Last year, Compaq donated an Alpha box to the Debian project to help them port their software. See here for some details.
For this, Compaq should be congratulated for. I hope they help out free software more in the future.
Two minor nits on RMS's position.
on
RMS on APSL
·
· Score: 1
This is basic IP law. If you look at the source and work on something similar, you're bound by the license.
IANAL, but to the best of my knowledge, this is not at all correct.
Copyrights protect the copying of the "expression of an idea". If you take, say, a story of "boy meets girl", change all the characters names, change the setting, change all the words in a non-mechanical way, but still have the same basic plot, then you are fine. You can use the same "idea" as long as you don't express it the same way. If you have never seen the original, but by some miricle come up with exactly the same words, then you are also OK because you haven't "copied the expression of an idea". If there are only a few ways of expressing an idea, then it is common to have most of them copyrighted, but that is OK.
So, by just looking at some source code, you are not bound by its license, unless you copy the expression of the ideas in that code. Companies often want people who have never seen the original work so that there is no chance that they can possibly copy the original work. That makes life much easier in court, but it is not critical. Companies also have to make sure that people haven't signed non-disclosure or non-compete agreements, which is not a copyright issue.
Patents, on the other hand, protect an invention. It makes no difference if you have seen the invention before, or if you came up with the same idea completely on your own. If the invention is patented, then you can't use it with the patent holders OK. So, in this case also, you can look at the source all you want, it won't change things.
It is so tempting to claim victory at anything that could possibly be seen as furthing your cause. (Note how often politicians claim credit for just about anything good, or place blaim on their rivals for anything bad.) I think in this case, ESR and OSI are claiming victory for something that is not entirely A Good Thing.
Bruce Perens says that he is talking with the folks at Apple about these issues. As in the case of Troll Tech and QT, I think there is a good chance that Apple will come around to a truely free license. Bruce, once a very vocal critic of the QT license, is now a vocal supporter of them because they did respond.
It is important to keep the flames down and only talk about the technical problems with the license. Other issues with Apple, whether it is the fact that they are part owned by MS, or they once filed the much hated (in the OSS communuty) "look and feel" lawsuit, or the quality of their GUI, or whatever, should remain seperate issues. IF they fix their license then Apple has done A Good Thing. If they don't, then we should consider it to be "not free software", but no worse than most shareware and as such, it should not get flamed, it should just get ignored.
Now, improvements made by others only is defined as the files where his copyright is on the top. If someone makes a (say) 2 line mod to a single file, but does not include their (c) symbol in the code as well, then they don't have a legal leg to stand on. The original (c) holder may take that code, put a different license on it and distribute it commercially/proprietry if they wish. If someone wants to enforce that you (being the original code writer) don't redistribute their changes, you need to put your little (c) mark on it as well. If you don't there's no point in crying over spilt milk.
This is incorrect. Since almost(?) the entire world has now signed the Berne Convention, you now have an implicit copyright on everything you write. You have to do something to give up this right, such as to do the work as a "work for hire", assigning it to the public domain, etc.
So, if I change two lines of code around in some file, no one else can use that patch w/o my permission. Things quickly start to get funny here. If I don't spell out what copy rights I want to reserve and what I want to give away, then it is not clear what, if any rights you have to use my work. Much of the time you can argue that the author "implicitly" gave certain rights away. Such as if I post the patch to Usenet or the Web, with a comment that says "this patch should fix the multi-user race condition, let me know if it doesn't work", then I think that most courts would conclude that I have given up much of my copy rights. Similarly, by writing this article and submitting it to/., I have given implicit permission for it to appear on/. and to be transmitted to other people. I have also implicitly said that it is OK to create "diriviative works" by quoting my article in your follow up. What isn't clear is whether I have given implicit permission for it to be copied anywhere else.
Now if my patch is applied to a GPL'ed or similar OSS license, then it can be argued, and the courts would probably agree that I implicitly licensed my patch under the same license. This means that you can't take my patch and change the license. (The BSD license, however, would allow you to take my patch and make it proprietary.)
These messy areas where you have to argue that the author gave "implicit permission" to do something causes a lot of problems in the OSS community. The Debian folks are very strict about honoring the copyrights and licenses of other people. You won't find any pirated warez on their machines. However, they frequently run into packages that have lines like this:
(c) Copyright by Foo Bar, all rights reserved. This package is in the public domain.
This causes the Debian folks to pull their hair out. If it is in the Public Domain, then that means that it isn't copyrighted at all. If you say that "all rights are reserved", then that means that you can't copy/distributed/modify it at all. This is clearly a self contradicting "license." What the courts would do with such a "license" is anyones guess. As a result, Debian won't distribute such a package without the author changing their "license."
A good place to start for more information about copyrights is here and a related one one copyright myths here.
Finally, the rules for copyrights and for patents are very different. Patents prevent people from using an inventions, even if it was independantly discovered, copyrights apply only to the "expression of an idea." If you come up, independantly, with the same two line patch, then there is nothing that my copyright on my patch can do to prevent you from using yours.
BTW, Rob, if you want free discussions going on here, I challange you to keep this post instead of removing it Otherwise this message board is no better than a company sponsered board
I doubt that Rob removed your post. I doubt that Rob had anything to do with what happened to your post.
I am not going to come out of the AC (Thats my choice)
Being an AC is your choice, but it does have consequences such as not being able to automatically set your "viewing threshold" and see that your previous message had its score dropped to -1. This was most like done by a moderator other than Rob.
*Why* did a moderator drop your score? I don't know, but one guess is that it sounded somewhat belligerent. Had you not posted it as an AC, it would still have had a score of zero and would be visible by default.
I have a bone to grind with some of the posters here.. While saying that VA and Linux-hw use quality components.. they flame Dell for not using quality components. Do you all have evidance for that.
Kit Cosper didn't say anything about the quality of parts that Dell uses, he certainly didn't flame them.
Can anyone give me examples of where is it used in a large program?
[PLUG ALERT]
How about this for a large program that uses TCL/TK? A couple million lines of code, hundreds of users world wide, and the price? Well, let's just say that people don't mind buying 500MHz-PIII's with 256MB RAM, 24" monitors, etc. 'cause it really doesn't effect the total much.:->
P.S. Please don't/. it. It's where I play quake from.
I don't see anything wrong with commercial (!= proprietary) interests in open source software. Even RMS has held this position, and he had said this for at least 15 years, and repeats it often.
The next big jump for open source software is when coporations start developing OSS to "scratch their own itch." In the past, a lot of OSS came out of universities and goverment projects, where this kind of idea arises more naturally. In the future, I hope that lots of business look to a problem they have, see that there is some OSS that already does much of what they want to have done, and then extend it to do what they really need done. Of course, the are oblidged to release these modifications.
Once OSS reaches a "critical mass", I think that lots of businesses will be choosing this route of "extending the OSS that is out there." It has a lot of advantages over in-house development because you get a good start for free and you often get bug fixes and other enhancements from other people for free. It also has a lot of advantages over proprietary software because the business has control over their own destiny, and the cost of deploying gobs of workstations is much smaller.
Comercial software is not inherently evil.
Businesses are not inherently evil.
MicroSoft is not inherently evil. (It is irrelevant.)
Linux is seen by many people as a "light weight" server: good for small print spoolers, email systems, file servers, small web sites, small ISPs, low-performance routers, etc. They think that if you have a serious server, you need a serious OS.
There is more than a grain of truth to this. Linux doesn't scale to a 64-processor, multi-GB RAM, multi-TB disk systems. It is just that these same people don't realize how much you can do with a 4-processor, 1GB RAM, 200GB disk Linux box, and how many PC users would think that is a large server.
I wouldnt mind betting linux.org, redhat.com, suse.com are getting those sorts of figures off their boxes.
Are you sure that those would be anywhere near 1million hits/day? I mean,/. is "only" 600khd, and I can't imaging that they would be hit as much as/. is.
Sure, they get a lot of ftp traffic too, but ftp traffic doesn't count toward hits/day. (Besides, ftp traffic is "easy" compared to most web traffic.)
I still have a copy of Dell Unix on one of my machines and it is still bootable. I even bought some commerical apps for it. Dell Unix was clearly the "darling of the Internet" for PC UNIX's for about a year or so, before it imploded.
While Dell clearly didn't quite "get the Unix" market back then, they were one of the very few PC vendors that even tried to "get it." That, in it self, is reason to hold out a lot of hope for Dell.
Please note that it was a Republican in the article.
Yes, I did notice. I have found it most humorous that your comment is the first I've seen that has mentioned the political party, either here on/. or on ZDnet.
Imagine if it had been a *democrat* that has said that!
Personally, I think democrats are mostly "tax and spend", while republicans are "borrow and spend", the latter being slightly worse 'cause you have to pay interest and it costs future generations.
Last year, the politicians in Washington (on both sides of the isle) claim the federal goverment has a $69billion surplus. In fact, if you take out the "surplus" that is going into the Social Security "trust fund", there was really a $29billion deficit. If you take into account all the other borrowing that goverment did, there was a $113billion deficit.
Because it is based wrong data. Ted Lewis wrote: "Unix has more than 10 million lines of code, while Linux has only 1.5 million." Now, Linus has just recently said: "Right now its in the amount of 15,000,000 now" (lines of kernel code) (see that transcript of IRC with Linus, it was here yesterday)
Linus didn't type that number, someone else was transcribing his phone conversation to IRC. That latter person misstyped. The source to the Linux kernel is about 1.5MLOC.
However, the article is still wrong because the error test wasn't comparing kernels, but the utilities, and the GNU utilities are not dramatically different in size than the commerical utilities, in terms of KLOC.
I have the impression that, at one point, when you bought OS/360 you got source, and that many businesses did, in fact, use that to add features or fix bugs - and that some of those bug fixes and/or features eventually got into mainstream OS/360.
Yes, this happened in the late '70s. For example, IBM's product ASP (Automatic Spooler Program" was modified by the University of Houston and became "HASP", which was later rolled back into an IBM product called JES. Or at least that's what I recall. Some of the details are kind of fuzzy after all these years.
I also have the impression that, at one point, IBM decided to go with "object code only", and that some customers didn't like this; if my impression is correct, I'm curious what what the history of that was, what the motivation for OCO was, and whether OCO turned out to be a good or bad idea overall.
Again, it has been 20 years, so I don't remember clearly, but yes, the customer base screamed like hell about going object code only. That is when they started to release the code on microfiche.
In so many ways, the early-mid mainframe era had a lot of similarities to what the current/future OSS movement hopes to become. Many large businesses and Universities payed people to develop code that was useful for them. This code got released to the world, and modifications can back to the original users. Since the original people could justify the work based off of their immediate benefits, the added features/bug fixes that came back were just a bonus.
One of the key differences is that with the GPL, no one can take it back to being a closed system.
I think what really caused IBM to kill the freely available source code was that it allowed "Plug compatible" companies (think "PC Clone manufacture") to compete too easily with them. Competition is such a horrible thing, ya know.
The mini-computer industry had a lot of source code floating around too, but I don't think they had the resources to actively send tapes and such around the country/world like the mainframe people did, nor did as many people get to go to conferences. With the Internet, this cost has obviously dropped to almost zero.
The one point I am personally unsure of is how IBM can fit into the Open Source movement. That is, we pulled down some 14 BILLION dollars in software sales alone last year. How can we maintain those numbers with "free" software solutions?
The problem with this additude is that the market can just pass you by. It is just this additude that cost IBM dearly in the late '80s and early '90s and why DEC, at one time the second largest computer comany, is now owned by a PC manufacture.
There are many reasons why Open Source Software is taking off, but one of the clear ones is lower cost of ownership. Not just because OSS is "free" (as in beer), but more importantly because it is "free" (as in speech). A business can control its own destiny with OSS software. If they need a feature, they can decide if it is worth the cost to add it, they don't have to beg some other company to do it for them. If an obscure bug bites them hard and few other people, they can choose to fix it, instead of pleading with some other company who might not even be able to reproduce it. Instead of writing a custom system from scratch, or paying some other company to, in effect, write a custom version of their software, they can leverage an OSS system.
Expect the total cost per machine to drop because of OSS. It doesn't matter if IBM is involved or not, they will get hit by this lower cost either way. If they don't get on the front wave of the OSS in business movement, they are going to have a hard time catching up.
I hope IBM doesn't make the same mistake that they made in the '80s and try and preserve there current cash-cows only to see their cash-cow fade away.
There is a chance that even though the total software dollars per machine may drop, the number of machines sold will increase enough to offset this. Of course, this point is irrelevant. You can't change the effect that OSS will have anyway.
Our customers are happy to buy binaries-only at this point so there is no need to change.
Your customers might not be complaining too loadly now, but I bet everyone of them would love to have access to, and possibly modify the source to every peice of software they run.
You don't have to want to "improve" the software for you to really need the source. Every time HP-UX or NT gives me some weird error that isn't well decribed in the doc, I'm in a tough place. When BSDI (which we have source to) gives us problems, we can read the source or even put debug statements in to try and figure out what the real problem is. Last I knew (mid '80s), IBM's mainframe OS's came with source on microfiche for similar reasons.
Cool. I submitted this article 2 days ago, with the comment that "Workstations are where the apps are."
I really think that shipping workstations is much more important than shipping servers. Once people are comfortable with using an operating system, it is easy for them to strip it down and put it in a corner as a server. Servers, on the other hand, are often viewed as these mysterious boxes that only the sysadmin knows how to run, so why would I want that on my desk?
The desktop is where most of the applications need to run. That means that most developers must learn how to program for the desktop OS. That means that if they need to develop for a server, they will have a strong bias toward using the same OS that they already know. It also means that whey they develope stuff for fun, they will develope for the desktop OS.
This is critical folks. If Linux (or any version of Unix) fails to capture the desktop, it will ultimately fail to capture anything.
Ok, my subject line is somewhat of a troll, but not entirely. When Linux was new and the *BSD systems were also just becoming freely available, one of the things that caused Linux to succeed was the large number of PC programmers who flocked to Linux.
The early *BSD systems didn't even recognize the standard DOS partition scheme. Linux, on the other hand, quickly supported the FAT filesystem and worked on DOSEMU, and with UMSDOS, you could even run Linux off of a MS-DOS filesystem! The *BSD developers were, in a large part, old-time UNIX gurus who were more comfortable with a VAX or a Sun workstation than a PC. Linux users came from the Minix/MS-DOS/Windows crowd and were sneered at as not known how a "true Unix" should work. This additude still exists among many *BSD people.
I'm an old-time Unix user, but I don't fear an influx of Windows programmers, any more than I feared the "hords of unwashed masses" that have "invaded" the Internet over the last 15 years that I have been on it. Yes, the Internet has changes much because of them, and in many ways for the worse, but if there were still only a few million people on the entire Internet, there would be no Linux, no WWW, no/., no CNN, no place to buy my christmass presents, no ebay, etc. The world would be locked into "Prodigy classic" or old style Compuserve.
1. Create a large 10MB test file on an IDE partition.
2. Do lots of checksums using sum or md5sum.
Q: Do you get the same checksum each time?
I once had a problem like this where I would get a corrupt byte from the disk about once very 50-100MB of data read. It happened on two different disks, I tried three different IDE controllers, I swapped RAM, I ran RAM tests, I made sure my bus speeds were within spec. One byte every 50-100MB might not be very high, but it was enough to crash my system once or twice a week.
It turned out that I needed to set my RAM speed in the BIOS to the slowest setting, and everything worked. The RAM test programs didn't detect anything, I think because the errors only occured if I was doing heavy disk access at the same time I was doing very CPU intensive things.
Moral of the story: PC hardware is crap and Unix tends to push the hardware much futher than other systems.
Set your BIOS settings down to the very slowest settings and see if the problem goes away. Try swapping components, and try a *good* memory tester (Linux has one called mem86 or something).
also the reservered blocks percentage (man mke2fs). On large partitions you can waste alot of space if you keep the default 5% reservered blocks percentage.
Reducing the amount of reserved space may save you some space, but it can cost you a *lot* of time. Having 5% reserved space will mean that there will almost always be a "free block" within about 20 blocks from the end of the file.
Unless you want your expensive raid system to spend lots of time seeking, you should keep the 5% min free value, or even increase it to 10% like the BSD folks use. You certainly don't want to constantly run the filesystem close to being full most of the time.
In case you hadn't noticed, we're expecting a fairly hefty budget surplus this year
Oh, well, yes, the US has had a "budget surplus" this year, but that is only because of the extra money that is going into the "Social Security Trust Fund". If you took that off budget, which legally it is supposed to be, we would have and a "hefty budget deficit" this year. ("hefty" being defined as being about the same size as the surplus that you call "hefty").
The US government went about $113 billion futher in debt last fiscal year, when everything is totalled up. Check out the Concord Coalition web site for more info. They are a group made up of both democratic and republicans that know all too well how washington works.
Ok, so your competition now has the exact same software as you, and you are out many thousands of dollars since you paid for the development of the software. Now, assuming you and your competitor are equal in most other areas, which company is going to go out of business first, you or your competitor? Where is the incentive for anyone to pay for new software development projects?
Right now, you have somewhat of a point, although I think a business can justify starting an Open Source package from scratch. See this web page from the Open Source folks.
An easier case to consider is this: Assume that somehow, a large number of Open Source packages have been developed that do almost all of the common things that a business might use. Now you are a business, and you need software to do something. You have several choices:
You can hire a bunch of programmers to develop a package that does exactly what you want to do. This will cost a lot, and have continual update costs
You can buy a "generic" commerical package that does much of what you want to do, but not exactly. This won't cost too much, but you will probably have to continually buy updates as the company stops selling the old software and only sells you new, but slightly incompatible software. You may also have to change around the way your business works in order to mesh with the comerical software, which might make you much less efficent.
You can take the Open Source software that is out there and modify it to do exactly what you want. Of course, you must release these modifications, but your up-front costs are still going to be much lower than building your own, and your long term costs may be much lower than buying a comerical package. Plus, you can make it do what your business wants to do, so you can operate more efficently.
I think that once the Open Source Software reaches a critical mass, you are going to see both comercial companys that create "generic" software, and in-house "closed" development fade away. Really specialized software will probably surive in a closed form for much longer, and I think there will be "software" companies that special in modifying OSS to meet custom needs.
I can see the argument that the government should fund the OSS in order to reach this critical mass and thus provide both businesses and the public with cheaper, higher quality software. In particular, I think the government should strongly consider using OSS for all its internal development. Why should I pay taxes to the gov so it can buy closed source custom software? Why should I pay taxes so that the gov can send lots of money to MicroSoft instead of using Linux or *BSD?
Slashdot Mirror
Watch the scores. It appears that the vast majority of moderators up scores, rather than kill them.
For this, Compaq should be congratulated for. I hope they help out free software more in the future.
Copyrights protect the copying of the "expression of an idea". If you take, say, a story of "boy meets girl", change all the characters names, change the setting, change all the words in a non-mechanical way, but still have the same basic plot, then you are fine. You can use the same "idea" as long as you don't express it the same way. If you have never seen the original, but by some miricle come up with exactly the same words, then you are also OK because you haven't "copied the expression of an idea". If there are only a few ways of expressing an idea, then it is common to have most of them copyrighted, but that is OK.
So, by just looking at some source code, you are not bound by its license, unless you copy the expression of the ideas in that code. Companies often want people who have never seen the original work so that there is no chance that they can possibly copy the original work. That makes life much easier in court, but it is not critical. Companies also have to make sure that people haven't signed non-disclosure or non-compete agreements, which is not a copyright issue.
Patents, on the other hand, protect an invention. It makes no difference if you have seen the invention before, or if you came up with the same idea completely on your own. If the invention is patented, then you can't use it with the patent holders OK. So, in this case also, you can look at the source all you want, it won't change things.
Bruce Perens says that he is talking with the folks at Apple about these issues. As in the case of Troll Tech and QT, I think there is a good chance that Apple will come around to a truely free license. Bruce, once a very vocal critic of the QT license, is now a vocal supporter of them because they did respond.
It is important to keep the flames down and only talk about the technical problems with the license. Other issues with Apple, whether it is the fact that they are part owned by MS, or they once filed the much hated (in the OSS communuty) "look and feel" lawsuit, or the quality of their GUI, or whatever, should remain seperate issues. IF they fix their license then Apple has done A Good Thing. If they don't, then we should consider it to be "not free software", but no worse than most shareware and as such, it should not get flamed, it should just get ignored.
This is incorrect. Since almost(?) the entire world has now signed the Berne Convention, you now have an implicit copyright on everything you write. You have to do something to give up this right, such as to do the work as a "work for hire", assigning it to the public domain, etc.
So, if I change two lines of code around in some file, no one else can use that patch w/o my permission. Things quickly start to get funny here. If I don't spell out what copy rights I want to reserve and what I want to give away, then it is not clear what, if any rights you have to use my work. Much of the time you can argue that the author "implicitly" gave certain rights away. Such as if I post the patch to Usenet or the Web, with a comment that says "this patch should fix the multi-user race condition, let me know if it doesn't work", then I think that most courts would conclude that I have given up much of my copy rights. Similarly, by writing this article and submitting it to /., I have given implicit permission for it to appear on /. and to be transmitted to other people. I have also implicitly said that it is OK to create "diriviative works" by quoting my article in your follow up. What isn't clear is whether I have given implicit permission for it to be copied anywhere else.
Now if my patch is applied to a GPL'ed or similar OSS license, then it can be argued, and the courts would probably agree that I implicitly licensed my patch under the same license. This means that you can't take my patch and change the license. (The BSD license, however, would allow you to take my patch and make it proprietary.)
Please note that the string "(c)" has no legal meaning. At one time, to copyright something, you have to use a c with a circle around it "©", or spell out "Copyright", but now with the Berne convention, neither of these are required. So saying "(c) Wayne Schlitt, all rights reserved" has never done anything for me.
These messy areas where you have to argue that the author gave "implicit permission" to do something causes a lot of problems in the OSS community. The Debian folks are very strict about honoring the copyrights and licenses of other people. You won't find any pirated warez on their machines. However, they frequently run into packages that have lines like this:
This causes the Debian folks to pull their hair out. If it is in the Public Domain, then that means that it isn't copyrighted at all. If you say that "all rights are reserved", then that means that you can't copy/distributed/modify it at all. This is clearly a self contradicting "license." What the courts would do with such a "license" is anyones guess. As a result, Debian won't distribute such a package without the author changing their "license."A good place to start for more information about copyrights is here and a related one one copyright myths here.
Finally, the rules for copyrights and for patents are very different. Patents prevent people from using an inventions, even if it was independantly discovered, copyrights apply only to the "expression of an idea." If you come up, independantly, with the same two line patch, then there is nothing that my copyright on my patch can do to prevent you from using yours.
BTW, Rob, if you want free discussions going on here, I challange you to keep this post instead of removing it Otherwise this message board is no better than a company sponsered board
I doubt that Rob removed your post. I doubt that Rob had anything to do with what happened to your post.
I am not going to come out of the AC (Thats my choice)
Being an AC is your choice, but it does have consequences such as not being able to automatically set your "viewing threshold" and see that your previous message had its score dropped to -1. This was most like done by a moderator other than Rob.
*Why* did a moderator drop your score? I don't know, but one guess is that it sounded somewhat belligerent. Had you not posted it as an AC, it would still have had a score of zero and would be visible by default.
I have a bone to grind with some of the posters here.. While saying that VA and Linux-hw use quality components.. they flame Dell for not using quality components. Do you all have evidance for that.
Kit Cosper didn't say anything about the quality of parts that Dell uses, he certainly didn't flame them.
[PLUG ALERT]
How about this for a large program that uses TCL/TK? A couple million lines of code, hundreds of users world wide, and the price? Well, let's just say that people don't mind buying 500MHz-PIII's with 256MB RAM, 24" monitors, etc. 'cause it really doesn't effect the total much. :->
P.S. Please don't /. it. It's where I play quake from.
The next big jump for open source software is when coporations start developing OSS to "scratch their own itch." In the past, a lot of OSS came out of universities and goverment projects, where this kind of idea arises more naturally. In the future, I hope that lots of business look to a problem they have, see that there is some OSS that already does much of what they want to have done, and then extend it to do what they really need done. Of course, the are oblidged to release these modifications.
Once OSS reaches a "critical mass", I think that lots of businesses will be choosing this route of "extending the OSS that is out there." It has a lot of advantages over in-house development because you get a good start for free and you often get bug fixes and other enhancements from other people for free. It also has a lot of advantages over proprietary software because the business has control over their own destiny, and the cost of deploying gobs of workstations is much smaller.
Comercial software is not inherently evil.
Businesses are not inherently evil.
MicroSoft is not inherently evil. (It is irrelevant.)
The Governement is not inherently evil.
There is more than a grain of truth to this. Linux doesn't scale to a 64-processor, multi-GB RAM, multi-TB disk systems. It is just that these same people don't realize how much you can do with a 4-processor, 1GB RAM, 200GB disk Linux box, and how many PC users would think that is a large server.
Are you sure that those would be anywhere near 1million hits/day? I mean, /. is "only" 600khd, and I can't imaging that they would be hit as much as /. is.
Sure, they get a lot of ftp traffic too, but ftp traffic doesn't count toward hits/day. (Besides, ftp traffic is "easy" compared to most web traffic.)
While Dell clearly didn't quite "get the Unix" market back then, they were one of the very few PC vendors that even tried to "get it." That, in it self, is reason to hold out a lot of hope for Dell.
This was already posted to /. over a week ago. Look here for the comments from last time...
Yes, I did notice. I have found it most humorous that your comment is the first I've seen that has mentioned the political party, either here on /. or on ZDnet.
Imagine if it had been a *democrat* that has said that!
Personally, I think democrats are mostly "tax and spend", while republicans are "borrow and spend", the latter being slightly worse 'cause you have to pay interest and it costs future generations.
Last year, the politicians in Washington (on both sides of the isle) claim the federal goverment has a $69billion surplus. In fact, if you take out the "surplus" that is going into the Social Security "trust fund", there was really a $29billion deficit. If you take into account all the other borrowing that goverment did, there was a $113billion deficit.
See The Concord Coalition for more details.
I once knew PI to 51 digits, but that was 20 years ago, and now I can only reliably remember 23...
Somewhere, I have a link to a story about someone who claims to have memorized 42,000 digits.
UNIX: Any of a collection of similar operating systems that you can run the programming language 'perl' on.
Yesterday, while looking at dejanews.com, I found out that someone is still using it as a .signature. Somehow, I find this project to be most amusing.
Linus didn't type that number, someone else was transcribing his phone conversation to IRC. That latter person misstyped. The source to the Linux kernel is about 1.5MLOC.
However, the article is still wrong because the error test wasn't comparing kernels, but the utilities, and the GNU utilities are not dramatically different in size than the commerical utilities, in terms of KLOC.
Yes, this happened in the late '70s. For example, IBM's product ASP (Automatic Spooler Program" was modified by the University of Houston and became "HASP", which was later rolled back into an IBM product called JES. Or at least that's what I recall. Some of the details are kind of fuzzy after all these years.
I also have the impression that, at one point, IBM decided to go with "object code only", and that some customers didn't like this; if my impression is correct, I'm curious what what the history of that was, what the motivation for OCO was, and whether OCO turned out to be a good or bad idea overall.
Again, it has been 20 years, so I don't remember clearly, but yes, the customer base screamed like hell about going object code only. That is when they started to release the code on microfiche.
In so many ways, the early-mid mainframe era had a lot of similarities to what the current/future OSS movement hopes to become. Many large businesses and Universities payed people to develop code that was useful for them. This code got released to the world, and modifications can back to the original users. Since the original people could justify the work based off of their immediate benefits, the added features/bug fixes that came back were just a bonus.
One of the key differences is that with the GPL, no one can take it back to being a closed system.
I think what really caused IBM to kill the freely available source code was that it allowed "Plug compatible" companies (think "PC Clone manufacture") to compete too easily with them. Competition is such a horrible thing, ya know.
The mini-computer industry had a lot of source code floating around too, but I don't think they had the resources to actively send tapes and such around the country/world like the mainframe people did, nor did as many people get to go to conferences. With the Internet, this cost has obviously dropped to almost zero.
The problem with this additude is that the market can just pass you by. It is just this additude that cost IBM dearly in the late '80s and early '90s and why DEC, at one time the second largest computer comany, is now owned by a PC manufacture.
There are many reasons why Open Source Software is taking off, but one of the clear ones is lower cost of ownership. Not just because OSS is "free" (as in beer), but more importantly because it is "free" (as in speech). A business can control its own destiny with OSS software. If they need a feature, they can decide if it is worth the cost to add it, they don't have to beg some other company to do it for them. If an obscure bug bites them hard and few other people, they can choose to fix it, instead of pleading with some other company who might not even be able to reproduce it. Instead of writing a custom system from scratch, or paying some other company to, in effect, write a custom version of their software, they can leverage an OSS system.
Expect the total cost per machine to drop because of OSS. It doesn't matter if IBM is involved or not, they will get hit by this lower cost either way. If they don't get on the front wave of the OSS in business movement, they are going to have a hard time catching up.
I hope IBM doesn't make the same mistake that they made in the '80s and try and preserve there current cash-cows only to see their cash-cow fade away.
There is a chance that even though the total software dollars per machine may drop, the number of machines sold will increase enough to offset this. Of course, this point is irrelevant. You can't change the effect that OSS will have anyway.
Our customers are happy to buy binaries-only at this point so there is no need to change.
Your customers might not be complaining too loadly now, but I bet everyone of them would love to have access to, and possibly modify the source to every peice of software they run.
You don't have to want to "improve" the software for you to really need the source. Every time HP-UX or NT gives me some weird error that isn't well decribed in the doc, I'm in a tough place. When BSDI (which we have source to) gives us problems, we can read the source or even put debug statements in to try and figure out what the real problem is. Last I knew (mid '80s), IBM's mainframe OS's came with source on microfiche for similar reasons.
I really think that shipping workstations is much more important than shipping servers. Once people are comfortable with using an operating system, it is easy for them to strip it down and put it in a corner as a server. Servers, on the other hand, are often viewed as these mysterious boxes that only the sysadmin knows how to run, so why would I want that on my desk?
The desktop is where most of the applications need to run. That means that most developers must learn how to program for the desktop OS. That means that if they need to develop for a server, they will have a strong bias toward using the same OS that they already know. It also means that whey they develope stuff for fun, they will develope for the desktop OS.
This is critical folks. If Linux (or any version of Unix) fails to capture the desktop, it will ultimately fail to capture anything.
The early *BSD systems didn't even recognize the standard DOS partition scheme. Linux, on the other hand, quickly supported the FAT filesystem and worked on DOSEMU, and with UMSDOS, you could even run Linux off of a MS-DOS filesystem! The *BSD developers were, in a large part, old-time UNIX gurus who were more comfortable with a VAX or a Sun workstation than a PC. Linux users came from the Minix/MS-DOS/Windows crowd and were sneered at as not known how a "true Unix" should work. This additude still exists among many *BSD people.
I'm an old-time Unix user, but I don't fear an influx of Windows programmers, any more than I feared the "hords of unwashed masses" that have "invaded" the Internet over the last 15 years that I have been on it. Yes, the Internet has changes much because of them, and in many ways for the worse, but if there were still only a few million people on the entire Internet, there would be no Linux, no WWW, no
2. Do lots of checksums using sum or md5sum.
Q: Do you get the same checksum each time?
I once had a problem like this where I would get a corrupt byte from the disk about once very 50-100MB of data read. It happened on two different disks, I tried three different IDE controllers, I swapped RAM, I ran RAM tests, I made sure my bus speeds were within spec. One byte every 50-100MB might not be very high, but it was enough to crash my system once or twice a week.
It turned out that I needed to set my RAM speed in the BIOS to the slowest setting, and everything worked. The RAM test programs didn't detect anything, I think because the errors only occured if I was doing heavy disk access at the same time I was doing very CPU intensive things.
Moral of the story: PC hardware is crap and Unix tends to push the hardware much futher than other systems.
Set your BIOS settings down to the very slowest settings and see if the problem goes away. Try swapping components, and try a *good* memory tester (Linux has one called mem86 or something).
Good luck
Reducing the amount of reserved space may save you some space, but it can cost you a *lot* of time. Having 5% reserved space will mean that there will almost always be a "free block" within about 20 blocks from the end of the file.
Unless you want your expensive raid system to spend lots of time seeking, you should keep the 5% min free value, or even increase it to 10% like the BSD folks use. You certainly don't want to constantly run the filesystem close to being full most of the time.
Oh, well, yes, the US has had a "budget surplus" this year, but that is only because of the extra money that is going into the "Social Security Trust Fund". If you took that off budget, which legally it is supposed to be, we would have and a "hefty budget deficit" this year. ("hefty" being defined as being about the same size as the surplus that you call "hefty").
The US government went about $113 billion futher in debt last fiscal year, when everything is totalled up. Check out the Concord Coalition web site for more info. They are a group made up of both democratic and republicans that know all too well how washington works.
Right now, you have somewhat of a point, although I think a business can justify starting an Open Source package from scratch. See this web page from the Open Source folks.
An easier case to consider is this: Assume that somehow, a large number of Open Source packages have been developed that do almost all of the common things that a business might use. Now you are a business, and you need software to do something. You have several choices:
You can hire a bunch of programmers to develop a package that does exactly what you want to do. This will cost a lot, and have continual update costs
You can buy a "generic" commerical package that does much of what you want to do, but not exactly. This won't cost too much, but you will probably have to continually buy updates as the company stops selling the old software and only sells you new, but slightly incompatible software. You may also have to change around the way your business works in order to mesh with the comerical software, which might make you much less efficent.
You can take the Open Source software that is out there and modify it to do exactly what you want. Of course, you must release these modifications, but your up-front costs are still going to be much lower than building your own, and your long term costs may be much lower than buying a comerical package. Plus, you can make it do what your business wants to do, so you can operate more efficently.
I think that once the Open Source Software reaches a critical mass, you are going to see both comercial companys that create "generic" software, and in-house "closed" development fade away. Really specialized software will probably surive in a closed form for much longer, and I think there will be "software" companies that special in modifying OSS to meet custom needs.
I can see the argument that the government should fund the OSS in order to reach this critical mass and thus provide both businesses and the public with cheaper, higher quality software. In particular, I think the government should strongly consider using OSS for all its internal development. Why should I pay taxes to the gov so it can buy closed source custom software? Why should I pay taxes so that the gov can send lots of money to MicroSoft instead of using Linux or *BSD?
Wow. What a troll. I'm impressed. almost everthing is wrong. Here is a zinger: in a communist system, you don't pay taxes! Whee!