I like the ideas behind the debian one. It is as good of a choice of the IRIX one, maybe more of a logical choice as it does not have as many issues regarding licensing. I *really* like the IRIX window manager and assorted GUI tools. They paved the way for ease of use under UNIX, and don't want the work they have done to fade into a niche unless it has to..:-) Basically the SGI environment allows you to command line, and GUI or both depending upon your mood or need. Pretty painless either way for most common tasks.
In either case, the key needs to be flawless execution, and very straightforward interfaces, at least on the GUI side.
This package manager has a number of good features, and works well with both command line, and gui interfaces. Here are a couple of things that this tool has been able to do for years.
-interactive conflict resolution. The user gets to see what will get broken or overwritten, and has a chance to choose from different options.
-Removal of software even in the middle of installing it. I once ran out of disk on a few of these machines, switched modes, selected other software not needed, marked it for removal and continued without a hitch.
-It follows links, and nfs mounts without any hassle on the part of the user.
-Interactive display (gui mode) of all operations, disk space requirements, and overhead for pending operations.
If you have a chance to look at this tool, it is worth the look. With a few standards for packages, and dependancy listings, it would be a very good thing. Something the MS crowd is just now beginning to realize that they need. Every windows user I have shown this to asks if they can get it for their machine. I have used many of the various package managers for Unix and Linux, and this one is the easiest, and most robust out there. SGI are you reading? Maybe this should get onto the project list...
Man, You are not kidding! Wonder just how long these facts will stick around. Rainbow magazine was great. I learned a lof from those pages.
You know there were quite a number of good hacks for that machine. I used to use one that increased the already fast tape drive to 2400 baud. Good for the day. It was very tolerant as well as it used actual d-a a-d conversion for the data. Most of the others used digital, or cheap square waves that were both overly noisy, and very intolerant of speed changes. The COCO tape would deal with all of that, and had filenames!
I never used the keyboard code hack, but did see one coco overclocked. I can't remember which one 2 or 3, but basically you lost the video. If you were running os9, then you could get the additional speed, and just use a terminal via serial port.
I think hitachi made 6809E clones. The will clock to about 12Mhz these days. Still a pretty good choice if you are going to have to use assembler to build your project. There is probably no easier instruction set.
I have been unhappy with this trend toward bigger, more powerful machines. Todays PC's are HUGE. Not all of them need to be. Servers & such should be because you are going to want to load them up with everything, but for most people there is too much wasted space. There have been a few really nice small form factor machines released. (SGI Indy, Mac Performa, Ergo Brick, others...) Each of them had nicely intergrated features that provided most of what you need for computing. That part I like, but they lacked standards, and power (except for the Indy:)
My ideal machine would be book sized, maybe a bit bigger, and have ports for everything. It would be able to run on a few batteries, or perhaps a very standardized rechargable unit. It would run a very reliable OS, and include features for power management, so that I would not have to do the whole shutdown startup suspend thing. It would just be running, if I was not asking it to do much, then it would not consume much power. This is the thing that Crusoe really adds to the mix. If the OS understands what the users demands are, it can maintain standards of interactivity while not wasting resources.
This machine would be very networked. Wireless, IR, Ethernet, USB, serial (gotta have a console for the very lowest energy consumption!). Basically I don't want to give up any connectivity. All of these would be dynamic of course. If they are not used, they are turned off, or maybe done in software with only a configurable hardware interface that could adapt to the technology being presented.
Each machine would have display and keyboard capabilities that you can take or leave. If a bunch of them are networked together, no need to use their keyboards, just use the one closest to you. Power users could just get 3 or four of these and use one as traffic cop for the others while they are processing.
The OS should be able to communicate and utilize other resources of nearby machines with only a small direction from the users. Basically you should be able to stack a few of them up for hard things, and they should be able to get things done. Maybe you just might want to quickly distribute some data for a meeting, or maybe use your machine as a server for another machine already connected to a large display device. Maybe the data would only need to move when there are bandwidth problems, or you want the other party to own it.
Good media outputs and inputs. None of this marginal good enough crap. Audio should be 48Khz 75Db. Video should be S-video / Digital minimum. If we bite the bullet now on this, it will get cheap fast. These things would enable lots of applications that people would want. Combine your favorite commercials for jokes, maybe grab stills from your home movies for prints to send to friends, in fact why not just let them have the whole thing.
Maybe a few high school students want to assemble a great piece of music. They combine all of their machines to form a small mixing studio. Each of them provides input to one or two of them laying around, or borrowed to process and combine the data. Those same students discover a story of interest, and as a group possess the power to combine their vision into a compelling story that gets air time on a local radio, tv, or web broadcaster. Maybe they say the hell with it, and stream it themselves... It should be easy.
Companies would like them because they are disposable. Networking the office could really be done using I-R or lowpower RF. Just bring the machine near a desk, and maybe it would ask for a password, or permission to connect to the server who is wondering if the user needs anything. Data stays put. When a user goes home, sensitive stuff might just stay where it needs to be. Their machine goes with them however.
No matter how any of this goes, I just look at some of the cooler small form factor machines that were made, and they have way more appeal than the clunkers we have today. Lighter too.
Anyone remember this chip? Had probably the coolest and most powerful instruction set of any of the 8 bitters. I learned assembly on one of these, THEN had to go back and learn it again on a 6502 machine.... (bummer) There was a multiuser version of os/9 ported to the tandy color computer series of machines. I guess it is another example of the best tech not getting the limelight... Not relevant, but tandy actually clocked down their color computer series of machines in order to lessen their competetive edge compared to the tandy 1000 series. There was a register that you could poke to boot the clock back to the normal speeds.
Anyway I remember wanting to get a PET with the 6809 option, but decided not to because there was more 6502 software.
This would be a great idea. A lot of us know about prior art, but it will mean nothing if there is not a place for that information to be logged and searchable. Even if it were not verified, it still could cast lots of doubt on any patent litigation...
1. M$ Makes windows only run on Intel compatable PC's
2. They make office only run on windows.
3. To get any part of that functionality you basically have to buy into it all.
This means that anyone wanting to use office needs to have windows, and that means throwing out anything that is not a PC (or mac maybe). No matter how effective your computing solution is, there will be enormous pressure to get rid of it so that you can communicate with the other windows users. Basically this forces the technical computer users and administrators to buy in because of the intergration of office and windows.
How many companies run windows without office? I would be willing to wager that almost none of the larger ones do this.
What if office would run on other platforms? How many of those decisions to throw out platform X would have been made? Think about the advantages X servers and clients have for casual users and licensing. Clever utilization of compute resources and applications could really lower overall computing costs. Instead everybody gets a box, and it just happens to be a pretty good one because of the top-heavy software.
The intergration of applications and the OS is wrong. I am happy that Netscape was featured in the fact finding, but really the core of the problem is Office. It removes the question of choice from the equation because a machine can realistically only run one operating system at a time. (I know about VMWare and it rocks, but is not for everyone.)
I think that the DOJ should enforce this seperation to some degree. I don't want people telling us that certian applications need to be ported to certian other operating systems. That would result in some pretty stupid combinations, but there should be something in place to allow this to happen if the market wants it. Engineers, creative artists, industrial designers, and other technical professionals have spent the last few years paying to "upgrade" from whatever OS they were using (IRIX! among others...) to windows because they were unable to communicate with their peers in the office who were just running windows. Somehow this strikes me as really wrong. PC hardware is just now catching up to the level of performance that other hardware/OS combinations have had for a while now. They might cost more, but maybe that would not be true if they were able to move more boxes.
Should everybody be using a PC? Probably not. Because of windows intergration, creative hardware is discouraged. One OS does not fit all, and neither does one piece of hardware.
I think that is is there for control. They can vary the content by zone, and they also can have a say in who makes disk players, and what they can and can't do. Think about this: If there were no real encryption, what would prevent the average joe from producing DVD's? The mastering process would be just as it is now for CD's. Expensive at first, but less as demand grows. Also said average joe would be able to distribute their media anywhere...
Holistic system design. The OS was built during a time when good small powerful code was necessary for the operation of the machine. The display portion of the OS was pretty well dictated by the display capabilities of the computer, not the other way around. This is why the Amiga was revolutionary during its time. Everybody else was looking at building general solutions that pretty much limited what was possible with respect to multimedia solutions. (Can't resist saying that it also ran on a Motorola chip which at the time made Intel chips look pale in comaprison...) Also I guess you can't forget Jay Miner. He was the one who developed the initial chip designs. He also developed the early chip designs for the Atari 8 bit machines. On a low level all of his chip sets are similar. They were designed to give the programmer both a high level, and low level of control over what is being done with the display. It is these features that makes the Amiga distinctive even to this day. A developer could choose to make a general OS call for a bitmap, and overlay a few sprites to get the job done, or maybe tie the display to low-level interrupt driven CPU code to generate new and interesting displays, pretty much anything was possible even at the low clock speeds of the time. These types of graphics engines are very hackable. The smarter you are, the more the display will do for you. Contrast that to todays current crop of display systems. They are fairly rigid structures with general intent. Given todays general purpose computing enviroment, it is pretty hard to build something that is Amiga like using the current crop of off the shelf stuff.
Todays general purpose PC is just that. General purpose. The Amigas, SGI IRIX machines, and some others that I don't know about really excel at certain things because one entity was able to design the system, and tweak the OS to achieve pretty high goals.
I do not have an Amiga today (wish I still did!), but I still have a number of IRIX machines to work with, and they run circles around PC's in the areas where the system was designed as a whole to perform. (Just try to compose a letter, or edit a spread sheet though!) I do not believe that general purpose machines will ever reach this level of excellence just because of their design. Their focus is just too broad. Even if the OS was perfect, the machine would not be.
Making the Amiga OS open source will no doubt expose some great tools, and Linux will benefit, but none of it will really shine the way that machine, or others like it do because it will be for the most part be running on an X86 Intel driven PC.
We need a machine that runs linux that is not necessarily a PC. Another chicken and egg problem, but an interesting one. Lets say we take a couple of cases where a machine like this could complement the current PC. Most users don't need all these features in one box, so why not build them in smaller ones that they can use with the box that they have? Maybe use that box as a sort of traffic cop to control the other smaller ones that they might have. Linux is GREAT at this. It has evolved on the net, and thus is very net-centric.
1. Modeling engines. I think of the Sony engine that will be released in a short while. It is capable of realtime visualization tasks that would really boost productivity for engineers, animators, simulators and the like. They need a "real machine" with lots of disk, network, and CPU to hold the data and deal with their computing enviroment. Lets say that this "engine" connects with a network wire to the box that they already have. Maybe even utilizes a pass through cable to take advantage of their display. Make it big enough to be able to do the job it is supposed to do, (crunch on geometry) yet small and cheap enough that it is not a big deal to own one. Lets face it people will spend $3000 or more on just Graphics cards. Why not a modeling rendering Linux running engine with a chipset to die for?
2. Media stations. Having a large boxy computer to perform many different multimedia tasks is expensive and wasteful. Have a series of networked appliances that do various things. Compositing, editing, audio mixing, and processing. The interested party just gets them in the quantity needed, networks them, and a larger "server" type Linux machine could administer and direct them. They would be very portable, kind of like SCSI devices are today only you should not have to shut everything them down to swap them in and out. The users of this type of computing enviroment would also be capable of many parallel workflows that would require lots of expensive general purpose machines to duplicate.
My whole point here is that each item could be very flexible and powerful but well focused. Having the operating system and tools open sourced would allow the vendors of these sort of tools to compete on system design and function. They could use anyone's CPU, graphics, IO, whatever, or they could roll their own. Each unit could be very Amiga like in its design yielding the best performance for the least dollars.
Things have to change. Over the last couple of years it seems that all the workstation vendors have been able to produce similar capability. Now it is all about price. Reading the Sony story earlier makes me think about what is the next big thing? Probably Intel is in a sort of rut. Maybe the way they design CPU's needs to change. As long as there is very little difference between an inexpensive white box, and an expensive workstation, Intel will be in trouble. If we are at the end of Moores Law, then how come Sony is able to do what they are doing? I believe it is because they are able to approach the design of their systems holisticially without all of the compatability baggage required with todays PC. They are using the silicon, and software, and careful focus to provide very specific capabilities. Intel on the other hand is really just providing general purpose silicon. Maybe we are seeing the limits of what general purpose computing machines can do for a while. Not Moores Law.
I also have been thinking about this for a while. I myself have 4 foster kids. They are just getting started using Computers. Teaching them has given me food for lots of thought... What do they need to know, and how do I present it to them without all of the baggage that I had to go through to learn what I have? Their schools use of the computer is just too limited. I mean that not every kid really needs to know computers from the guts out, but they should know more than just how to use one. Here are a few things that I have been working through with my children... Some of this stuff will end up as short volunteer lectures at my local schools.
Computer culture is something that is priceless. With the right "manners" a child will grow to use a computer for anything they want without fear. If they learn to share early, maybe they will help others the same way that I have. (It is also interesting that the kids that I went to school with that shared knowedge, and code and games! I still talk to and learn from.) This sort of thing is needed when kids start learning. They are learning all social aspects of everything else, why not the use of computers? My own children seem pretty receptive. Simple ethical issues can lead to lengthly productive discussions. The same sort of discussions that arise from such questions like: "I found some chalk at the school, is it ok to keep it?" Computer culture is learned by peer example at the higher education levels, but not really at the lower ones, at least not in the schools in my area.
What is the internet? Their answers to this really blew me away. They said that it was just like TV only you get to pick what you see. People who knew the internet before the invention of the WWW are very different from those who joined in after the fact. All aspects of the internet should be introduced along with the appropriate culture. The sharing culture of the internet is slowly being eroded away. There are also many things surrounding fact vs opinion that are very confusing to young people. They really can't see the difference between a hack and a legit web site. Does a site contain factual information? Is a particular opinion popular? With who? What does that mean to me, and should I go along....
Low level computer understanding. My kids ask me things like: Why does the computer have to start up? (after having been turned off..) Why can't I just type what I want and have it go find it? Where do things go when the computer is off? Why are some peoples computers different? Why does the computer crash when it is supposed to do exactly what it is told? These sort of things are easy to ignore at first, but if you ask young kids any of these, their answers will surprise you.
Maybe this is all simple stuff, (to us it should be) but to the young it matters. More than our educational system is ready to prepare them for.
I think that the graphics are coming. There have got to be lots of legal issues regarding how they do graphics, and what they can or cannot release openly. The open source methodology may also slow them a bit. They just cannot rewrite what is there. That would not be good for those who have worked hard over the years. They are having to intergrate what they know, and sell it to those who are helping. If you have ever sat down at one of their workstations, it would become obvious that they know how to do X + OpenGL. The problem is that not everyone has been able to check out their technology. The only reason that NT happened so quickly is that they only had to Tweak OpenGL, and the underling HAL. Linux needs more work than this.
I hope they will be OK as well.
Interesting comments from users in CAD/Graphics
on
SGIs Linux Future
·
· Score: 1
Looks like this rambles a bit, but hey it has been a while. Most of it is relevant I think...
The 320 initially was regarded as a "non expandable" propritatary solution. Over the last few months people have begun to see value in the hardware. Entry level machines have the same graphics sub system as do the full boat systems. You can get the whole box for the price of the Wildcat 3D card in some cases. They are expandable where it counts. You can add RAM, and CPU. These will directly affect the overall performance of the graphics system as well as other things. They also have the ability to fine tune the system in ways that their competetors cannot because they have created their own HAL for NT to run under. It is pretty obvious that they made the machine just the way that they wanted to, but are offering customers a choice as to what they want to run. Right now it happens to be NT. The CAD/Graphics community simply has more interconnectivity on that platform and that is hurting all things Unix. (Even though it clearly is better, faster, more reliable.....)
I find it interesting that the CAD users will look hard at compute/floating point performance, and less at graphics. The things that they do are simply not as demanding of the graphics as are the other 3D apps. The other competetors stand up to the SGI machines pretty well in this space, although SGI is now gaining a price advantage. Maybe those earlier problems with manufacturing have gone away?
The graphics people happen to like the box quite well. Given all the different features that are intergrated, they get a lot of bang for the buck. There is no beating the unified memory scheme for rendering complex things. Their own applications run quite nicely on their machines, and give a compelling price/performance ratio. I think that they are competeting quite nicely in this area. Many of the graphics people that I have talked to miss high end batch rendering and scripting though. NT is still NT, and that means being in front of the machine to do real work.
It is obvious that they have built this machine to run the OS that the customer wants to run, whatever that is. Over the next few years interoperability will become the dominant force in computing. Those that have Unix have been buying PC's because of this. I have said this before, but maybe it bears repeating. The real battle in both of these markets is not really against NT, but the applications that run on it. Office in particular. Once a company starts using it, then it is only a matter of time before Unix in any form starts to die in R&D, Engineering, and Graphics.
At least SGI is adopting a stratgy that allows them to provide computing flexibility where the customer wants it. Right now that means co-existing with NT, and many users that I know are happy to see it. For those lucky enough to be running X based applications, that means that they can utilize the reliability of SGI Irix, and the luxury of having one PC on the desktop. All of which could be SGI. When Linux hits, maybe a bunch of them will find a reason to run it on hardware that was designed to take advantage of it.
If they are "Just another NT company", then why are they doing the things that they are?
They have been very active in developing their IRIX Operating system. 6.5 is very stable, and very feature complete, and very friendly. There is a GUI for most common tasks. If you are into any kind of media application, it is very hard to compete with IRIX given that the BASE OS provides many functional media tools, that are add-ons for everyone else.
From what I see, they are seriously trying to develop solutions in the Linux space as well as the NT one. Just because they have released an NT product does not make them a sellout. They did what they always do. Make killer hardware, and code to run on it. They did produce their own HAL for NT. This is clearly different from "all the other companies" who do a little tweak here and there to get performance. Their Visual Workstation was designed to run other things anyway. It boots with an ARC loader, and comes with simple install tools. Just like the IRIX based machine the O2. If you know both machines, you can't miss the likeness in design. (Internal design that is...) You need to take another look at one of these before deciding that they are just another NT company. (NT still does suck even on a great machine tho:-J )
These machines will run Linux, and they will have full Open-GL support under X. For graphics they will most likely set the standard for GLX under Linux.
Network computing has been pretty fair in their coverage of higher-end computing. The fact that they recognize the value of Open Source and that it involves a process, not just Linux is great stuff to see go mainstream. Don't forget to check out the excellent perl article later in the issue. Nice easy example of Perl flexibility. Program starts simple and gets complex yet really does not change much in size....
I also work for a competetor to PTC. We have a growing number of NT based sites. Those that bought our product on Unix are seeing less and less value. I gave Vmware a spin a while back and found it comfortable. I was able to do everything that I needed. I could run CAD via rlogin, E-mail & schedule & word & misc network tasks via Vmware, and administer any machine on our network, even the NT ones using VNC. Pretty powerful conbination. Now that SAMBA is getting some attention also, true interoperability solutions are a reality. This is the way of the future for me. Those that are aware of what network (true not M$ terminal server) computing can be like should have a choice. You guys could just provide a preconfigured Vmware file provided the user has an NT license, and they would be good to go. Maybe this could be a user group, or reseller thing if the lawyers have a problem with the license.... Maybe you are SDRC, maybe not... Need a beta tester?
Ok I have to say a few things about this. It't long so if you don't feel for MCAD, then blow past this one....
There are a few big unix based MCAD packages right now doing everything that they can to get their software to be windows NT friendly. I find it interesting that these companies will give up the strengths that the UNIX development brought them in exchange for compatability, and the "chance" to sell the same software cheaper.
In the last few years the CAD market has matured such that high priced UNIX workstations are no longer seen as a value. Linux could very well change this. (I for one really hope this happens!) Nobody minds buying a PC, but a Sun or SGI machine is a tough sell these days. Basically there has to be no other alternative, otherwise the sale will goto an NT machine.
There are a number of smaller, newer competetors in the MCAD arena, and they are entirly based on NT, and they are using "Sell lots of cheap software and make money later" business model. They are totally embracing the Microsoft development, marketing, and implementation model. The older established MCAD vendors need something new as a differentator; otherwise, they will just continue to lose market share until they own niche markets. Linux based systems could change this, but there are some things that need to be addressed before software ports are truly attractive to these companies, and their users.
1. Good solid 3d Open-GL. This is a requirement. Solid modelers consume graphic resources when they are not computing model geometry. A users performance is directly related to the 3D capabilities of the workstation in question.
2. Office compatiblity. Many engineers now have to schedule, e-mail, and perform basic office tasks using M$ software. This is one of the primary reasons UNIX machines are being displaced. Most, if not all of the companies do not like two machines on one engineers desk. Many engineers do not really want to know anything about the software they use. It gets in the way. Not that they are not capable users, they just want to be users only. When M$ wins on the office desktop, UNIX in the R&D department is on borrowed time.
3. Standard Linux distros. The biggest selling points, in this market (true or not:-P ) going for NT are: - Ability to standardize on inexpensive hardware. - Ease of Administration. I know that they do not yet realize what this one will cost them in the long run, but for now this is key. - Compatability. (also mentioned above.)
4. Good user tools. What does the average user do? Get good OpenSource tools that work together across Distros that accomplish common user tasks. Easy stuff should be easy. (Printing, Plotting, moving files, finding files that sort of thing.) M$ currently makes easy stuff easy for the average Joe user. Problem is Hard things are Impossible.
These MCAD comapnies that come from UNIX roots have a fair number of reasons to port also.
1. Marketing. If any one of them actually does it, then they will get a lot of attention in a market that generally is hard to get to listen.
2. Absolute control of the User Interface. The Windows GUI was meant for the average computer user. All the Solid Modelers I know really know their package. A few clicks here and there matter. If a package must conform to the windows "Look and Feel" then their ability to compete on this basis is sharply limited.
3. Leverage of existing tools and technology. X windows is great. Someday customers better realize that they could purchase a few very capable servers, and let their power users use them, from any machine in the building. NT does not allow for Network computing, and this market needs it.
4. Ability to push compute envelope beyond current Intel Technology. Right now if you top out your new shiny NT workstation with a large assembly, you are stuck. If you were using X you could just rlogin into a more powerful machine for that task, or maybe you could go with an Alpha machine, or maybe PowerPC. Either way, top performance would be available to those that are willing to pay for it. NT currently is a dead end here right now. BTW Cad performs poorly on Multi Processor machines due to its serial nature.
5. The other NT based competetors will have to do one hell of a re-write in order to compete, or compete on other strengths. Their reliance on pure Microsoft tools will insure this. Using Linux correctly will raise the bar in many areas that are currently taken for granted by much of this market.
I guess I need to clarify what I said before. I don't want my GUI to look like theirs, and I certianly am not advocating use of their standards. The average user will want to be able to do most of the things that the MS GUI is capable of, and they will want to do it visually. That's all.
Pay by the hour computer gaming probably is not as bad as DIVX, but it is still pretty bad. The user is still tied to their pay-per-service scheme. I believe that DIVX is in trouble because it is a hard sell, and they need some other gimmick to keep it afloat.
Does anyone know how this will work? Do they have a license manager that you type codes into, or does the thing need a tcp-ip connection to mamma the whole time you play? How many people will tie up their phone for that length of time? What happens to the user who starts their game, then needs to do something else, and doesn't get back to it for a couple of days? Oops! Better pay again? NO WAY! There are not that many people that are STUPID enough to buy into this to keep it afloat. Then again, they can always gouge their existing DIVX customers again, and again....
Probably this would deflate a few egos also. Imagine writing a hotly worded column, then getting... NOTHING! Maybe while they are here looking for that sweet flame, they can follow a link or two and learn something.
Slashdot Mirror
I like the ideas behind the debian one. It is as good of a choice of the IRIX one, maybe more of a logical choice as it does not have as many issues regarding licensing. I *really* like the IRIX window manager and assorted GUI tools. They paved the way for ease of use under UNIX, and don't want the work they have done to fade into a niche unless it has to.. :-) Basically the SGI environment allows you to command line, and GUI or both depending upon your mood or need. Pretty painless either way for most common tasks.
:-)
In either case, the key needs to be flawless execution, and very straightforward interfaces, at least on the GUI side.
You should pitch what you like, I'm gonna
This package manager has a number of good features, and works well with both command line, and gui interfaces. Here are a couple of things that this tool has been able to do for years.
-interactive conflict resolution. The user gets to see what will get broken or overwritten, and has a chance to choose from different options.
-Removal of software even in the middle of installing it. I once ran out of disk on a few of these machines, switched modes, selected other software not needed, marked it for removal and continued without a hitch.
-It follows links, and nfs mounts without any hassle on the part of the user.
-Interactive display (gui mode) of all operations, disk space requirements, and overhead for pending operations.
If you have a chance to look at this tool, it is worth the look. With a few standards for packages, and dependancy listings, it would be a very good thing. Something the MS crowd is just now beginning to realize that they need. Every windows user I have shown this to asks if they can get it for their machine. I have used many of the various package managers for Unix and Linux, and this one is the easiest, and most robust out there. SGI are you reading? Maybe this should get onto the project list...
Man, You are not kidding! Wonder just how long these facts will stick around. Rainbow magazine was great. I learned a lof from those pages.
You know there were quite a number of good hacks for that machine. I used to use one that increased the already fast tape drive to 2400 baud. Good for the day. It was very tolerant as well as it used actual d-a a-d conversion for the data. Most of the others used digital, or cheap square waves that were both overly noisy, and very intolerant of speed changes. The COCO tape would deal with all of that, and had filenames!
I never used the keyboard code hack, but did see one coco overclocked. I can't remember which one 2 or 3, but basically you lost the video. If you were running os9, then you could get the additional speed, and just use a terminal via serial port.
I think hitachi made 6809E clones. The will clock to about 12Mhz these days. Still a pretty good choice if you are going to have to use assembler to build your project. There is probably no easier instruction set.
Later,
I have been unhappy with this trend toward bigger, more powerful machines. Todays PC's are HUGE. Not all of them need to be. Servers & such should be because you are going to want to load them up with everything, but for most people there is too much wasted space. There have been a few really nice small form factor machines released. (SGI Indy, Mac Performa, Ergo Brick, others...) Each of them had nicely intergrated features that provided most of what you need for computing. That part I like, but they lacked standards, and power (except for the Indy :)
My ideal machine would be book sized, maybe a bit bigger, and have ports for everything. It would be able to run on a few batteries, or perhaps a very standardized rechargable unit. It would run a very reliable OS, and include features for power management, so that I would not have to do the whole shutdown startup suspend thing. It would just be running, if I was not asking it to do much, then it would not consume much power. This is the thing that Crusoe really adds to the mix. If the OS understands what the users demands are, it can maintain standards of interactivity while not wasting resources.
This machine would be very networked. Wireless, IR, Ethernet, USB, serial (gotta have a console for the very lowest energy consumption!). Basically I don't want to give up any connectivity. All of these would be dynamic of course. If they are not used, they are turned off, or maybe done in software with only a configurable hardware interface that could adapt to the technology being presented.
Each machine would have display and keyboard capabilities that you can take or leave. If a bunch of them are networked together, no need to use their keyboards, just use the one closest to you. Power users could just get 3 or four of these and use one as traffic cop for the others while they are processing.
The OS should be able to communicate and utilize other resources of nearby machines with only a small direction from the users. Basically you should be able to stack a few of them up for hard things, and they should be able to get things done. Maybe you just might want to quickly distribute some data for a meeting, or maybe use your machine as a server for another machine already connected to a large display device. Maybe the data would only need to move when there are bandwidth problems, or you want the other party to own it.
Good media outputs and inputs. None of this marginal good enough crap. Audio should be 48Khz 75Db. Video should be S-video / Digital minimum. If we bite the bullet now on this, it will get cheap fast. These things would enable lots of applications that people would want. Combine your favorite commercials for jokes, maybe grab stills from your home movies for prints to send to friends, in fact why not just let them have the whole thing.
Maybe a few high school students want to assemble a great piece of music. They combine all of their machines to form a small mixing studio. Each of them provides input to one or two of them laying around, or borrowed to process and combine the data. Those same students discover a story of interest, and as a group possess the power to combine their vision into a compelling story that gets air time on a local radio, tv, or web broadcaster. Maybe they say the hell with it, and stream it themselves... It should be easy.
Companies would like them because they are disposable. Networking the office could really be done using I-R or lowpower RF. Just bring the machine near a desk, and maybe it would ask for a password, or permission to connect to the server who is wondering if the user needs anything. Data stays put. When a user goes home, sensitive stuff might just stay where it needs to be. Their machine goes with them however.
No matter how any of this goes, I just look at some of the cooler small form factor machines that were made, and they have way more appeal than the clunkers we have today. Lighter too.
Anyone remember this chip? Had probably the coolest and most powerful instruction set of any of the 8 bitters. I learned assembly on one of these, THEN had to go back and learn it again on a 6502 machine.... (bummer) There was a multiuser version of os/9 ported to the tandy color computer series of machines. I guess it is another example of the best tech not getting the limelight... Not relevant, but tandy actually clocked down their color computer series of machines in order to lessen their competetive edge compared to the tandy 1000 series. There was a register that you could poke to boot the clock back to the normal speeds.
Anyway I remember wanting to get a PET with the 6809 option, but decided not to because there was more 6502 software.
This would be a great idea. A lot of us know about prior art, but it will mean nothing if there is not a place for that information to be logged and searchable. Even if it were not verified, it still could cast lots of doubt on any patent litigation...
Good for Sony. As things slowly become global, people realize that other cultures have a lot to offer. What model DVD did you get?
The thing I want fixed the most is this:
1. M$ Makes windows only run on Intel compatable PC's
2. They make office only run on windows.
3. To get any part of that functionality you basically have to buy into it all.
This means that anyone wanting to use office needs to have windows, and that means throwing out anything that is not a PC (or mac maybe). No matter how effective your computing solution is, there will be enormous pressure to get rid of it so that you can communicate with the other windows users. Basically this forces the technical computer users and administrators to buy in because of the intergration of office and windows.
How many companies run windows without office? I would be willing to wager that almost none of the larger ones do this.
What if office would run on other platforms? How many of those decisions to throw out platform X would have been made? Think about the advantages X servers and clients have for casual users and licensing. Clever utilization of compute resources and applications could really lower overall computing costs. Instead everybody gets a box, and it just happens to be a pretty good one because of the top-heavy software.
The intergration of applications and the OS is wrong. I am happy that Netscape was featured in the fact finding, but really the core of the problem is Office. It removes the question of choice from the equation because a machine can realistically only run one operating system at a time. (I know about VMWare and it rocks, but is not for everyone.)
I think that the DOJ should enforce this seperation to some degree. I don't want people telling us that certian applications need to be ported to certian other operating systems. That would result in some pretty stupid combinations, but there should be something in place to allow this to happen if the market wants it. Engineers, creative artists, industrial designers, and other technical professionals have spent the last few years paying to "upgrade" from whatever OS they were using (IRIX! among others...) to windows because they were unable to communicate with their peers in the office who were just running windows. Somehow this strikes me as really wrong. PC hardware is just now catching up to the level of performance that other hardware/OS combinations have had for a while now. They might cost more, but maybe that would not be true if they were able to move more boxes.
Should everybody be using a PC? Probably not. Because of windows intergration, creative hardware is discouraged. One OS does not fit all, and neither does one piece of hardware.
Looks different than all the other bikes. Has so many fun toys attached that you never really get to ride it...
I think that is is there for control. They can vary the content by zone, and they also can have a say in who makes disk players, and what they can and can't do. Think about this: If there were no real encryption, what would prevent the average joe from producing DVD's? The mastering process would be just as it is now for CD's. Expensive at first, but less as demand grows. Also said average joe would be able to distribute their media anywhere...
Holistic system design. The OS was built during a time when good small powerful code was necessary for the operation of the machine. The display portion of the OS was pretty well dictated by the display capabilities of the computer, not the other way around. This is why the Amiga was revolutionary during its time. Everybody else was looking at building general solutions that pretty much limited what was possible with respect to multimedia solutions. (Can't resist saying that it also ran on a Motorola chip which at the time made Intel chips look pale in comaprison...) Also I guess you can't forget Jay Miner. He was the one who developed the initial chip designs. He also developed the early chip designs for the Atari 8 bit machines. On a low level all of his chip sets are similar. They were designed to give the programmer both a high level, and low level of control over what is being done with the display. It is these features that makes the Amiga distinctive even to this day. A developer could choose to make a general OS call for a bitmap, and overlay a few sprites to get the job done, or maybe tie the display to low-level interrupt driven CPU code to generate new and interesting displays, pretty much anything was possible even at the low clock speeds of the time. These types of graphics engines are very hackable. The smarter you are, the more the display will do for you. Contrast that to todays current crop of display systems. They are fairly rigid structures with general intent. Given todays general purpose computing enviroment, it is pretty hard to build something that is Amiga like using the current crop of off the shelf stuff.
Todays general purpose PC is just that. General purpose. The Amigas, SGI IRIX machines, and some others that I don't know about really excel at certain things because one entity was able to design the system, and tweak the OS to achieve pretty high goals.
I do not have an Amiga today (wish I still did!), but I still have a number of IRIX machines to work with, and they run circles around PC's in the areas where the system was designed as a whole to perform. (Just try to compose a letter, or edit a spread sheet though!) I do not believe that general purpose machines will ever reach this level of excellence just because of their design. Their focus is just too broad. Even if the OS was perfect, the machine would not be.
Making the Amiga OS open source will no doubt expose some great tools, and Linux will benefit, but none of it will really shine the way that machine, or others like it do because it will be for the most part be running on an X86 Intel driven PC.
We need a machine that runs linux that is not necessarily a PC. Another chicken and egg problem, but an interesting one. Lets say we take a couple of cases where a machine like this could complement the current PC. Most users don't need all these features in one box, so why not build them in smaller ones that they can use with the box that they have? Maybe use that box as a sort of traffic cop to control the other smaller ones that they might have. Linux is GREAT at this. It has evolved on the net, and thus is very net-centric.
1. Modeling engines. I think of the Sony engine that will be released in a short while. It is capable of realtime visualization tasks that would really boost productivity for engineers, animators, simulators and the like. They need a "real machine" with lots of disk, network, and CPU to hold the data and deal with their computing enviroment. Lets say that this "engine" connects with a network wire to the box that they already have. Maybe even utilizes a pass through cable to take advantage of their display. Make it big enough to be able to do the job it is supposed to do, (crunch on geometry) yet small and cheap enough that it is not a big deal to own one. Lets face it people will spend $3000 or more on just Graphics cards. Why not a modeling rendering Linux running engine with a chipset to die for?
2. Media stations. Having a large boxy computer to perform many different multimedia tasks is expensive and wasteful. Have a series of networked appliances that do various things. Compositing, editing, audio mixing, and processing. The interested party just gets them in the quantity needed, networks them, and a larger "server" type Linux machine could administer and direct them. They would be very portable, kind of like SCSI devices are today only you should not have to shut everything them down to swap them in and out. The users of this type of computing enviroment would also be capable of many parallel workflows that would require lots of expensive general purpose machines to duplicate.
My whole point here is that each item could be very flexible and powerful but well focused. Having the operating system and tools open sourced would allow the vendors of these sort of tools to compete on system design and function. They could use anyone's CPU, graphics, IO, whatever, or they could roll their own. Each unit could be very Amiga like in its design yielding the best performance for the least dollars.
Things have to change. Over the last couple of years it seems that all the workstation vendors have been able to produce similar capability. Now it is all about price. Reading the Sony story earlier makes me think about what is the next big thing? Probably Intel is in a sort of rut. Maybe the way they design CPU's needs to change. As long as there is very little difference between an inexpensive white box, and an expensive workstation, Intel will be in trouble.
If we are at the end of Moores Law, then how come Sony is able to do what they are doing? I believe it is because they are able to approach the design of their systems holisticially without all of the compatability baggage required with todays PC. They are using the silicon, and software, and careful focus to provide very specific capabilities. Intel on the other hand is really just providing general purpose silicon. Maybe we are seeing the limits of what general purpose computing machines can do for a while. Not Moores Law.
Wonder what he would say?
I also have been thinking about this for a while. I myself have 4 foster kids. They are just getting started using Computers. Teaching them has given me food for lots of thought... What do they need to know, and how do I present it to them without all of the baggage that I had to go through to learn what I have? Their schools use of the computer is just too limited. I mean that not every kid really needs to know computers from the guts out, but they should know more than just how to use one. Here are a few things that I have been working through with my children... Some of this stuff will end up as short volunteer lectures at my local schools.
Computer culture is something that is priceless. With the right "manners" a child will grow to use a computer for anything they want without fear. If they learn to share early, maybe they will help others the same way that I have. (It is also interesting that the kids that I went to school with that shared knowedge, and code and games! I still talk to and learn from.) This sort of thing is needed when kids start learning. They are learning all social aspects of everything else, why not the use of computers? My own children seem pretty receptive. Simple ethical issues can lead to lengthly productive discussions. The same sort of discussions that arise from such questions like: "I found some chalk at the school, is it ok to keep it?" Computer culture is learned by peer example at the higher education levels, but not really at the lower ones, at least not in the schools in my area.
What is the internet? Their answers to this really blew me away. They said that it was just like TV only you get to pick what you see. People who knew the internet before the invention of the WWW are very different from those who joined in after the fact. All aspects of the internet should be introduced along with the appropriate culture. The sharing culture of the internet is slowly being eroded away. There are also many things surrounding fact vs opinion that are very confusing to young people. They really can't see the difference between a hack and a legit web site. Does a site contain factual information? Is a particular opinion popular? With who? What does that mean to me, and should I go along....
Low level computer understanding. My kids ask me things like: Why does the computer have to start up? (after having been turned off..) Why can't I just type what I want and have it go find it? Where do things go when the computer is off? Why are some peoples computers different? Why does the computer crash when it is supposed to do exactly what it is told? These sort of things are easy to ignore at first, but if you ask young kids any of these, their answers will surprise you.
Maybe this is all simple stuff, (to us it should be) but to the young it matters. More than our educational system is ready to prepare them for.
I think that the graphics are coming. There have got to be lots of legal issues regarding how they do graphics, and what they can or cannot release openly. The open source methodology may also slow them a bit. They just cannot rewrite what is there. That would not be good for those who have worked hard over the years. They are having to intergrate what they know, and sell it to those who are helping. If you have ever sat down at one of their workstations, it would become obvious that they know how to do X + OpenGL. The problem is that not everyone has been able to check out their technology. The only reason that NT happened so quickly is that they only had to Tweak OpenGL, and the underling HAL. Linux needs more work than this.
I hope they will be OK as well.
Looks like this rambles a bit, but hey it has been a while. Most of it is relevant I think...
The 320 initially was regarded as a "non expandable" propritatary solution. Over the last few months people have begun to see value in the hardware. Entry level machines have the same graphics sub system as do the full boat systems. You can get the whole box for the price of the Wildcat 3D card in some cases. They are expandable where it counts. You can add RAM, and CPU. These will directly affect the overall performance of the graphics system as well as other things. They also have the ability to fine tune the system in ways that their competetors cannot because they have created their own HAL for NT to run under. It is pretty obvious that they made the machine just the way that they wanted to, but are offering customers a choice as to what they want to run. Right now it happens to be NT. The CAD/Graphics community simply has more interconnectivity on that platform and that is hurting all things Unix. (Even though it clearly is better, faster, more reliable.....)
I find it interesting that the CAD users will look hard at compute/floating point performance, and less at graphics. The things that they do are simply not as demanding of the graphics as are the other 3D apps. The other competetors stand up to the SGI machines pretty well in this space, although SGI is now gaining a price advantage. Maybe those earlier problems with manufacturing have gone away?
The graphics people happen to like the box quite well. Given all the different features that are intergrated, they get a lot of bang for the buck. There is no beating the unified memory scheme for rendering complex things. Their own applications run quite nicely on their machines, and give a compelling price/performance ratio. I think that they are competeting quite nicely in this area. Many of the graphics people that I have talked to miss high end batch rendering and scripting though. NT is still NT, and that means being in front of the machine to do real work.
It is obvious that they have built this machine to run the OS that the customer wants to run, whatever that is. Over the next few years interoperability will become the dominant force in computing. Those that have Unix have been buying PC's because of this. I have said this before, but maybe it bears repeating. The real battle in both of these markets is not really against NT, but the applications that run on it. Office in particular. Once a company starts using it, then it is only a matter of time before Unix in any form starts to die in R&D, Engineering, and Graphics.
At least SGI is adopting a stratgy that allows them to provide computing flexibility where the customer wants it. Right now that means co-existing with NT, and many users that I know are happy to see it. For those lucky enough to be running X based applications, that means that they can utilize the reliability of SGI Irix, and the luxury of having one PC on the desktop. All of which could be SGI. When Linux hits, maybe a bunch of them will find a reason to run it on hardware that was designed to take advantage of it.
If they are "Just another NT company", then why are they doing the things that they are?
:-J )
They have been very active in developing their IRIX Operating system. 6.5 is very stable, and very feature complete, and very friendly. There is a GUI for most common tasks. If you are into any kind of media application, it is very hard to compete with IRIX given that the BASE OS provides many functional media tools, that are add-ons for everyone else.
From what I see, they are seriously trying to develop solutions in the Linux space as well as the NT one. Just because they have released an NT product does not make them a sellout. They did what they always do. Make killer hardware, and code to run on it. They did produce their own HAL for NT. This is clearly different from "all the other companies" who do a little tweak here and there to get performance. Their Visual Workstation was designed to run other things anyway. It boots with an ARC loader, and comes with simple install tools. Just like the IRIX based machine the O2. If you know both machines, you can't miss the likeness in design. (Internal design that is...) You need to take another look at one of these before deciding that they are just another NT company. (NT still does suck even on a great machine tho
These machines will run Linux, and they will have full Open-GL support under X. For graphics they will most likely set the standard for GLX under Linux.
Network computing has been pretty fair in their coverage of higher-end computing. The fact that they recognize the value of Open Source and that it involves a process, not just Linux is great stuff to see go mainstream. Don't forget to check out the excellent perl article later in the issue. Nice easy example of Perl flexibility. Program starts simple and gets complex yet really does not change much in size....
I also work for a competetor to PTC. We have a growing number of NT based sites. Those that bought our product on Unix are seeing less and less value. I gave Vmware a spin a while back and found it comfortable. I was able to do everything that I needed. I could run CAD via rlogin, E-mail & schedule & word & misc network tasks via Vmware, and administer any machine on our network, even the NT ones using VNC. Pretty powerful conbination. Now that SAMBA is getting some attention also, true interoperability solutions are a reality. This is the way of the future for me. Those that are aware of what network (true not M$ terminal server) computing can be like should have a choice.
You guys could just provide a preconfigured Vmware file provided the user has an NT license, and they would be good to go. Maybe this could be a user group, or reseller thing if the lawyers have a problem with the license.... Maybe you are SDRC, maybe not... Need a beta tester?
Cheers,
Ok I have to say a few things about this. It't long so if you don't feel for MCAD, then blow past this one....
:-P ) going for NT are:
There are a few big unix based MCAD packages right now doing everything that they can to get their software to be windows NT friendly. I find it interesting that these companies will give up the strengths that the UNIX development brought them in exchange for compatability, and the "chance" to sell the same software cheaper.
In the last few years the CAD market has matured such that high priced UNIX workstations are no longer seen as a value. Linux could very well change this. (I for one really hope this happens!) Nobody minds buying a PC, but a Sun or SGI machine is a tough sell these days. Basically there has to be no other alternative, otherwise the sale will goto an NT machine.
There are a number of smaller, newer competetors in the MCAD arena, and they are entirly based on NT, and they are using "Sell lots of cheap software and make money later" business model. They are totally embracing the Microsoft development, marketing, and implementation model. The older established MCAD vendors need something new as a differentator; otherwise, they will just continue to lose market share until they own niche markets. Linux based systems could change this, but there are some things that need to be addressed before software ports are truly attractive to these companies, and their users.
1. Good solid 3d Open-GL. This is a requirement. Solid modelers consume graphic resources when they are not computing model geometry. A users performance is directly related to the 3D capabilities of the workstation in question.
2. Office compatiblity. Many engineers now have to schedule, e-mail, and perform basic office tasks using M$ software. This is one of the primary reasons UNIX machines are being displaced. Most, if not all of the companies do not like two machines on one engineers desk. Many engineers do not really want to know anything about the software they use. It gets in the way. Not that they are not capable users, they just want to be users only. When M$ wins on the office desktop, UNIX in the R&D department is on borrowed time.
3. Standard Linux distros. The biggest selling points, in this market (true or not
- Ability to standardize on inexpensive hardware.
- Ease of Administration. I know that they do not yet realize what this one will cost them in the long run, but for now this is key.
- Compatability. (also mentioned above.)
4. Good user tools. What does the average user do? Get good OpenSource tools that work together across Distros that accomplish common user tasks. Easy stuff should be easy. (Printing, Plotting, moving files, finding files that sort of thing.) M$ currently makes easy stuff easy for the average Joe user. Problem is Hard things are Impossible.
These MCAD comapnies that come from UNIX roots have a fair number of reasons to port also.
1. Marketing. If any one of them actually does it, then they will get a lot of attention in a market that generally is hard to get to listen.
2. Absolute control of the User Interface. The Windows GUI was meant for the average computer user. All the Solid Modelers I know really know their package. A few clicks here and there matter. If a package must conform to the windows "Look and Feel" then their ability to compete on this basis is sharply limited.
3. Leverage of existing tools and technology. X windows is great. Someday customers better realize that they could purchase a few very capable servers, and let their power users use them, from any machine in the building. NT does not allow for Network computing, and this market needs it.
4. Ability to push compute envelope beyond current Intel Technology. Right now if you top out your new shiny NT workstation with a large assembly, you are stuck. If you were using X you could just rlogin into a more powerful machine for that task, or maybe you could go with an Alpha machine, or maybe PowerPC. Either way, top performance would be available to those that are willing to pay for it. NT currently is a dead end here right now. BTW Cad performs poorly on Multi Processor machines due to its serial nature.
5. The other NT based competetors will have to do one hell of a re-write in order to compete, or compete on other strengths. Their reliance on pure Microsoft tools will insure this. Using Linux correctly will raise the bar in many areas that are currently taken for granted by much of this market.
--Happily running MCAD and Design Apps on an SGI.
I guess I need to clarify what I said before. I don't want my GUI to look like theirs, and I certianly am not advocating use of their standards. The average user will want to be able to do most of the things that the MS GUI is capable of, and they will want to do it visually. That's all.
Pay by the hour computer gaming probably is not as bad as DIVX, but it is still pretty bad. The user is still tied to their pay-per-service scheme. I believe that DIVX is in trouble because it is a hard sell, and they need some other gimmick to keep it afloat.
Does anyone know how this will work? Do they have a license manager that you type codes into, or does the thing need a tcp-ip connection to mamma the whole time you play? How many people will tie up their phone for that length of time? What happens to the user who starts their game, then needs to do something else, and doesn't get back to it for a couple of days? Oops! Better pay again? NO WAY! There are not that many people that are STUPID enough to buy into this to keep it afloat. Then again, they can always gouge their existing DIVX customers again, and again....
Probably this would deflate a few egos also. Imagine writing a hotly worded column, then getting... NOTHING! Maybe while they are here looking for that sweet flame, they can follow a link or two and learn something.
Just a test.
You would think that they would get basic graphics right, but they are busy trying to get their feet back underneath of them.
I'm going to miss the cube too. Even though it looks sort of 80'ish, I like it.