Slashdot Mirror


User: smash

smash's activity in the archive.

Stories
0
Comments
7,084
First seen
Last seen
Profile
(view on slashdot.org)

Comments · 7,084

  1. Re:A few choice quotes from Theo de Raadt on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    And the counterpoint to that is: by releasing under the GPL, you are restricting the ability of your code to make the world a better place. BSD allows the code to be used ANYWHERE, rather than commercia vendors being forced to reinvent the wheel. Reinventing the wheel is a waste of time/money that could be better used to write their own code to enhance their product for the benefit of their users. The free BSD code is still out there, a commercial entity making use of it does not detract from its existance. BSD people don't care about others using their code to make money in commerical projects. If this happens, all the better.

  2. Re:Why so harsh? on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    Do you know how many run Windows?

  3. Re:But it did... on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    Does it matter? Either way, the code was returned, and the BSD objective of ensuring wider distribution of stable well tested code was met (vs the gnu objective of "commercial software must lose").

  4. Re:Linux made UNIX easier on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    Try using BSD and reading the docs. It is not linux so your linux knowledge does not apply. However in terms of ease of use, they are on par (i've used both since around 2000, started on linux in 96).

  5. Re:Linus is right on about microkernels on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    Win7 does this. Had a dicky video card crashing due to heat. Black screen, 2 second delay, screen comes back. System still up. Not that windows 7 is my favourite OS at all, but it is one thing microsoft has got fairly right in recent years. vista/7 onwards are not the same as XP or previous in this respect.

  6. Re:Linus is right on about microkernels on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    How about surviving a crash and safely writing data to disk, instead of hard crashing due to a network fault? Sure in a firewall that may not be such a concern, but if you're running a DB server it sure is. Horses for courses, etc...

  7. Re:Linus is right on about microkernels on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    I do tend to agree with this. Right now if my display is borked but I can get in via ssh I'm still pretty-much SOL. With a microkernel design I could probably reboot the entire display stack (just tell those drivers that the computer is booting up - unload them from memory and load them back in and all that). It would wipe out all my video buffers and maybe kill my X11 session, but it isn't a reboot. If you go a step further you could have a watchdog of some kind or some other way of triggering this and maybe do that proactively.

    Win7 actually does this. I had a fan die on my 8800GT and the video was crashing regularly. Screen goes black, video driver reloads, screen comes back. This is far preferable to complete crash, losing work, stopping network traffic, etc, etc.

  8. Re:The BSD community just doesn't accept stupidity on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    Not to be a simple "me too" but... "me too". The freebsd documentation is excellent (vs GNU "we don't like man, we prefer info" - which is shit). Linux has API changes every foo weeks which breaks stuff. FreeBSD might be a little slower out of the gate, but once something is implemented, its generally done properly and the documentation from 5 years ago is pretty much usable today.

    I started out with Linux in 1995. Was exposed to solaris in 1998 and tried out BSD in 2001. Haven't looked back. Linux is just too "different" to everything else and generally not for any real good reason. If you're a linux user trying BSD (who has never been exposed to other unix), it works quite differently and will screw with your head for a bit. However it is more "the unix way" and makes more sense once you understand it. And as above, the handbook/docs are excellent.

  9. Re:Disagree on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    Dtrace offers nothing not offered by systemtap. Indeed, systemtap does things dtrace doesn't.

    You're kidding, right? How's that GNU cool-aid?

  10. Re:Frozen, I tells you on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    (b) Hurd development was an elitist, academic cluster-fuck for 20 years.

    When did this change?

  11. Re:BSD far more common via Mac OS X on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    I think you have unix confused with plan9

  12. Re:This is getting old on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    This might have been true 20 years ago, but most people don't run dos apps any more. Windows beats linux because an idiot can install/mis-use it.

    And consultants can charge money to maintain/babysit it.

  13. Re:This is getting old on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    The problem with the BSD license is the fact that any company can take the code and make it their own without giving anything back so it is impossible to determine how much of Apple's code is BSD or not since their code is propitiatory.

    This is a feature. BSD aligned people want ALL software to be better through the use of standards based, well-tested code, and release their code for free use/distribution as a result. The GNU types want commercial software to lose.

    There's a fundamental philisophical difference here, and personally i think the BSD way is preferable.

  14. Re:This is getting old on Andrew Tanenbaum On Minix, Linux, BSD, and Licensing · · Score: 1

    Or - to rephrase - as nice as Mac's BSD under the skin OS is, how much has it actually improved the BSD experience of anyone else?

    Clang. Grand central. Biggest use of d-trace outside of solaris. Zeroconf, webkit, etc. Apple releases plenty of open source, and exposes the freebsd userland to a far greater audience than the tools would otherwise see. This means that bugs are more likely to be uncovered.

    If you use the BSD code sure you can refuse to contribute upstream. However its shooting yourself in the foot as far as taking advantage of further free development goes.

  15. Re:Windows is not optimized for Bulldozer on Bulldozer Server Benchmarks Not Promising · · Score: 2

    Be this as it may: Windows is what 99% of people run. By the time Windows 8 hits RTM, the CPU landscape is going to look pretty different.

  16. Re:And moreover on Bulldozer Server Benchmarks Not Promising · · Score: 1

    When you're trying to fab 2x as many transistors per die as the next guy (intel) for comparable performance, yield is always going to be an issue.

  17. Re:Bulldozer outdated already ? on Bulldozer Server Benchmarks Not Promising · · Score: 1

    The cyrix was actually considerably better than intel at integer. This.... not so much.

  18. Re:Bulldozer outdated already ? on Bulldozer Server Benchmarks Not Promising · · Score: 4, Informative

    wrong. the 386-sx had a 16 bit memory bus (vs 32 bit on the DX). It had no FPU, that was a separate socket.

  19. Re:Bulldozer outdated already ? on Bulldozer Server Benchmarks Not Promising · · Score: 1

    Yeah, but FPU performance sucked rocks until the pentium came along and pipelined it.

  20. Re:AMD needs its swagger back on Bulldozer Server Benchmarks Not Promising · · Score: 1

    Uh... 2007, rather.

  21. Re:AMD needs its swagger back on Bulldozer Server Benchmarks Not Promising · · Score: 1

    My former primary machine, a core 2 Q6600 does that just fine and i bought that back in 2006.

  22. Re:AMD needs its swagger back on Bulldozer Server Benchmarks Not Promising · · Score: 2

    I'm sure intel can out-design intel any time it wants, but generally back in the real world they hit the mark pretty well with what actually matters spec sheet wise to get good performance. Be it smaller L1/L2 cache, or lack of onboard memory controller until well after AMD got it on die, intel has proven time and time again that they've been able to get more with less. The dozer has 2x-3x as many transistors and 50% more cores to barely beat a xeon that is about to be superseded. I wouldn't call that a win by any stretch.

  23. Re:I don't get it. It beat the Xeons?? on Bulldozer Server Benchmarks Not Promising · · Score: 1

    It also used 2-3x as many transistors to do so....

  24. Re:Lots of Real World Users (TM) try to use all co on Bulldozer Server Benchmarks Not Promising · · Score: 1

    The US Office of Management and Budget (OMB) has a virtual to physical server target of 15:1.

    Every large business, and most medium sized ones, are going to try to (at least) match that target.

    (athough memory seems to be a bigger constraint.)

    They're still not likely to use all the cores unless they have some peculiar workload. They'll run out of RAM and IO (on a single server) first.

  25. Re:Speak for yourself on Whither the Portable Optical Drive? · · Score: 1

    So, what computer have you been using those CDs/DVDs in for the last decade, genius?