the card out of one box go to the next room, plug it into another ray and you can immediatly continue the session exactly where you've left it.

    (Even if you treat your Ray with your Beaked Axe of *Slay* Hardware (2,6) (+8,+11) (+2) very badly you just have to get a new one and haven't lost even your last keypress. :-)

  • Some points:
    • PCs are noisy
    • PCs consume more energy
    • PCs need more room
    • You have to setup a PC before you can use it. (This adds $$$ to the price!)
    • PCs (especially the cheap ones!) tend to fail more often

  • "I sincerely believe that Thin Clients are the way of the future, (...)" I agree. In most networked environments people have always used (X)terminals. They're some sort of thin-client. Everybody is now acting as if thin clients are something new because people are used to windows/mac-based systems. Those systems were meant to be used as stand-alone systems, but nowadays most systems are networked and the technology is out of date. I think thin clients are going to take the world back again because they're a cheaper and simpler solution. Not only to buy, but especially to maintain.
  • This thin client drama that has emerged recently, is in fact quite astonishing to me. How can anybody remember history and still make it sound so new? Let me make my point here.

    I remember a long time ago when computers were extremely expensive things. They designed computers with software meant to be mult-user because it was the only way one could make computing efficient. The way one would communicate and interact with the computer was through a node or a `terminal'. These often got called _dumb_ terminals, because they were essentially a screen, a keyboard and some kind of interface to the computer system. It so showed that these systems were extremely efficient and stable, because all the effort of making it stable was put on the back end. The risk of a monitor getting a hickup wasn't all that great. In fact, in many places these kinds of computing scenarios are still in place and far from being exchanged.

    Then came the PC from IBM. The PC was the breakthrough because it was cheap and reasonably efficient. Now each employee could have their own computer. Since most data was processed through linefeed printing rather than databases, this was a good thing. People liked having `their own' computer. They could finally keep their data to themselves. Microsoft tried to mimic the file and directory structure of the Unix filesystem (UFS), but failed immensely.

    Ever since, the PC has been just that. A _personal_ computer. Each person has one. This becomes expensive since each person makes a claim of a computer for at least $1200. Calculate that with a 100 people and that is well over a million bucks. Then you still need a server to all of the sudden _SHARE_ data. This is of course a salutation because most organizations today make use of databases and printer sharing, but still. Computing is still computing. Wheather people today like pretty mouse cursors and colorful buttons, better than a green and black monitor. Actually, even I can understand that... Now Sun and many other companies have solved this by using X-terminals. A screen with a little box with rendering capabilities. The thought here is once again to move all processing to the backend, thus making the need for a desktop computer less important. It is also better for upgrades because your X-terminal doesn't go out of style. We still use X-terminals from 1989 and they work great. Wanna speed up things? Upgrade _ONE_ server.

    But as soon as Microsoft announces something (not entirely) new, the TERMINAL Server version of NT, this gets out of hand. Let me tell people that this has been around for ages. No longer can we remind people that Microsoft is using old and thought-through technologies, but lets name them _thin-clients_. A thin client is still a PC. With a slightly smaller demand for hardware, but still has a CPU, a VGA board, a possible sound card, a harddrive and memory. IT IS STILL A PC! A cheaper one perhaps, but still. Moving semaphores and most of the computation to the backend is merely a fallback to the old and tested way of doing it.

    Enough about history, and to the point. What is Compaq's goal with this? Are they making a Linux version just for the hell of it, or are they in fact going to do some work on it? Are they looking to integrate the Windows CE version with the NT Terminal Server and the Linux version with their AlphaServers on Tru64? As far as i understand, if this is supposed to be called a thin client in the sense of it being a `terminal', it must be integrated with some other product for the backend...



  • Their is one loophole in that: they could release their code as a binary-only module. AFAIK, this would be perfectly legal.

    (This doesn't mean that Compaq will actually take this option, of course.)

  • Noone is "claiming" that Terminals are new. It's just they're trying to come into the mainstream.
    Windows Terminals are certainly new - the idea isn't - but it's new to windows - even citrix based stuff like winframe/metaframe/multiwin are all only a few years old.

    What you're saying is synonomous to saying "what's all this hype about the internet - we had it a 100 years ago - it's called the telegraph".
  • This is what Apple is trying to do with netboot.
    The server is only a "online" disk.
    The programs are running on the clients.
  • Fuck you "Stop whining and code it yourself" parrots piss me off!

    Can't you come up with a response that's at least original, if not pertinent?

    Not every user wants to code their own shit, or has the programming ability to do so, it's the nature of the world - so get over it!
  • Thin clients rock on ice for some applications; the mistake that the vendors are consistently making is to pretend that they're a good office worker solution. Clue phone: they aren't. They work great for single-function terminals.

    Course, we mighty geeks never consider that market because it's beneath our notice. However, consider how many terminals are in use today at, say, airport desks, hotel front desks, warehouses (shipping and receiving), many customer service jobs, telemarketing, yatta yatta yatta. It's a huge market and it's an important one, and if I ever get stuck running something like that I want to be using thin clients, cause they're easy to swap out and cheap cheap cheap to maintain.

    And -- this is the really important thing -- easy to upgrade. Remotely. Yeah.

In seeking the unattainable, simplicity only gets in the way. -- Epigrams in Programming, ACM SIGPLAN Sept. 1982