If we define a Trojan Horse as a program which appears to do one thing -- perhaps productive, perhaps not -- and actually does something else, such as breach the heck out of your system security, could it not be argued that Windows, Internet Explorer and Outlook could be catagorized as Trojan Horses?
I can't think of any OS in recent history where simply getting the system to boot should qualify as any kind of victory. Getting an OS to boot may be a milepost, but only the very first milepost in a very long trip.
I also agree the recent "tech" stuff is pathetic. Remember when Byte magazine used to have schematic diagrams? Remember Byte magazine?? So much now is so absurdly platform specific, and so many reviewers claiming to be experts on things they are they shouldn't, and condeming "competing" or differing products they show no real experience in or understanding of.
I'm also not impressed that this reviewer would even mention an attempt to do an install on unreliable hardware, or a dial-up FTP install. I've tried a few FTP installs through my cable internet access. OpenBSD is the only one that really went well! Now, I think I will credit this more to the site that I pulled it down from than the product itself, but I have to say OpenBSD's FTP install can work, and can work very well, at least if you don't bring an expectation that it is "just like Linux" (or DOS, or Windows or...) to the process.
While I would agree there is a lot of stuff written on Linux, I'd question the value of most of it. The real issue is how much GOOD stuff is written on the subjects, and how much cr*p you have to dig through to find it.
I have been playing with Linux on-and-off since probably 1994 or 1995. I did my first OpenBSD and FreeBSD installs only a month or so ago. I felt more at home with OpenBSD and FreeBSD within minutes than I did with any version of Linux..ever.
Why? Simply because *BSD more closely matched the considerable Unix documentation that is out there. Linux is a changing target. By the time a release is properly documented, it is being replaced by a new release.
I also found the documentation available for *BSD to be more specific to the actual package I had than the Linux documentation. The Linux docs are considerable and some are excelent quality -- but the first challenge is trying to figure out what actually got installed and what is applicable and what isn't to the particular distribution that is loaded on the machine in front of me.
To me, it feels like Linux is a lot of pieces bolted together. Each piece is well documented, but how the pieces interact and work together is a challenge to figure out. OpenBSD seems to be a more coherent system.
I wanted to turn OpenBSD into a NAT system for my cable internet access. Within a relatively few hours of fiddling and fighting with wierd hardware, it was running and operating perfectly. I spent far more time figuring out the logic of the partitioning program (and cleaning up the mess after I discovered I was wrong...time after time! 8) and fighting with a SCSI controller supported by nothing in the world other than DOS and Windows 95 than I did getting NAT going.
Don't get me wrong... I respect Linux greatly, and I do not question that it is a much better desktop OS than *BSD is for a number of reasons. (One of which is the hype factor -- a perceived popular OS will have more software support, and this is important.) However, *BSD has a very significant place where stable, secure and reliable operation is more important than flash. OpenBSD has impressed the heck out of me very quickly. And, it sounds like the only "Internet" OS that really has security as something other than an afterthought.
Esp. in an area like the Internet and Internet patents, where things are so new and so unproven. Most likely, some of these cases are going to have to go through the courts (ALL the way through the courts!) to get set some guidelines set. Until then, the victor would probably be the people who have the deeper pockets to wave around the high-priced legal team and scare their competitors into submission.
The only potentially "cheap" thing I can think of doing is to find a legal team who would take the case pro-bono. It would (probably) have to be either an "easy" victory and/or a high visability case (i.e., very hard victory, but landmark-decision type case). It would also require a law practice which is not starving for money but wanting to increase their image in the area of Internet patent law. There are probably a lot of law offices which might be willing to take the right case like this, but again, you are stuck trying to find a law office of QUALITY.
Yes, MS-DOS had an undelete program, but it made no effort to avoid overwriting the old files. MS-DOS (and apparently, Windows 9x) uses the lowest number available block (closest to track 0, outer edge of the disk) first when writing files. So, if you delete a file (physically) at the beginning of the disk, it is overwritten VERY quickly. If you delete a file physically near the end of the disk and there is lots of free space earlier up on the disk, you may be able to recover that file for quite some time. Your ability to recover a file under DOS was based on luck and what you have done since then. There is no deliberate plan to avoid overwriting deleted files.
Hmm....now that I think about it, DOS v6 *did* include an optional device driver which would track deleted files in a comparatively primative way, but it was there (and I assume it worked, at least if you remembered to load it BEFORE you made a mistaken deletion!). I'm curious if it worked only from the command line (i.e., it watched for use of the DELETE command, like the Win9x recycle bin) or if it worked at the API level (so a program (like MS-Office 4.x) that deleted a file would have it tracked). No idea -- I never played with the deletion tracking feature of DOS. (In case you were wondering, the Netware deletion tracking tracked at the API level..doesn't matter if you DEL *.*'ed or the program deleted the file, Netware tracked it. Netware even tracks the file if you delete its parent directory).
Virtually any file system on any OS can have files recovered from deletion if new files haven't actually overwriten the "ashes" of the old files. This has not a thing to do with the MS-DOS FAT file system. DOS's file system was simple enough, however, that it was fairly easy to write undeletion programs. The first one I saw was a customer of mine wrote one on MS-DOS v1 (a FANTASICLY clever kid, but a bit clumsy when it came to deleting files). Under CP/M, undeleting files was SO easy, I always just used a disk editor (I think undelete programs existed, they were just barely needed).
Netware is rare in that it actually TRACKS those old "ashes" and keeps them around and intact, with all information (including the userID of the deleter!) so they can be reliably and successfully recovered as long as possible (or until told to do otherwise).
I do seem to have wandered well off-topic.. I wasn't writing in praise of Netware (I am very willing to do this, as people have undoubtably noticed!), as I have the knawing suspicion in this forum, most people wouldn't care. My point was to bring up Yet Another potentially anti-competitive action of Microsoft. The idea of writing a temporary file before deleting the original would give you an improved likelyhood of being able to undelete a newly replaced file even under DOS if you got to it soon enough. However, this Temporary/Delete/Rename process did WONDERFUL things when combined with the Netware deletion tracking system.
I don't do NT. I have yet to see an NT network I'd have been proud to put my name on. (I told this to an NT installer once. He listened to my points, agreed with them. Then said "Yeah, but you make much more MONEY with NT!")
I haven't become comfortable enough with Linux and FreeBSD's problem recovery process to put it at my clients as a file/print sharing system yet. But hopefully, soon. 8)
I became aware of another potentially anti-competitive thing MS did some years back, which, as far as I am aware no one else has mentioned in the press before. If it wasn't anti-competitive, it was at least stupid, and MS users are the ultimate victims.
A standard proceedure for saving files used in "single user" apps, like word processors and spreadsheets, was developed many, many years ago, and used almost universally. The proceedure was you DON'T just overwrite the old file, you do the following: * write the new file to a temporary file name * Delete the old file -- or rename it to.BAK (first deleting any old.BAK file). * Rename the temporary file to the original file name.
This made sure that at no time was the only copy of the file on the disk deleted or dammaged. If the program crashed durring the save cycle, the original copy still existed. Great idea. Saved a LOT of people, and this is why it was used almost universally, and at Microsoft through Office v4.3.
Starting with Office 95, however, they changed the saving proceedure: * Reset file pointer to beginning of file * Overwrite existing file with new data
Obviously, this leaves the system at a fairly delicate point, should the app or OS crash after the file has started to be overwritten.
So far, things look just stupid. However, if we remember one of MSs chief rivals, Novell, and a feature of Netware 3.x and above, things look a little less innocent. Netware has a very nifty "undelete" feature, they refer to as "Salvage"ing files. Unlike most other OSs where this feature is provided by third party add-ons, with Netware, this is a CORE part of the OS. A deleted file is marked as deleted and put into the free disk space pool, but it isn't overwritten until the entire disk has been used, at which point the oldest deleted files are overwritten (Actually, Macintosh does much the same thing). Novell provides you a nifty utility (SALVAGE in NW3.x, FILER in NW4+) which can bring these files back up to the point they are physically overwritten (or Netware is told to "PURGE" the deleted files).
This is a really neat feature, or at least, WAS until MS re-did their file writing process. This ment you could bring back virtually ANY revision of a document, often weeks old. File get corrupted at 4:00pm? Don't loose an entire day's work by going to last night's tape, just SALVAGE the previous save! As far as I am aware, no other common OS has this kind of recovery feature standard (as I indicated earlier, Mac has 90% of it, but you have to get a third party program to actually recover the file). This was a clear advantage of Netware over NT as a file server, and it is something I walked people through many times, to much praise, I might add. 8) Obviously, SALVAGE doesn't replace a good backup system, but it is an appreciated feature.
It doesn't work any more. Word and Excel in Office 95 and later seem to go out of their way to never "Delete" the old file (thus, making it availble to SALVAGE). I'm not pleased with this. And, knowing Excel's tendancy to corrupt files, it is a very sorely missed feature.
In case you were wondering, Lotus Smart Suite DOES follow the "Temp file/Delete/Rename" proceedure, so it isn't an issue of Windows 9x API.
As I said, I can't prove that it was an attempt to minimize a Netware feature or if it was simple incompetence, but it is gone. It appears that MS has shot themselves in their customer's foot.
Someone else (fairmang) here was kind enough to provide some hard numbers as to how much power would be available. His conclusion was 1mw. And that has to be divied up on that list of sub-systems (which was the point of the list, the things this tiny amount of power would have to do).
Processor: 1mw is a very small amount of power for a processor. For reference, I believe some of the Pentium II class chips are running around 15W. 15,000 times as much. Not a whole lot can be done with 1mw of power other than your watch. Keep in mind, your watch isn't trying to keep its own weight in mid-air, but rather just counting the oscilations of a quartz crystal.
A 1mw transmitter is a very small amount of power OUTPUT for a radio transmitter, and transmitter circuits aren't particularly efficient.
As for propulsion, gyros are a means of orientation and perhaps navigation, not propulsion (unless, perhaps, they are off balance!). Assuming all the 1mw of power went into propulsion, at near 100% efficiency (HA!), I doubt it could fight the thermal drafts around it, assuming it got airborn (a helium baloon tied to it might help it combat gravity...). Think about trying to fight the thermal drafts around a lightbulb...putting out thousands of times the energy our fly has available to keep it on or near course.
Anyone with an aeronautical engineering background know if there is some minimum energy expenditure required to keep an object of weight X aloft? (There would be more to the calculation than just weight, obviously) Or to fight a (say) 0.05m/s headwind? (My background is Electrical Eng...I can say with authority that 1mw of power is very, very little for the electronics. I can't say with authority that 1mw is currently or theoretically sufficient or insufficient to levitate this thing) If so, this thing could possibly be proven to be impossible as presented OR possible, given sufficiently efficient parts.
The thing might actually come close to being possible with near-term technology if instead of trying to have it generate or carry its own power, it were to be powered by an external radio signal -- the mylar wings could be used as a radio antenna. Feed it with a powerful, high-frequency radio source (a microwave oven magnetron would be close to the right frequency, at least in the neighborhood), you could extract enough power from this antenna as you needed, assuming you 1) kept the wings oriented properly to the radio source (going directly towards or away from the radio source) and 2) didn't mind frying things in line with the power supply's output.
>People who say that it can't be done should not interfere with those that are doing it Point well taken. That's a phrase I live by. I would not care if they are working with their own money, or with the money of investors who have volenteered to put it into this project. However, they are working on what, from what I can see, is very likely a thermodynamically impossible project on tax payer money. I'm not really happy about that idea. I'm more concerned about that then I am about any potential spying that this thing could ever do in my life time.
I wouldn't mind if they were working on pure research which *might* some day lead to a robofly, and after it gets airborn, wouldn't it be cool if they could control it, and then add a camera, and a transmitter... but that's not what they are saying. They said it would be hopefully flying in 2004, and the person who sold the Navy on this admitted that he didn't even know how a biological fly flies at the time he commited to the time schedule. For someone like that to say commit to a project like that just sounds too close to fraud to me.
Waste and corruption in government? Who would have guessed? (Yes, I realize $2.5M is small peanuts in the overall scheme of things)
1) It doesn't look like the thing is even flying yet. Much less having independant operation, much less sending back a signal, much less actually having a camera mounted on it. Sounds like this article is VERY, VERY premature. Sounds a bit like writing up Star Trek as a news story...reality is probably going to look a lot different.
2) They claimed it is solar powered. Now, I *really* have trouble imagining a solar power collection system providing enough power to actually make it fly. I believe plants have the most efficient solar energy collection and conversion systems around (I could be quite wrong on this, actually), and I've never seen a flying plant. No biological fly uses solar power. I can't believe this thing is going to fly purely on solar power, and I really question how they are going to put ANY power plant on the thing so small and yet power everything that needs to be powered: * Processor (and it will probably be quite a processor!) * Propulsion system * Video camera * Transmitter * Receiver (gotta be able to update its directions, eh?) * Some kind of energy storage system, as if it is supposed to go indoors, in rubble, etc., or work at night, it won't be able to be purely solar powered. This of course means that it will have to collect more than 100% of its energy requirements to "bank" the extra for when solar isn't available.
Up to the point the point where I saw "Robofly will be powered by the sun", I thought it was just some interesting research that may or may not lead anywhere near its original goal, but upon seing that it was to be solar powered, I'm starting to think it sounds more like fraud. Or a reporter who got something VERY wrong. Personally, I'd find any energy STORAGE system that could propel a tiny flying machine for more than a few seconds very, very facinating.
When I hit "Reply to This", I see your message above this message I'm typing. Not sure what settings cause that but I can (and did!) scroll back up and re-read your message while I respond to it.
If you ment to quote back the origional message *IN* the reply, I *strongly* (but hopefully, nicely 8) disagree. I don't enjoy reading the same message time and time again. I really don't enjoy reading MY message (and all its typos) time and time again!
I've also found that many people who have the option of quoting back the entire message often start to just insert one or two word replies mid-paragraph -- replies, which when actually found, are rarely worth the effort. It increases the amount of data without increasing the amount of meaning. And, it encourages people to write poorly (or encourages poor writers to write, not sure which)
As a long time user of micro-computers, I'm a very firm believer that there is no such thing as a revolution in this business. Usually, it is nothing more than a step of evolution which got good press.
A few common examples of "revolutionary" computers and the evolutions which led up to them: Macintosh: Decended DIRECTLY from Lisa and PARC stuff. Also, in the consumer market, check out the Epson QX-10 with Valdocs, an early What-you-see operating and working environment. Amiga: Any number of hardware graphics and sound subsystems which existed LONG before Amiga put them together.
Anyway, to my point. There were some interesting available source things that substantially pre-date the current Open Source movement.
My first "big" computer was a Heathkit H-100 (kit version of the Zenith Z-100) which I purchased in 1983, and didn't really retire until 1989 or so. 8088 and 8085 processors (Dual processors, not like you would think now, however), 5.25" floppies. Ran MS-DOS but it was NOT an IBM compatable. It was a very different design and it was in design concurently with the original IBM PC.
This was an Open Hardware system. When first shipped, the system came with a nice set of technical manuals (this was later dropped, a few owners thought it was required reading and went mad, I think). The entire main board, the video board, and the floppy controller were documented and detailed with very useful schematic diagrams and written circuit descriptions. Every chip was socketed on the main board. The Boot ROM source code was provided.
When the system first shipped, it had two operating systems, MS-DOS v1 and CP/M-80 v2.2 (for the 8085 processor). One very facinating feature of this system which was common at the time but nearly forgotten now was the concept of boot ROMs. The H/Z-100 used the boot ROM to load the OS, ONLY. The BIOS, something which has come to mean "ROM" on modern PCs, then ment the thing that tied the OS kernel (the part provided by MS or Digital Research) to the specific hardware. Unlike the modern PC (well, 32 bit OSs are kinda changing this), the BIOS of the Z-100 was in RAM, not ROM, loaded at boot time.
This lead to some rather interesting possibilities. One could modify the BIOS to support new hardware or alternate hardware configurations. I modified both CP/M and MS-DOS v1 (I *think* I did MS-DOS v2, don't recall for sure, however) to support four 5.25" floppy drives, and modified the hardware to support both 40 track and 80 track (the "high capacity" of the day) floppy drives. I'm not aware of anyone making commercially/publicly available "reved" BIOSs, nor do I think it would have been legal under the license agreements, nor am I aware of any other users other than me who ever touched the BIOS source code, but I would suspect that this BIOS code availablity might have helped with the number of OSs that ended up being supported on the H/Z-100 (MS-DOS v1 ->3.1, CP/M-80 v2.2 and 3.0, CP/M-86 (two different companies produced this) MPM-86 (again, two different companies), Concurrent DOS, UCSD p-System. These are just the OSs that *I* had (although, I don't know of any others that were available). At least FOUR radically different OS families, WOW!
This was an Open System. The only part of the system which wasn't open was the data separator card of the hard disk controller set (the HD controller was a two card set -- the controller itself and the data separator. Apparently, Zenith managed a few things on the data separator they were very proud of, and refused to document it, although they DID document the controller. Go figure. Of course, the card I ended up doing some interesting mods to was the data separator card, which required some interesting reverse engineering, but that's another story. 8)
This wasn't a unique trait of the H/Z-100. At that time, both CP/M and MS-DOS included an assembler, which gave a standard, baseline development tool. Quite a few other systems and machines included BIOS source. I remember having earned great praise when I altered the driver code for a Corvus hard disk to make it turn into drives A: and B: after boot so a very slow floppy app could be moved to a hard disk (this would have been in 1983).
This idea of providing BIOS source appears to have been lost when IBM put the BIOS in ROM, and later versions of DOS lost the included assembler.
At the time, there were quite a few other things that more closely resembled Open Source than this "available source" of the BIOS. At the time, there was a very substantial "public domain" software movement. Granted, there weren't any real apps in the public domain, but there were some pretty cool utilities. This seems to have died with the introduction of "Shareware" and the loss of a "universal" development language on the PC (alternately, one could look at this as the boom of alternative programming languages to the standard, OS supplied assemblers).
Anyway, I guess if there is a point to this, I would say there are two relevant to the Open Source movement. One, is it isn't really a totally new concept. Two, the previous examples of things similar to Open Source faded out for various reasons, it would probably be foolish to assume that Open Source is here to stay without continued effort from its advocates. I think people barely noticed the loss of source code in the mid-1980s.
What happens if a developer (no names, but the temptation was high to name one) starts developing software on an over-clocked chip? And we all suddenly need that much performance to just get by?
A DVD needed to distribute the OS. *shudder*
Hey! My shortest/. post ever! (and most meaningless, probably) Nick.
I'm not talking about connections, I'm talking about backup strategies.
There are a few things I insist on with backups: * Rotation * Off-site (at least ability to do it)
Rotation means you are not backing up twice in a row on the same media. Why? The most likely time for a drive to fail is DURRING the backup, it is the very time the most intensive disk activity takes place, and the only time the ENTIRE disk system is read. IF the drive fails durring the backup, you have 1) a dead drive and the need for a good backup and 2) just clobbered your only good backup.
Further, the most common reason to go to the backup is NOT drive failure! It is user error (Oops...didn't mean to delete that!) or data corruption (program error). VERY OFTEN, the problem won't be found for days (weeks? MONTHS??). You *HAVE* to have some kind of history to go back to. I tell my barely trainable clients to use a one-week rotation (five media), the more sophisticated will often do a second rotation of weekly backups (for example, four Wednesday tapes), and the truly enlightened will pull a monthly tape for permanent archiving.
Off-site: For serious use, some kind of off-site backup is important. Most people just toss their tapes on top their computer and walk away. In the event of fire, flood, theft, roof leak, whatever, very often the server is lost -- ALONG WITH THE BACKUPS! In many data-oriented businesses, they can set up in someone's basement, but they need their data. These people need some kind of off-site backup to rebuild their business in the event of a site-disaster.
I would argue that an improper backup strategy is worse than no backups at all. If you aren't doing backups, you usually know you are doing something stupid and living on the edge. Doing improper backups causes people to think they are "mostly safe", and that "most" of their data will come back...but that is VERY rarely the case.
By the way, RAID has nothing to do with backups. RAID will keep you running in the event of a drive failure. Doesn't do squat in the event of data corruption, accidental deletion, theft, fire, etc. It also doesn't help you if your system has a controller failure and you can't find a compatable RAID controller to restore your system to operation (people tend to forget that. They rarely have a spare controller in the closet). I yell at any of my clients who mention RAID and Backups in the same sentance.
Part of me keeps having this thought that many of the people here are "hobbiest users", and perhaps this really doesn't apply. Reality keeps reminding me just how long it would take to just set up 60+G of data, assuming you actually have the original data someplace. If you are actually CREATING data, forget it. Hobbiest users need backups, too.
I'm very tough on people using backups. I will and have dropped clients who don't do proper backups. I can afford to have anyone NOT as a client, I can't afford to have them as an unhappy client, and if they aren't doing their backups properly, they will be an unhappy client someday.
O.k., 73G (or more accurately, a lot less!) on one drive is cool, but what about backup?
Virtually all backup devices are measured using "compressed" capacity, which is bogus, at best. Fraud might be a better word. Usually, on a big drive, you have either huge databases (which typically compress well) or lots of graphics and sound files, which compress hardly at all. Selling a drive based on its compressed capacity is kinda like measuring the interior space of a car by including the roof rack and the potential trailer you could tow.
I'm a little concerned that backup technology isn't really keeping up with HD technology. I'm even more concerned that hardly anyone pays attention to backup technology around here -- I only saw one person ask how you would back the thing up, and the one reply was to another drive. Having one on-line backup is NOT a backup strategy!
It scares the heck out of me to see people buying 10+G drives without a thought to backup. Even if it is all programs, trying to rebuild a system after a data loss is very, very time consuming.
I'm also a little ticked off over the quality of backup devices. I've replaced far more tape drives in my client's servers than I have hard disks. Really pathetic. A friend has assured me DLTs are much better than the DATs I normally recommend, and that may be true, but $3000 for a drive and $100 per media, well, that's a few DAT drives. I gave up on Travan drives on servers -- I've had astronomical failure rates on both drives and tapes, but curiously, they seem to do o.k. on Windows 9x workstations.
Actually, I had a couple people tell me RAM prices were on their way up a few days BEFORE the earthquake.
The earthquake increased the effect, but prices were already on their way up. I've seen this repeatedly -- something happens some place that has something to do with semiconductor production, and RAM prices shoot up before the pipeline has a chance to go dry, and hang there as long as possible. That's the way a market economy works. You charge what you can get.
Observation (getting further off topic): We always seem to need to add $100-$150 worth of RAM to our computers: Back when 48K was common, we paid $150 or so to up it to 64k. When 64k was common, we paid $120 or so to up it to 128k. ... (I'll spare you the rest of this progression) Now, we want to add another 128M to our systems, but I'll wait until the price drops back down to $150... 8)
Warning: The comments contained here are from a very new OpenBSD user... (Operation system in place for about a week. Got interested in it about three weeks ago).
No, OpenBSD's install is not as friendly as the new Linux installs. The disk partitioning is really exciting. Hint: Start with a totally empty disk. When you manage to get partitions where you want them, put a small DOS partition on and work around it. Another hint: If you don't grab a calculator or at least a piece of paper, you are either very good at memorizing numbers and doing math in your head or you just nuked that partition you were trying to work around. The first time I installed it on a Compaq (with the configuration partition), I wiped the disk completely clean. Not even the OpenBSD partition survived the install! I think it took me two or three additional attempts (i.e., reloads) before I got it right. Suddenly, it becomes easy, even logical 8). I'm *very* serious about practicing on a "spare" hard disk first.
I found the FTP install actually worked very nicely with OpenBSD, something I could NOT get to work with the otherwise more friendly FreeBSD.
I will dispute any claims that documentation is better on Linux than *BSD. The Linux documentation is certainly BIGGER, at least on mall store shelves, but hardly better. You see, Linux is a rapidly moving target. The details of every distribution change every release, and in fact, some of the products might even be perceived as completely different OSs by a non-X familiar person. While I have been playing with Linux for YEARS (key word: Playing. I'm not yet a Linux master), I've been pleased that OpenBSD seems to be more "familiar" quickly based on my Unix experience. *BSD is much more standard Unix than Linux is. A book on RedHat v5.2 will not do much to help you get RedHat 6.0 up and running.
This is good from a learning perspective. Unix books cover both OpenBSD and FreeBSD much better than they do Linux. On the other hand, I really miss 'bash'. HOWEVER, *I* will impliment bash, I will know how it got there, I will know how to remove it or change it or whatever. You loose this in the modern Linux distributions. It is all done for you.
Another problem with Linux documentation is the quality. Everyone sees it as a cash cow, everyone is writing a book. Most of them stink. Several years ago, I bought a book on Linux. Durring my first five minutes with the book, I spotted THREE blatent FACTUAL errors (Linus wrote Minix then improved it to Linux was one of them). A few are good. Most are obsolete.
As far as I can tell for multiple booting, to boot OpenBSD, set the partition active. Therefore, it will probably work with most boot management software.
Hardware support: Well, OpenBSD seems to support "standard" stuff. IDE, main-stream SCSI. Haven't played with the sound. It recognized every network card I threw at it (at least four totally different cards). Forget USB for the time being, and I wouldn't wait up.
A *BIG* comment: I know a lot of people (I do it myself once in a while) seem to think the magic of a new system is installing it. This is not true. Not at all. As long as you think installing it is the major hurdle, you haven't really learned the product. The magic is what you do ONCE it is installed. How you set it up for efficient use. How you back it up. How you recover it when you woof a hard disk (or a few strategic files). How you recover it when you woof or upgrade a main board. I clean up a lot of networks set up by people who didn't understand that installing the software is just the very, very tip of the iceberg.
"...unlike Windows NT where Compaq, Data General, Hewlett-Packard, IBM, and Unisys provide 99.9 percent system-level uptime guarantees for Windows NT-based servers."
Assuming it takes 10 minutes to identify a crashed server, power-cycle it, and wait for it to reboot (this time varies GREATLY different systems), 99.9% availabity would be one crash per week (I did the math). I'm *SURE* they are not counting "planned" down times (You know, you have to reboot now that you made that change), and I am pretty sure they aren't counting hardware or power failures or other things that are quite outside of MS's control.
I don't consider a crash per week acceptable. Sorry.
If I don't hit 99.9% of the other cars on the road when I'm driving, am I a good driver?
I have a client who had a Novell server go over 900 days without reboot (A two-day power failure finally took it down). I think I can say quite certainly _No_ NT server in history has done that. That is the way things are supposed to work. Linux seems a heck of a lot closer to my ideals than NT is.
I can't speak for DG, IBM, or Unisys, but I have seen Compaq's "Guaranteed four hour service" program in action at a couple clients. Never seen a client broke for less than 24 hours with that program, either. I don't trust Compaq's "guarantees" when it comes to servicing their own products, how can they guarantee someone else's software??
In the 1950s through the 1970s, people bought cars based on features and looks, not based on reliability and quality. There were a few cars that were well built and lasted, but for the most part, they rotted on the lots, while their flashy but unreliable competition sold.
People used to line up to see the new models. People with money lined up to BUY the new models. Knowing full well it probably didn't start any better than the old one.
It wasn't that Detroit was incapable of BUYING a quality product. It was that the consumer was unwilling to buy the boring but solid product. It wasn't until the late 70s and early 80s that consumers suddenly demanded QUALITY over LOOKS from Detroit, and Detroit responded (in the auto industry, the problem was the speed at which the change in consumer attitude took place. You can't change the manufacturing criteria from flash to quality overnight..and no one was sure if this was a passing fad or a real trend)
The facts are, if the consumer demands something with their money, they will get it. Complaints don't mean a thing.
Another example: Airlines. People complain about late flights, they complain about lousy service, but they book the cheapest flight. Duh! Leaving on time costs money! Great service costs money! If the consumer buys the cheapest product, they can complain until they are blue in the face, nothing will change, at least not for the better.
It is simple economics. There is great demand for flashy software. There is little demand for quality software. While that is the case, that's what the consumer gets. The software industry better hope consumers don't change their mind on flash vs. quality overnight, like happened with the auto industry.
I've been supporting bad software for many years now. I'm starting to detect a change in attitudes towards business people (although, I am probably guilty of contaminating the sample!). I think this is good.
Consumer complaints don't enter into economic decisions. Dollars do.
I have near zero experience with OpenBSD so far, but I did try a couple installations.
Yeah, the OpenBSD disk layout program is really bizzare, at least to those of us who have used MS-DOS systems too much. When reading through the FreeBSD documentation that came with the Walnut Creek package, they shed some light on the *BSD disk model -- it appears to have been developed absolutely independantly of the IBM PC's HD layout, and they have different ideas about how things are done. When BSD was ported to the PC platform, the disk partitioning model was ported, too. Knowing FDISK probably hurts when laying out an OpenBSD system.
As I understand it, you basicly create a IBM/MS style partition which reserves space for OpenBSD. THIS becomes the BSD "disk". Now, this "disk" is partitioned. Kinda like "Extended partitions", but totally different 8)
My first attempt to load OpenBSD was on a generic clone, empty IDE hard disk, 3Com network card, FTP install. Worked great. I then tried putting it on a machine which I wanted to have it run on (Compaq Deskpro XL 575). Had more problems, as there was the infamous Compaq system maintence partition on the drive already, and a small DOS partition. But, the install appeared to go well. Until reboot. At which point, the system refused to boot, and upon investigation, the disk was blank, except for the OpenBSD boot loader. No OpenBSD partition, no DOS partition, no Compaq system partition (and those are a pain to reload).
I went back to FreeBSD after that. I'll try OpenBSD again, esp. once I found the explaination of the BSD disk concepts in the Walnut Creek FreeBSD book, but yes, for newbies, it would be best to practice on a BLANK hard disk at first. And of course, your backup are up to date, aren't they??
As someone else said, FreeBSD has a much nicer disk preparation program.
It *appears* it depends upon the service provider. Some of the service providers apparently actively LOOK for people running web/other Internet servers, while others are more indifferent. My service provider said it was against the rules, but upon asking, they told me they don't actively look, but they will investigate any curious or excessive activity.
My web server has been on line for about a year now, without any complaints from the cable company, other than the time some (*censored*) cracked my system and set up a cracking shop, using my machine as a base to crack others. Got a nasty note from the cable company which I couldn't figured out until I realized the odd things happening on my Linux box were really odd things, not just ignorance on my part!
I do wish to point out my web server is STRICTLY for experimentation and occasionally getting stuff to friends. I do think using Cable Internet Access to do commercial or high-volume stuff is really tacky. We get high bandwidth at a low cost because it is assumed we will be rarely using most of the bandwidth.
I'm with the @Home service, but a friend who is waiting for Cable Internet Access who would also be getting @Home through a different cable company, sounds like he would be getting a totally different set of terms of service than I have.
Best I can say is anyone who says "This is how it is" is probably wrong SOMEWHERE. The rules are different everywhere. The rules are still being written...
Here, they recently put a cap on the "upstream" link (128kbps supposedly), which I am fine with...means I don't have to worry quite so much in case someone finds something interesting on my server...
Oh, yeah... There are a LOT of things about ADD (ADHD, whatever) that are in the "traditional" literature that is Just Plain Wrong.
A couple of my favorites (to some degree, this is my speculation. Were I degreed in the medical arts, it would be an official guess, but in my case, it is just my guess): 1) "You grow out of it". No, you learn to adapt. School is probably one of the most ADD hostile areas of anyone's life. Leave school, Wow! problem Gone or at least Greatly Diminished.
2) It doesn't affect women. Bull. Many years ago, it was supposedly 20 men to one woman (sorry..at the time, you out-grew it, too, so it was boys to girls). A few years ago, it was seven to one. Last number I heard was three to one. There's a trend here! My guess: Boys are diagnosed and treated, girls are stereotyped. Until fairly recently, girls were expected to grow up to be mothers and wives -- a girl who "didn't fit in" was allowed to fall through the cracks and fail, after all, she had the Mrs. safety net.
There is (possibly) some correlation between hair color and ADD. Guess what hair colors are more likely to have ADD? Redheads and Blondes! A blond boy is considered ADD, a blonde girl is a ditz.
I know a woman who showed such obvious ADD and hyperactivity it was unmistakable, but since she is a woman, it was attributed to hormones, her disasterous marriage, etc. Everything but the obvious. She had even seen some professionals which really should have jumped on it -- it was obvious.
When I was in school, after considerable testing I was labeled "Learning Disabled" with no more accurate diagnosis. This was in 1979, dyslexia was just starting to be recognized, but ADD was still a euphamism for "Hyperactive" and the diagnosis of hyperactivity was so CLEARLY wrong (I tell people if I were any less hyperactive, I'd be diagnosed as DEAD!). In high school, we basicly decided I was dyslexic. About four years ago (LONG after my very difficult but ultimately successful school (I lived) life was over), I finally realized that I was also ADD, something that we also realized ran in my family like a bull through a china shop.
Another observation: The lines between ADD and many other disorders are very fuzzy. If the diagnosis explains things, it is useful. It seems most "disorders" ("reorders"?) of the mind are diagnosed by symptom matching, rather than by isolating and diagnosing a root cause. The manifestations vary widely. Personally, I noticed an initially unexpected similarity between ADD and Tourette's syndrom: a difference in focus and intensity, but seems to me to be of similar root causes. Most people think I'm off my rocker when I say that (hey, I'm used to it...), but I've also been told that some research along those lines is starting to be done.
Ah, so much more to say...but I fear I've wandered more than far enough off topic...
Slashdot Mirror
If we define a Trojan Horse as a program which appears to do one thing -- perhaps productive, perhaps not -- and actually does something else, such as breach the heck out of your system security, could it not be argued that Windows, Internet Explorer and Outlook could be catagorized as Trojan Horses?
Just a thought.
Nick.
Hear Hear!!
I can't think of any OS in recent history where simply getting the system to boot should qualify as any kind of victory. Getting an OS to boot may be a milepost, but only the very first milepost in a very long trip.
I also agree the recent "tech" stuff is pathetic. Remember when Byte magazine used to have schematic diagrams? Remember Byte magazine?? So much now is so absurdly platform specific, and so many reviewers claiming to be experts on things they are they shouldn't, and condeming "competing" or differing products they show no real experience in or understanding of.
I'm also not impressed that this reviewer would even mention an attempt to do an install on unreliable hardware, or a dial-up FTP install. I've tried a few FTP installs through my cable internet access. OpenBSD is the only one that really went well! Now, I think I will credit this more to the site that I pulled it down from than the product itself, but I have to say OpenBSD's FTP install can work, and can work very well, at least if you don't bring an expectation that it is "just like Linux" (or DOS, or Windows or...) to the process.
Nick.
I would beg to differ on this.
While I would agree there is a lot of stuff written on Linux, I'd question the value of most of it. The real issue is how much GOOD stuff is written on the subjects, and how much cr*p you have to dig through to find it.
I have been playing with Linux on-and-off since probably 1994 or 1995. I did my first OpenBSD and FreeBSD installs only a month or so ago. I felt more at home with OpenBSD and FreeBSD within minutes than I did with any version of Linux..ever.
Why? Simply because *BSD more closely matched the considerable Unix documentation that is out there. Linux is a changing target. By the time a release is properly documented, it is being replaced by a new release.
I also found the documentation available for *BSD to be more specific to the actual package I had than the Linux documentation. The Linux docs are considerable and some are excelent quality -- but the first challenge is trying to figure out what actually got installed and what is applicable and what isn't to the particular distribution that is loaded on the machine in front of me.
To me, it feels like Linux is a lot of pieces bolted together. Each piece is well documented, but how the pieces interact and work together is a challenge to figure out. OpenBSD seems to be a more coherent system.
I wanted to turn OpenBSD into a NAT system for my cable internet access. Within a relatively few hours of fiddling and fighting with wierd hardware, it was running and operating perfectly. I spent far more time figuring out the logic of the partitioning program (and cleaning up the mess after I discovered I was wrong...time after time! 8) and fighting with a SCSI controller supported by nothing in the world other than DOS and Windows 95 than I did getting NAT going.
Don't get me wrong... I respect Linux greatly, and I do not question that it is a much better desktop OS than *BSD is for a number of reasons. (One of which is the hype factor -- a perceived popular OS will have more software support, and this is important.) However, *BSD has a very significant place where stable, secure and reliable operation is more important than flash. OpenBSD has impressed the heck out of me very quickly. And, it sounds like the only "Internet" OS that really has security as something other than an afterthought.
Nick.
Yeah... Agreed....
Esp. in an area like the Internet and Internet patents, where things are so new and so unproven. Most likely, some of these cases are going to have to go through the courts (ALL the way through the courts!) to get set some guidelines set. Until then, the victor would probably be the people who have the deeper pockets to wave around the high-priced legal team and scare their competitors into submission.
The only potentially "cheap" thing I can think of doing is to find a legal team who would take the case pro-bono. It would (probably) have to be either an "easy" victory and/or a high visability case (i.e., very hard victory, but landmark-decision type case). It would also require a law practice which is not starving for money but wanting to increase their image in the area of Internet patent law. There are probably a lot of law offices which might be willing to take the right case like this, but again, you are stuck trying to find a law office of QUALITY.
I don't think I helped much, eh?
Nick.
Yes, MS-DOS had an undelete program, but it made no effort to avoid overwriting the old files. MS-DOS (and apparently, Windows 9x) uses the lowest number available block (closest to track 0, outer edge of the disk) first when writing files. So, if you delete a file (physically) at the beginning of the disk, it is overwritten VERY quickly. If you delete a file physically near the end of the disk and there is lots of free space earlier up on the disk, you may be able to recover that file for quite some time. Your ability to recover a file under DOS was based on luck and what you have done since then. There is no deliberate plan to avoid overwriting deleted files.
Hmm....now that I think about it, DOS v6 *did* include an optional device driver which would track deleted files in a comparatively primative way, but it was there (and I assume it worked, at least if you remembered to load it BEFORE you made a mistaken deletion!). I'm curious if it worked only from the command line (i.e., it watched for use of the DELETE command, like the Win9x recycle bin) or if it worked at the API level (so a program (like MS-Office 4.x) that deleted a file would have it tracked). No idea -- I never played with the deletion tracking feature of DOS. (In case you were wondering, the Netware deletion tracking tracked at the API level..doesn't matter if you DEL *.*'ed or the program deleted the file, Netware tracked it. Netware even tracks the file if you delete its parent directory).
Virtually any file system on any OS can have files recovered from deletion if new files haven't actually overwriten the "ashes" of the old files. This has not a thing to do with the MS-DOS FAT file system. DOS's file system was simple enough, however, that it was fairly easy to write undeletion programs. The first one I saw was a customer of mine wrote one on MS-DOS v1 (a FANTASICLY clever kid, but a bit clumsy when it came to deleting files). Under CP/M, undeleting files was SO easy, I always just used a disk editor (I think undelete programs existed, they were just barely needed).
Netware is rare in that it actually TRACKS those old "ashes" and keeps them around and intact, with all information (including the userID of the deleter!) so they can be reliably and successfully recovered as long as possible (or until told to do otherwise).
I do seem to have wandered well off-topic.. I wasn't writing in praise of Netware (I am very willing to do this, as people have undoubtably noticed!), as I have the knawing suspicion in this forum, most people wouldn't care. My point was to bring up Yet Another potentially anti-competitive action of Microsoft. The idea of writing a temporary file before deleting the original would give you an improved likelyhood of being able to undelete a newly replaced file even under DOS if you got to it soon enough. However, this Temporary/Delete/Rename process did WONDERFUL things when combined with the Netware deletion tracking system.
Nick.
Y'all can wake up now, I'm finished.
>Will I be flamed if I say Novell rules!?
Not by me, you won't!
Netware is my choice for file and print services.
I don't do NT. I have yet to see an NT network I'd have been proud to put my name on. (I told this to an NT installer once. He listened to my points, agreed with them. Then said "Yeah, but you make much more MONEY with NT!")
I haven't become comfortable enough with Linux and FreeBSD's problem recovery process to put it at my clients as a file/print sharing system yet. But hopefully, soon. 8)
Nick.
I became aware of another potentially anti-competitive thing MS did some years back, which, as far as I am aware no one else has mentioned in the press before. If it wasn't anti-competitive, it was at least stupid, and MS users are the ultimate victims.
.BAK (first deleting any old .BAK file).
A standard proceedure for saving files used in "single user" apps, like word processors and spreadsheets, was developed many, many years ago, and used almost universally. The proceedure was you DON'T just overwrite the old file, you do the following:
* write the new file to a temporary file name
* Delete the old file -- or rename it to
* Rename the temporary file to the original file name.
This made sure that at no time was the only copy of the file on the disk deleted or dammaged. If the program crashed durring the save cycle, the original copy still existed. Great idea. Saved a LOT of people, and this is why it was used almost universally, and at Microsoft through Office v4.3.
Starting with Office 95, however, they changed the saving proceedure:
* Reset file pointer to beginning of file
* Overwrite existing file with new data
Obviously, this leaves the system at a fairly delicate point, should the app or OS crash after the file has started to be overwritten.
So far, things look just stupid. However, if we remember one of MSs chief rivals, Novell, and a feature of Netware 3.x and above, things look a little less innocent. Netware has a very nifty "undelete" feature, they refer to as "Salvage"ing files. Unlike most other OSs where this feature is provided by third party add-ons, with Netware, this is a CORE part of the OS. A deleted file is marked as deleted and put into the free disk space pool, but it isn't overwritten until the entire disk has been used, at which point the oldest deleted files are overwritten (Actually, Macintosh does much the same thing). Novell provides you a nifty utility (SALVAGE in NW3.x, FILER in NW4+) which can bring these files back up to the point they are physically overwritten (or Netware is told to "PURGE" the deleted files).
This is a really neat feature, or at least, WAS until MS re-did their file writing process. This ment you could bring back virtually ANY revision of a document, often weeks old. File get corrupted at 4:00pm? Don't loose an entire day's work by going to last night's tape, just SALVAGE the previous save! As far as I am aware, no other common OS has this kind of recovery feature standard (as I indicated earlier, Mac has 90% of it, but you have to get a third party program to actually recover the file). This was a clear advantage of Netware over NT as a file server, and it is something I walked people through many times, to much praise, I might add. 8) Obviously, SALVAGE doesn't replace a good backup system, but it is an appreciated feature.
It doesn't work any more. Word and Excel in Office 95 and later seem to go out of their way to never "Delete" the old file (thus, making it availble to SALVAGE). I'm not pleased with this. And, knowing Excel's tendancy to corrupt files, it is a very sorely missed feature.
In case you were wondering, Lotus Smart Suite DOES follow the "Temp file/Delete/Rename" proceedure, so it isn't an issue of Windows 9x API.
As I said, I can't prove that it was an attempt to minimize a Netware feature or if it was simple incompetence, but it is gone. It appears that MS has shot themselves in their customer's foot.
Nick.
Someone else (fairmang) here was kind enough to provide some hard numbers as to how much power would be available. His conclusion was 1mw. And that has to be divied up on that list of sub-systems (which was the point of the list, the things this tiny amount of power would have to do).
Processor: 1mw is a very small amount of power for a processor. For reference, I believe some of the Pentium II class chips are running around 15W. 15,000 times as much. Not a whole lot can be done with 1mw of power other than your watch. Keep in mind, your watch isn't trying to keep its own weight in mid-air, but rather just counting the oscilations of a quartz crystal.
A 1mw transmitter is a very small amount of power OUTPUT for a radio transmitter, and transmitter circuits aren't particularly efficient.
As for propulsion, gyros are a means of orientation and perhaps navigation, not propulsion (unless, perhaps, they are off balance!). Assuming all the 1mw of power went into propulsion, at near 100% efficiency (HA!), I doubt it could fight the thermal drafts around it, assuming it got airborn (a helium baloon tied to it might help it combat gravity...). Think about trying to fight the thermal drafts around a lightbulb...putting out thousands of times the energy our fly has available to keep it on or near course.
Anyone with an aeronautical engineering background know if there is some minimum energy expenditure required to keep an object of weight X aloft? (There would be more to the calculation than just weight, obviously) Or to fight a (say) 0.05m/s headwind? (My background is Electrical Eng...I can say with authority that 1mw of power is very, very little for the electronics. I can't say with authority that 1mw is currently or theoretically sufficient or insufficient to levitate this thing) If so, this thing could possibly be proven to be impossible as presented OR possible, given sufficiently efficient parts.
The thing might actually come close to being possible with near-term technology if instead of trying to have it generate or carry its own power, it were to be powered by an external radio signal -- the mylar wings could be used as a radio antenna. Feed it with a powerful, high-frequency radio source (a microwave oven magnetron would be close to the right frequency, at least in the neighborhood), you could extract enough power from this antenna as you needed, assuming you 1) kept the wings oriented properly to the radio source (going directly towards or away from the radio source) and 2) didn't mind frying things in line with the power supply's output.
>People who say that it can't be done should not interfere with those that are doing it
Point well taken. That's a phrase I live by. I would not care if they are working with their own money, or with the money of investors who have volenteered to put it into this project. However, they are working on what, from what I can see, is very likely a thermodynamically impossible project on tax payer money. I'm not really happy about that idea. I'm more concerned about that then I am about any potential spying that this thing could ever do in my life time.
I wouldn't mind if they were working on pure research which *might* some day lead to a robofly, and after it gets airborn, wouldn't it be cool if they could control it, and then add a camera, and a transmitter... but that's not what they are saying. They said it would be hopefully flying in 2004, and the person who sold the Navy on this admitted that he didn't even know how a biological fly flies at the time he commited to the time schedule. For someone like that to say commit to a project like that just sounds too close to fraud to me.
Waste and corruption in government? Who would have guessed? (Yes, I realize $2.5M is small peanuts in the overall scheme of things)
Nick.
A couple things caught my eye about this article.
1) It doesn't look like the thing is even flying yet. Much less having independant operation, much less sending back a signal, much less actually having a camera mounted on it. Sounds like this article is VERY, VERY premature. Sounds a bit like writing up Star Trek as a news story...reality is probably going to look a lot different.
2) They claimed it is solar powered. Now, I *really* have trouble imagining a solar power collection system providing enough power to actually make it fly. I believe plants have the most efficient solar energy collection and conversion systems around (I could be quite wrong on this, actually), and I've never seen a flying plant. No biological fly uses solar power. I can't believe this thing is going to fly purely on solar power, and I really question how they are going to put ANY power plant on the thing so small and yet power everything that needs to be powered:
* Processor (and it will probably be quite a processor!)
* Propulsion system
* Video camera
* Transmitter
* Receiver (gotta be able to update its directions, eh?)
* Some kind of energy storage system, as if it is supposed to go indoors, in rubble, etc., or work at night, it won't be able to be purely solar powered. This of course means that it will have to collect more than 100% of its energy requirements to "bank" the extra for when solar isn't available.
Up to the point the point where I saw "Robofly will be powered by the sun", I thought it was just some interesting research that may or may not lead anywhere near its original goal, but upon seing that it was to be solar powered, I'm starting to think it sounds more like fraud. Or a reporter who got something VERY wrong. Personally, I'd find any energy STORAGE system that could propel a tiny flying machine for more than a few seconds very, very facinating.
Nick.
WOW! Great link!!
Original Source Material!
(What a concept!)
It appears Thomas Gold has anticipated most of the comments and criticisms that have been posted here on Slashdot.
I have this knawing suspicion I am going to accomplish very little today.
Nick.
When I hit "Reply to This", I see your message above this message I'm typing. Not sure what settings cause that but I can (and did!) scroll back up and re-read your message while I respond to it.
If you ment to quote back the origional message *IN* the reply, I *strongly* (but hopefully, nicely 8) disagree. I don't enjoy reading the same message time and time again. I really don't enjoy reading MY message (and all its typos) time and time again!
I've also found that many people who have the option of quoting back the entire message often start to just insert one or two word replies mid-paragraph -- replies, which when actually found, are rarely worth the effort. It increases the amount of data without increasing the amount of meaning. And, it encourages people to write poorly (or encourages poor writers to write, not sure which)
Nick.
An Etch-a-Sketch only crashes when you drop it. 8-)
As a long time user of micro-computers, I'm a very firm believer that there is no such thing as a revolution in this business. Usually, it is nothing more than a step of evolution which got good press.
A few common examples of "revolutionary" computers and the evolutions which led up to them:
Macintosh: Decended DIRECTLY from Lisa and PARC stuff. Also, in the consumer market, check out the Epson QX-10 with Valdocs, an early What-you-see operating and working environment.
Amiga: Any number of hardware graphics and sound subsystems which existed LONG before Amiga put them together.
Anyway, to my point. There were some interesting available source things that substantially pre-date the current Open Source movement.
My first "big" computer was a Heathkit H-100 (kit version of the Zenith Z-100) which I purchased in 1983, and didn't really retire until 1989 or so. 8088 and 8085 processors (Dual processors, not like you would think now, however), 5.25" floppies. Ran MS-DOS but it was NOT an IBM compatable. It was a very different design and it was in design concurently with the original IBM PC.
This was an Open Hardware system. When first shipped, the system came with a nice set of technical manuals (this was later dropped, a few owners thought it was required reading and went mad, I think). The entire main board, the video board, and the floppy controller were documented and detailed with very useful schematic diagrams and written circuit descriptions. Every chip was socketed on the main board. The Boot ROM source code was provided.
When the system first shipped, it had two operating systems, MS-DOS v1 and CP/M-80 v2.2 (for the 8085 processor). One very facinating feature of this system which was common at the time but nearly forgotten now was the concept of boot ROMs. The H/Z-100 used the boot ROM to load the OS, ONLY. The BIOS, something which has come to mean "ROM" on modern PCs, then ment the thing that tied the OS kernel (the part provided by MS or Digital Research) to the specific hardware. Unlike the modern PC (well, 32 bit OSs are kinda changing this), the BIOS of the Z-100 was in RAM, not ROM, loaded at boot time.
This lead to some rather interesting possibilities. One could modify the BIOS to support new hardware or alternate hardware configurations. I modified both CP/M and MS-DOS v1 (I *think* I did MS-DOS v2, don't recall for sure, however) to support four 5.25" floppy drives, and modified the hardware to support both 40 track and 80 track (the "high capacity" of the day) floppy drives. I'm not aware of anyone making commercially/publicly available "reved" BIOSs, nor do I think it would have been legal under the license agreements, nor am I aware of any other users other than me who ever touched the BIOS source code, but I would suspect that this BIOS code availablity might have helped with the number of OSs that ended up being supported on the H/Z-100 (MS-DOS v1 ->3.1, CP/M-80 v2.2 and 3.0, CP/M-86 (two different companies produced this) MPM-86 (again, two different companies), Concurrent DOS, UCSD p-System. These are just the OSs that *I* had (although, I don't know of any others that were available). At least FOUR radically different OS families, WOW!
This was an Open System. The only part of the system which wasn't open was the data separator card of the hard disk controller set (the HD controller was a two card set -- the controller itself and the data separator. Apparently, Zenith managed a few things on the data separator they were very proud of, and refused to document it, although they DID document the controller. Go figure. Of course, the card I ended up doing some interesting mods to was the data separator card, which required some interesting reverse engineering, but that's another story. 8)
This wasn't a unique trait of the H/Z-100. At that time, both CP/M and MS-DOS included an assembler, which gave a standard, baseline development tool. Quite a few other systems and machines included BIOS source. I remember having earned great praise when I altered the driver code for a Corvus hard disk to make it turn into drives A: and B: after boot so a very slow floppy app could be moved to a hard disk (this would have been in 1983).
This idea of providing BIOS source appears to have been lost when IBM put the BIOS in ROM, and later versions of DOS lost the included assembler.
At the time, there were quite a few other things that more closely resembled Open Source than this "available source" of the BIOS. At the time, there was a very substantial "public domain" software movement. Granted, there weren't any real apps in the public domain, but there were some pretty cool utilities. This seems to have died with the introduction of "Shareware" and the loss of a "universal" development language on the PC (alternately, one could look at this as the boom of alternative programming languages to the standard, OS supplied assemblers).
Anyway, I guess if there is a point to this, I would say there are two relevant to the Open Source movement. One, is it isn't really a totally new concept. Two, the previous examples of things similar to Open Source faded out for various reasons, it would probably be foolish to assume that Open Source is here to stay without continued effort from its advocates. I think people barely noticed the loss of source code in the mid-1980s.
Nick.
What happens if a developer (no names, but the temptation was high to name one) starts developing software on an over-clocked chip? And we all suddenly need that much performance to just get by?
/. post ever!
A DVD needed to distribute the OS.
*shudder*
Hey! My shortest
(and most meaningless, probably)
Nick.
I'm not talking about connections, I'm talking about backup strategies.
There are a few things I insist on with backups:
* Rotation
* Off-site (at least ability to do it)
Rotation means you are not backing up twice in a row on the same media. Why? The most likely time for a drive to fail is DURRING the backup, it is the very time the most intensive disk activity takes place, and the only time the ENTIRE disk system is read. IF the drive fails durring the backup, you have 1) a dead drive and the need for a good backup and 2) just clobbered your only good backup.
Further, the most common reason to go to the backup is NOT drive failure! It is user error (Oops...didn't mean to delete that!) or data corruption (program error). VERY OFTEN, the problem won't be found for days (weeks? MONTHS??). You *HAVE* to have some kind of history to go back to. I tell my barely trainable clients to use a one-week rotation (five media), the more sophisticated will often do a second rotation of weekly backups (for example, four Wednesday tapes), and the truly enlightened will pull a monthly tape for permanent archiving.
Off-site: For serious use, some kind of off-site backup is important. Most people just toss their tapes on top their computer and walk away. In the event of fire, flood, theft, roof leak, whatever, very often the server is lost -- ALONG WITH THE BACKUPS! In many data-oriented businesses, they can set up in someone's basement, but they need their data. These people need some kind of off-site backup to rebuild their business in the event of a site-disaster.
I would argue that an improper backup strategy is worse than no backups at all. If you aren't doing backups, you usually know you are doing something stupid and living on the edge. Doing improper backups causes people to think they are "mostly safe", and that "most" of their data will come back...but that is VERY rarely the case.
By the way, RAID has nothing to do with backups. RAID will keep you running in the event of a drive failure. Doesn't do squat in the event of data corruption, accidental deletion, theft, fire, etc. It also doesn't help you if your system has a controller failure and you can't find a compatable RAID controller to restore your system to operation (people tend to forget that. They rarely have a spare controller in the closet). I yell at any of my clients who mention RAID and Backups in the same sentance.
Part of me keeps having this thought that many of the people here are "hobbiest users", and perhaps this really doesn't apply. Reality keeps reminding me just how long it would take to just set up 60+G of data, assuming you actually have the original data someplace. If you are actually CREATING data, forget it. Hobbiest users need backups, too.
I'm very tough on people using backups. I will and have dropped clients who don't do proper backups. I can afford to have anyone NOT as a client, I can't afford to have them as an unhappy client, and if they aren't doing their backups properly, they will be an unhappy client someday.
Nick.
The Backup Nazi.
O.k., 73G (or more accurately, a lot less!) on one drive is cool, but what about backup?
Virtually all backup devices are measured using "compressed" capacity, which is bogus, at best. Fraud might be a better word. Usually, on a big drive, you have either huge databases (which typically compress well) or lots of graphics and sound files, which compress hardly at all. Selling a drive based on its compressed capacity is kinda like measuring the interior space of a car by including the roof rack and the potential trailer you could tow.
I'm a little concerned that backup technology isn't really keeping up with HD technology. I'm even more concerned that hardly anyone pays attention to backup technology around here -- I only saw one person ask how you would back the thing up, and the one reply was to another drive. Having one on-line backup is NOT a backup strategy!
It scares the heck out of me to see people buying 10+G drives without a thought to backup. Even if it is all programs, trying to rebuild a system after a data loss is very, very time consuming.
I'm also a little ticked off over the quality of backup devices. I've replaced far more tape drives in my client's servers than I have hard disks. Really pathetic. A friend has assured me DLTs are much better than the DATs I normally recommend, and that may be true, but $3000 for a drive and $100 per media, well, that's a few DAT drives. I gave up on Travan drives on servers -- I've had astronomical failure rates on both drives and tapes, but curiously, they seem to do o.k. on Windows 9x workstations.
Nick.
Actually, I had a couple people tell me RAM prices were on their way up a few days BEFORE the earthquake.
... (I'll spare you the rest of this progression)
The earthquake increased the effect, but prices were already on their way up. I've seen this repeatedly -- something happens some place that has something to do with semiconductor production, and RAM prices shoot up before the pipeline has a chance to go dry, and hang there as long as possible. That's the way a market economy works. You charge what you can get.
Observation (getting further off topic): We always seem to need to add $100-$150 worth of RAM to our computers:
Back when 48K was common, we paid $150 or so to up it to 64k.
When 64k was common, we paid $120 or so to up it to 128k.
Now, we want to add another 128M to our systems, but I'll wait until the price drops back down to $150... 8)
Hard disk prices just seem to drop.
Nick.
Warning: The comments contained here are from a very new OpenBSD user... (Operation system in place for about a week. Got interested in it about three weeks ago).
No, OpenBSD's install is not as friendly as the new Linux installs. The disk partitioning is really exciting. Hint: Start with a totally empty disk. When you manage to get partitions where you want them, put a small DOS partition on and work around it. Another hint: If you don't grab a calculator or at least a piece of paper, you are either very good at memorizing numbers and doing math in your head or you just nuked that partition you were trying to work around. The first time I installed it on a Compaq (with the configuration partition), I wiped the disk completely clean. Not even the OpenBSD partition survived the install! I think it took me two or three additional attempts (i.e., reloads) before I got it right. Suddenly, it becomes easy, even logical 8). I'm *very* serious about practicing on a "spare" hard disk first.
I found the FTP install actually worked very nicely with OpenBSD, something I could NOT get to work with the otherwise more friendly FreeBSD.
I will dispute any claims that documentation is better on Linux than *BSD. The Linux documentation is certainly BIGGER, at least on mall store shelves, but hardly better. You see, Linux is a rapidly moving target. The details of every distribution change every release, and in fact, some of the products might even be perceived as completely different OSs by a non-X familiar person. While I have been playing with Linux for YEARS (key word: Playing. I'm not yet a Linux master), I've been pleased that OpenBSD seems to be more "familiar" quickly based on my Unix experience. *BSD is much more standard Unix than Linux is. A book on RedHat v5.2 will not do much to help you get RedHat 6.0 up and running.
This is good from a learning perspective. Unix books cover both OpenBSD and FreeBSD much better than they do Linux. On the other hand, I really miss 'bash'. HOWEVER, *I* will impliment bash, I will know how it got there, I will know how to remove it or change it or whatever. You loose this in the modern Linux distributions. It is all done for you.
Another problem with Linux documentation is the quality. Everyone sees it as a cash cow, everyone is writing a book. Most of them stink. Several years ago, I bought a book on Linux. Durring my first five minutes with the book, I spotted THREE blatent FACTUAL errors (Linus wrote Minix then improved it to Linux was one of them). A few are good. Most are obsolete.
As far as I can tell for multiple booting, to boot OpenBSD, set the partition active. Therefore, it will probably work with most boot management software.
Hardware support: Well, OpenBSD seems to support "standard" stuff. IDE, main-stream SCSI. Haven't played with the sound. It recognized every network card I threw at it (at least four totally different cards). Forget USB for the time being, and I wouldn't wait up.
A *BIG* comment: I know a lot of people (I do it myself once in a while) seem to think the magic of a new system is installing it. This is not true. Not at all. As long as you think installing it is the major hurdle, you haven't really learned the product. The magic is what you do ONCE it is installed. How you set it up for efficient use. How you back it up. How you recover it when you woof a hard disk (or a few strategic files). How you recover it when you woof or upgrade a main board. I clean up a lot of networks set up by people who didn't understand that installing the software is just the very, very tip of the iceberg.
Try it. On an empty hard disk at first. 8)
Nick.
"...unlike Windows NT where Compaq, Data General, Hewlett-Packard, IBM, and Unisys provide 99.9 percent system-level uptime guarantees for Windows NT-based servers."
Assuming it takes 10 minutes to identify a crashed server, power-cycle it, and wait for it to reboot (this time varies GREATLY different systems), 99.9% availabity would be one crash per week (I did the math). I'm *SURE* they are not counting "planned" down times (You know, you have to reboot now that you made that change), and I am pretty sure they aren't counting hardware or power failures or other things that are quite outside of MS's control.
I don't consider a crash per week acceptable. Sorry.
If I don't hit 99.9% of the other cars on the road when I'm driving, am I a good driver?
I have a client who had a Novell server go over 900 days without reboot (A two-day power failure finally took it down). I think I can say quite certainly _No_ NT server in history has done that. That is the way things are supposed to work. Linux seems a heck of a lot closer to my ideals than NT is.
I can't speak for DG, IBM, or Unisys, but I have seen Compaq's "Guaranteed four hour service" program in action at a couple clients. Never seen a client broke for less than 24 hours with that program, either. I don't trust Compaq's "guarantees" when it comes to servicing their own products, how can they guarantee someone else's software??
Nick.
In the 1950s through the 1970s, people bought cars based on features and looks, not based on reliability and quality. There were a few cars that were well built and lasted, but for the most part, they rotted on the lots, while their flashy but unreliable competition sold.
People used to line up to see the new models. People with money lined up to BUY the new models. Knowing full well it probably didn't start any better than the old one.
It wasn't that Detroit was incapable of BUYING a quality product. It was that the consumer was unwilling to buy the boring but solid product. It wasn't until the late 70s and early 80s that consumers suddenly demanded QUALITY over LOOKS from Detroit, and Detroit responded (in the auto industry, the problem was the speed at which the change in consumer attitude took place. You can't change the manufacturing criteria from flash to quality overnight..and no one was sure if this was a passing fad or a real trend)
The facts are, if the consumer demands something with their money, they will get it. Complaints don't mean a thing.
Another example: Airlines. People complain about late flights, they complain about lousy service, but they book the cheapest flight. Duh! Leaving on time costs money! Great service costs money! If the consumer buys the cheapest product, they can complain until they are blue in the face, nothing will change, at least not for the better.
It is simple economics. There is great demand for flashy software. There is little demand for quality software. While that is the case, that's what the consumer gets. The software industry better hope consumers don't change their mind on flash vs. quality overnight, like happened with the auto industry.
I've been supporting bad software for many years now. I'm starting to detect a change in attitudes towards business people (although, I am probably guilty of contaminating the sample!). I think this is good.
Consumer complaints don't enter into economic decisions. Dollars do.
Nick.
I have near zero experience with OpenBSD so far, but I did try a couple installations.
Yeah, the OpenBSD disk layout program is really bizzare, at least to those of us who have used MS-DOS systems too much. When reading through the FreeBSD documentation that came with the Walnut Creek package, they shed some light on the *BSD disk model -- it appears to have been developed absolutely independantly of the IBM PC's HD layout, and they have different ideas about how things are done. When BSD was ported to the PC platform, the disk partitioning model was ported, too. Knowing FDISK probably hurts when laying out an OpenBSD system.
As I understand it, you basicly create a IBM/MS style partition which reserves space for OpenBSD. THIS becomes the BSD "disk". Now, this "disk" is partitioned. Kinda like "Extended partitions", but totally different 8)
My first attempt to load OpenBSD was on a generic clone, empty IDE hard disk, 3Com network card, FTP install. Worked great. I then tried putting it on a machine which I wanted to have it run on (Compaq Deskpro XL 575). Had more problems, as there was the infamous Compaq system maintence partition on the drive already, and a small DOS partition. But, the install appeared to go well. Until reboot. At which point, the system refused to boot, and upon investigation, the disk was blank, except for the OpenBSD boot loader. No OpenBSD partition, no DOS partition, no Compaq system partition (and those are a pain to reload).
I went back to FreeBSD after that. I'll try OpenBSD again, esp. once I found the explaination of the BSD disk concepts in the Walnut Creek FreeBSD book, but yes, for newbies, it would be best to practice on a BLANK hard disk at first. And of course, your backup are up to date, aren't they??
As someone else said, FreeBSD has a much nicer disk preparation program.
Nick.
It *appears* it depends upon the service provider. Some of the service providers apparently actively LOOK for people running web/other Internet servers, while others are more indifferent. My service provider said it was against the rules, but upon asking, they told me they don't actively look, but they will investigate any curious or excessive activity.
My web server has been on line for about a year now, without any complaints from the cable company, other than the time some (*censored*) cracked my system and set up a cracking shop, using my machine as a base to crack others. Got a nasty note from the cable company which I couldn't figured out until I realized the odd things happening on my Linux box were really odd things, not just ignorance on my part!
I do wish to point out my web server is STRICTLY for experimentation and occasionally getting stuff to friends. I do think using Cable Internet Access to do commercial or high-volume stuff is really tacky. We get high bandwidth at a low cost because it is assumed we will be rarely using most of the bandwidth.
I'm with the @Home service, but a friend who is waiting for Cable Internet Access who would also be getting @Home through a different cable company, sounds like he would be getting a totally different set of terms of service than I have.
Best I can say is anyone who says "This is how it is" is probably wrong SOMEWHERE. The rules are different everywhere. The rules are still being written...
Here, they recently put a cap on the "upstream" link (128kbps supposedly), which I am fine with...means I don't have to worry quite so much in case someone finds something interesting on my server...
Nick.
Oh, yeah... There are a LOT of things about ADD (ADHD, whatever) that are in the "traditional" literature that is Just Plain Wrong.
A couple of my favorites (to some degree, this is my speculation. Were I degreed in the medical arts, it would be an official guess, but in my case, it is just my guess):
1) "You grow out of it". No, you learn to adapt. School is probably one of the most ADD hostile areas of anyone's life. Leave school, Wow! problem Gone or at least Greatly Diminished.
2) It doesn't affect women. Bull. Many years ago, it was supposedly 20 men to one woman (sorry..at the time, you out-grew it, too, so it was boys to girls). A few years ago, it was seven to one. Last number I heard was three to one. There's a trend here! My guess: Boys are diagnosed and treated, girls are stereotyped. Until fairly recently, girls were expected to grow up to be mothers and wives -- a girl who "didn't fit in" was allowed to fall through the cracks and fail, after all, she had the Mrs. safety net.
There is (possibly) some correlation between hair color and ADD. Guess what hair colors are more likely to have ADD? Redheads and Blondes! A blond boy is considered ADD, a blonde girl is a ditz.
I know a woman who showed such obvious ADD and hyperactivity it was unmistakable, but since she is a woman, it was attributed to hormones, her disasterous marriage, etc. Everything but the obvious. She had even seen some professionals which really should have jumped on it -- it was obvious.
When I was in school, after considerable testing I was labeled "Learning Disabled" with no more accurate diagnosis. This was in 1979, dyslexia was just starting to be recognized, but ADD was still a euphamism for "Hyperactive" and the diagnosis of hyperactivity was so CLEARLY wrong (I tell people if I were any less hyperactive, I'd be diagnosed as DEAD!). In high school, we basicly decided I was dyslexic. About four years ago (LONG after my very difficult but ultimately successful school (I lived) life was over), I finally realized that I was also ADD, something that we also realized ran in my family like a bull through a china shop.
Another observation: The lines between ADD and many other disorders are very fuzzy. If the diagnosis explains things, it is useful. It seems most "disorders" ("reorders"?) of the mind are diagnosed by symptom matching, rather than by isolating and diagnosing a root cause. The manifestations vary widely. Personally, I noticed an initially unexpected similarity between ADD and Tourette's syndrom: a difference in focus and intensity, but seems to me to be of similar root causes. Most people think I'm off my rocker when I say that (hey, I'm used to it...), but I've also been told that some research along those lines is starting to be done.
Ah, so much more to say...but I fear I've wandered more than far enough off topic...
Nick.