You are a pro, of course. If you really prefer Linux for technical reasons (the license being one of them), you obviously would dump it without looking back once you realize that there is an alternative that fits your needs even better. No irrational feelings about loyality or having to defend your system of choice against its competiton just because it is your system of choice, no FUD, no flames, no bumper stickers.
The pros as he describes them don't just "not care" about their platform, they choose them depending on what fits the task best. Same thing.
And, indeed, the only sane way of thinking about software IMHO, even if lots of people (not at all limited to Linux, there are a lot of irrational windows supporters as well, for example, or BSD, BeOS, Solaris, AIX, you name it - and let's not go into text editors, unix desktops or programming languages) don't seem to be capable of it.
That said, I think the article is silly anyway.
Let's just say his goal obviously was to generate lots of page impressions rather than to enlarge human knowledge as a whole.
Oh and the read-only sysctl problem for FreeBSD that he mentions was probably due to securelevel's being on (meaning you can't modify kernel variables).
Nope, kern.maxproc is really read-only in a running system even in securelevel -1. You have to set it in/boot/loader.conf (which doesn't seem to be prominently documented anywhere, so not finding it is nothing to blame fefe for).
What a good, original, idea! Tell you what, why don't we have, say, 13 of them (13's a nice number), and put them in different places across the world?
That would be a great idea indeed. Like, why don't we put some more than just 3 of them in places that can not only be reached from Europe, Asia, Africa and Australia via one of the few connections to the Americas?
I agree that the situation sucks (or at least sucked until very recently), but it's hardly an oversight of the FreeBSD people. In fact, there has been a Java-on-FreeBSD project for some time. They have managed to port the Sun VM to FreeBSD a long time ago, but weren't permitted to distribute it - i.e. you as a mere user could use it, but you'd have to install the linux version first, then get the source of the linux version under NDA from Sun, then get a patchset, then compile it; distributing binary packages would have been illegal.
The good news is that now there is an official, redistributable, native Sun Java VM port, at least for FreeBSD 4 (of course, you have to download it from the FreeBSD site, not Sun's, FreeBSD isn't part of the "A" in "WORA"). It has finally passed Sun's test suite, which it didn't earlier mostly because nobody could pay Sun enough money to run it. It was too late for 5.1, and there was still a minor issue IIRC, but I'd expect it to be in 5.2.
Bottom line:
Java has worked fine on FreeBSD for ages
Installing it has been a pain, isn't anymore
Using proprietary technology tightly controlled by company with stupid, but influencial marketing department to develop free software is not a good idea.
Unless you have serious issues with disk space or use your own custom-designed processor for which only you have a compiler, what problems do you think having these tools around will cause? If someone roots you, I doubt the first thing they are going to do is to fire up emacs, write a remote shell, compile and install it. They can just do that at home, and upload the binaries. (They could also upload a compiler, if they had a reason to)
This is the same as not having a text editor, so that an intruder cannot change config files. It doesn't work. When they could use an editor or compiler on your firewall, you've already lost.
At least at the time his biography (which I won't recommend reading, the Linux kernel in a hex editor is more enlightening and entertaining literature) he drove a blue BMW Z3. And given that trademarks are only valid for a certain kind of possibly-competing products, I doubt that he could sue the owner, even if he would want to.
We really need to link ICANN more effectively to the
world, maybe each state or province in each country can elect 1 ICANN rep.
There are already elections for the ICANN board, but they have proven to be completely useless even if you get "our" people in. You might want to google for some of the statements of Andy Muller-Maghun (the first "u" is supposed to be an umlaut, but/. seems to fuck it up), a german hacker, spokesman of the Chaos Computer Club and former elected member of the ICANN board, until he resigned in frustration.
There are other DNS servers, some of them even work. And one could post here just as well if BIND would have been implemented correctly, preferrably from the very beginning, or at least with the BIND 9 "complete rewrite" which doesn't seem to have helped much so far.
Then again, this issue has nothing to do whatsoever with BINDs rotten codebase. Normally you can be pretty sure that they get the DNS part right and only fuck up in the C implementation part, this time it's the other way around.
And you generally don't have to pay for GPL'd software. The only thing the developer asks for in return is that you release any changes under the same license.
This seems to be what the author considers the problem: If you use proprietary code without a license and are caught, you have the option of either stopping to use it or to start paying license fees and go on. With GPLed software, you can stop using the code, or open your own one, which is less attractive.
What makes the GPL more dangerous for companies is that the risk of trying to get away with copyright violations is higher. Companies that honor the law have nothing to fear, but they are obviously not Forbes' target group.
Face it, you people have thousands of media items: VHS tapes, CDs, DVDs, miniDVs
No, I haven't. Facing the same problem, a quick analysis showed that everything I archived could be faster be restored by downloading it from the net than by finding the relevant backup medium, with the exception of my DSL password and my private key (I hope). Hence I have only one business-card sized CD that I carry with me.
It takes a strong character however to admit that even most of your toy coding projects can easily be "restored" by forgetting about them and using one of the fifty equivalent projects on sourceforge. And no, this is slashdot, I'm not supposed to have a private life.
And it has improved! I don't even have to recompile my kernel any more to build the FreeBSD port like I had for 1.0, because compiling an Office suite obviously requires some more radical settings than the vanilla kernel, perfectly suitable for huge database or file servers running stable under high load while staying responsive for interactive use, can provide.
They could at least get rid of the useless splash screen that makes/all/ of your virtual desktops unusable while that piece of "enterprise quality" code takes as long to boot up as it took Don Knuth to implement TeX.
The FreeBSD port isn't that impressive either, currently. And it's very existence is due solely to the great volunteer work of the porters, there doesn't seem to be much effort from the OOo team itself to make their software more easily portable.
Just one more example of Suns idea of portable software, Java is another one: Works on any platform as long as it's Windows, Solaris or Red Hat.
Most IDEs include a compiler or interpreter of some kind, that will parse your code and barf on type errors. If yours doesn't because the language is too weakly typed to provide meaningful error messages, the right thing to do is using one that has a useful type system, like any language in common use except assembly, C, C++, Basic, Perl or PHP.
Encoding types in variable names helps exactly as much as calling a C string "name_that_will_never_be_longer_than_128_chars" and omitting bounds checking.
Java wasn't even the first C-like language that was "object-oriented from the ground up."
No surprise, given that Java isn't "object-oriented from the ground up". It just forces anybody not employed by Sun Microsystems to not use any abstraction mechanism but classes and objects, just like it forbids operator overloading for anybody not employed by Sun Microsystems. Once you get a job at Suns design team, Java is a much more flexible language.
Maybe instead of asking your toolset to design for you - security is a design process, not a patch process, which is why it works for qmail but not for sendmail - you should just learn to write secure code.
Can you name any project of significant size implemented in C and not written by DJB that never suffered from buffer overflows, double frees or format string errors?
Sorry, but the argument that it's the stupid programmers fault and not the language fails if there is only one non-stupid programmer. Dan Bernstein is just too damn slow to write all the software the whole world needs himself. I'll go for a language where people only fuck up in the ways not due to a too low level of abstraction, like the bugs that occur in high-level-language programs and those of DJB (although he won't admit them usually).
Re:I'm not sure I want to use Windows XP that long
on
Longhorn in 2006
·
· Score: 1
So, why did you install the viruses if you didn't want them?
Excuse me?! FTP is an absolutely braindead protocol from todays point of view - even if you find an interoperable solution to get rid of the plain-text passwords, the multiple-tcp-connections design is a fucking pain for people who have to configure packet filters to make it work. The most popular FTP servers, like WU-FTP or ProFTPD are about as secure a code base as BIND or sendmail. If it were for me, FTP should take its friend telnet and get the fuck off the net, joining finger and rlogin in the nirvana of net services.
SFTP is a different matter however, but it's less an extension of FTP as an add-on to SSH to implement similar functionality in a completly different way. Not bad as a protocol, but it suffers from the lack of a robust SSH implementation.
It is sorta like pushing a train. If it is standing still then it will be hell to get moving. But once it does stopping it will be even harder
Um, no, it won't. It will even eventually stop on its own. Bad analogy.
Don't forget that not so long ago is you suggested using MS software in an office you would have been kicked out by both IT AND management.
"Not so long ago?" You mean when their only product was basic for the Altair? I doubt there was much time where the general idea of using software in an office was OK, but MS software wasn't. Remember that they did not only MSDOS, but also stuff almost completely forgotten today, from COBOL compilers for Apples to Unix clones for early PCs. They were relevant for home computers scince there were home computers, and for personal computers scince there were personal computers.
What is surprising is just how insanely dominant they became, and that they suddendly managed to get a foot in the server arena as well. I'm still surprised - sure, it was perfectly natural for me to use Microsoft software on my C64 (I didn't even know that it was Microsoft software back then), or MSDOS, or Windows 3.1 back in these days, but there always have been viable alternatives, from Atari to OS/2. Somehow, in the mid-90ies, it stopped being common to be asked "for which platform do you want it" when you bought software. Ever scince, I can't help but feeling like I'm in some cartoonesk hollywood movie when I think about Microsofts economical, social and political role.
Then again this kind of poll is useless. Sure they are dissatisified. You always says you want a better/cheaper product. If you say you are 100% satisfied they will up the price.
Indeed. The same people would likely laugh at you if you would suggest switching to any of the alternatives, or if they would switch, they would be just as dissatisfied, for other reasons.
If you want to know what people want, it is a bad idea to ask them. Look at what they do. Everybody likes to bitch, but few actually do something about it. If they would want to get rid of the security problems, the annoying licenses, the cost etc. they could, today - Mac OS X, Linux, BSD, Solaris or AIX all exist. Since they they decide to stay with Microsoft, they deserve what they get - I just wish that every new MS worm wouldn't harm non-MS users as well.
According to SoftMaker, all that was involved was a recompile. So it cost them extremely little to support FreeBSD.
As far as I understand, their main problem wasn't about porting, but about supporting another OS. I have little idea about what kind of framework they use, so I don't know how platform-independent their code really is, but it certainly has at least some potential to make support more complicated.
It would be really great if SoftMaker - or other company that made the same step, like Opera - would follow up with some data, like how many units they sold, how it affected their support expenses, etc. I can see why they wouldn't, but I would certainly be most interested in it.
I've used the demo TextMaker, and frankly it's awesome. [...] I'm certainly considering a purchase.
Well, more power to you, and SoftMaker for that matter. I don't want to be misunderstood: I think this is great - as a FreeBSD user and supporter I love these kinds of thing, and SoftMaker certainly gained some Geek credit points with it. I'm just surprised given the mostly negative feedback before. (And while I do not consider a purchase at all, simply because I can't even remember using an Office suite the last time, let alone for something OOo (which sucks more on FreeBSD than elsewhere, probably due to Sun's idea of portability - "any platform, as long as it's Windows, Solaris or Red Hat") or Gnome-Office wasn't sufficient, I realize that I'm simply not their target audience)
If this works, great. If it doesn't, we'll have a disgruntled commercial vendor and potentially (if they withdraw the offer) disgruntled end users. I'd hate this to happen because of unjustified expectations on either side.
Actually, I'm sure that any lawyer would probably tell you that, assuming what McBride says about his licenses is true, then showing ANY of the SysV sourcecode would put him in violation of his own licenses, which could land SCO in hot water with it's licensees.
IANAL either, but I don't get this. Even if they would print every single line of code that doesn't belong in the Linux kernel in a newspaper ad, what damage would be done? The code is already on thousands of hard disks, ready for review by anyone interested. Since the secret that should have been protected supposedly is the code itself, and not the fact that it is copyright-protected SCO code, the damage is already done: the code is public knowledge, no way around it.
If this really is what happened, they are certainly free to sue IBM, SGI or whoever is responsible for the damages, but not telling anybody which code we are talking about helps nobody, not even them, and it harms people who have about nothing to do with the original problem, namely all Linux users and developers.
It's nice to hear that, however, I'm mildly surprised. There was a poll at bsdforums IIRC that asked whether people would be willing to pay for it - the vast majority (including me) seemed to see no real need for it. There are already various free office suites that work quite well, and if anybody would need TextMaker, they could simply run the linux version - works fine for lots of other apps, from acrobat reader to a complete Lisp IDE including native-code compiler (generating Linux binaries, of course), debugger and profiler I happen to use with FreeBSDs Linux compatibility.
In other words, while I'm always happy to see my platform of choice supported, I wouldn't expect any significant commercial gain for SoftMaker. Most people won't be interested in a proprietary office suite - just as with Linux people, but the FreeBSD desktop market is obviously a lot smaller than even the Linux one - and others would have bought it anyway, even without a native port.
(Of course, if they would be hiring, I would be much more enthusiastic;-)
(As I would if this story wouldn't be about TextMaker, but VMWare, which is the only proprietary program that I really, really miss.)
Slashdot Mirror
To be honest, that would definitly be one of their more useful patches.
The pros as he describes them don't just "not care" about their platform, they choose them depending on what fits the task best. Same thing. And, indeed, the only sane way of thinking about software IMHO, even if lots of people (not at all limited to Linux, there are a lot of irrational windows supporters as well, for example, or BSD, BeOS, Solaris, AIX, you name it - and let's not go into text editors, unix desktops or programming languages) don't seem to be capable of it.
Let's just say his goal obviously was to generate lots of page impressions rather than to enlarge human knowledge as a whole.The good news is that now there is an official, redistributable, native Sun Java VM port, at least for FreeBSD 4 (of course, you have to download it from the FreeBSD site, not Sun's, FreeBSD isn't part of the "A" in "WORA"). It has finally passed Sun's test suite, which it didn't earlier mostly because nobody could pay Sun enough money to run it. It was too late for 5.1, and there was still a minor issue IIRC, but I'd expect it to be in 5.2.
Bottom line:
This is the same as not having a text editor, so that an intruder cannot change config files. It doesn't work. When they could use an editor or compiler on your firewall, you've already lost.
At least at the time his biography (which I won't recommend reading, the Linux kernel in a hex editor is more enlightening and entertaining literature) he drove a blue BMW Z3. And given that trademarks are only valid for a certain kind of possibly-competing products, I doubt that he could sue the owner, even if he would want to.
Then again, this issue has nothing to do whatsoever with BINDs rotten codebase. Normally you can be pretty sure that they get the DNS part right and only fuck up in the C implementation part, this time it's the other way around.
What makes the GPL more dangerous for companies is that the risk of trying to get away with copyright violations is higher. Companies that honor the law have nothing to fear, but they are obviously not Forbes' target group.
I am aware that both OOo and Java work on more platforms than this, if that's what you mean. It's just not that Sun would really care, and it shows.
It takes a strong character however to admit that even most of your toy coding projects can easily be "restored" by forgetting about them and using one of the fifty equivalent projects on sourceforge. And no, this is slashdot, I'm not supposed to have a private life.
They could at least get rid of the useless splash screen that makes /all/ of your virtual desktops unusable while that piece of "enterprise quality" code takes as long to boot up as it took Don Knuth to implement TeX.
Just one more example of Suns idea of portable software, Java is another one: Works on any platform as long as it's Windows, Solaris or Red Hat.
Encoding types in variable names helps exactly as much as calling a C string "name_that_will_never_be_longer_than_128_chars" and omitting bounds checking.
Sorry, but the argument that it's the stupid programmers fault and not the language fails if there is only one non-stupid programmer. Dan Bernstein is just too damn slow to write all the software the whole world needs himself. I'll go for a language where people only fuck up in the ways not due to a too low level of abstraction, like the bugs that occur in high-level-language programs and those of DJB (although he won't admit them usually).
So, why did you install the viruses if you didn't want them?
SFTP is a different matter however, but it's less an extension of FTP as an add-on to SSH to implement similar functionality in a completly different way. Not bad as a protocol, but it suffers from the lack of a robust SSH implementation.
What is surprising is just how insanely dominant they became, and that they suddendly managed to get a foot in the server arena as well. I'm still surprised - sure, it was perfectly natural for me to use Microsoft software on my C64 (I didn't even know that it was Microsoft software back then), or MSDOS, or Windows 3.1 back in these days, but there always have been viable alternatives, from Atari to OS/2. Somehow, in the mid-90ies, it stopped being common to be asked "for which platform do you want it" when you bought software. Ever scince, I can't help but feeling like I'm in some cartoonesk hollywood movie when I think about Microsofts economical, social and political role.
Indeed. The same people would likely laugh at you if you would suggest switching to any of the alternatives, or if they would switch, they would be just as dissatisfied, for other reasons.If you want to know what people want, it is a bad idea to ask them. Look at what they do. Everybody likes to bitch, but few actually do something about it. If they would want to get rid of the security problems, the annoying licenses, the cost etc. they could, today - Mac OS X, Linux, BSD, Solaris or AIX all exist. Since they they decide to stay with Microsoft, they deserve what they get - I just wish that every new MS worm wouldn't harm non-MS users as well.
It would be really great if SoftMaker - or other company that made the same step, like Opera - would follow up with some data, like how many units they sold, how it affected their support expenses, etc. I can see why they wouldn't, but I would certainly be most interested in it.
Well, more power to you, and SoftMaker for that matter. I don't want to be misunderstood: I think this is great - as a FreeBSD user and supporter I love these kinds of thing, and SoftMaker certainly gained some Geek credit points with it. I'm just surprised given the mostly negative feedback before. (And while I do not consider a purchase at all, simply because I can't even remember using an Office suite the last time, let alone for something OOo (which sucks more on FreeBSD than elsewhere, probably due to Sun's idea of portability - "any platform, as long as it's Windows, Solaris or Red Hat") or Gnome-Office wasn't sufficient, I realize that I'm simply not their target audience)If this works, great. If it doesn't, we'll have a disgruntled commercial vendor and potentially (if they withdraw the offer) disgruntled end users. I'd hate this to happen because of unjustified expectations on either side.
If this really is what happened, they are certainly free to sue IBM, SGI or whoever is responsible for the damages, but not telling anybody which code we are talking about helps nobody, not even them, and it harms people who have about nothing to do with the original problem, namely all Linux users and developers.
In other words, while I'm always happy to see my platform of choice supported, I wouldn't expect any significant commercial gain for SoftMaker. Most people won't be interested in a proprietary office suite - just as with Linux people, but the FreeBSD desktop market is obviously a lot smaller than even the Linux one - and others would have bought it anyway, even without a native port.
(Of course, if they would be hiring, I would be much more enthusiastic ;-)
(As I would if this story wouldn't be about TextMaker, but VMWare, which is the only proprietary program that I really, really miss.)