Steve Gibson is an idiot. He was personally affected by a DDoS, so he's now against raw sockets for emotional reasons. By his logic, Linux shouldn't be given to desktop users unless raw sockets are removed
It's not really RedHat's fault. It's the fact that they rely on a bunch of self-prescribed "programmers" who don't have the discipline to put any thought into the code (includes planning and meticulous logic analysis). No moderately-experienced programmer should ever have buffer-overflow problems bigger than "off-by-one" mistakes. But in wanting the code to "do something already", input routines are written quicky and shoddily.
I'll quit ranting now before I get nasty. Time to get some sleep.
You're right, and if he was a white famous football player, he would have landed his rear-end in jail, too. Remember, the Rodney King incident had just happened, so there was an "us vs them" mentality that wouldn't have been there if race wasn't an issue.
Heh. What's funny is that in Saskatchewan, it's illegal to use a title containing the word "Engineer" without being licensed by APEGS (Association of Professional Engineers & Geoscientists of Saskatchewan), so all those 31337 certificate-holders can call themselves MCSEs, but they can't expand that acronym anywhere.
Sort of exposes Microsoft's attempt at sophistication, no?
Say[ing] "no can do" to those demands is very hard for someone that takes pride in what they do.
You wouldn't have to do it that much. You say "it will take no less thanX amount of time" to management the first few times, but when you finally deliver, management will realize that it was worth it, because the result will be excellent. The rationale that "you'll have to spend more time/effort/money to work around bad software than you will by having to wait" seems to work pretty well, at least for me.
Finally! Someone with some sense and experience. I've been saying the same thing for the last while, and sadly, I've been mostly ignored.
I think the latest onset of buggy code has something to do with ESR's phrase "shut up and show me the code". Some confused people mess it up, and take it as, "shut up and code," which basically means "don't bother planning anything, just keep hacking until it works."
I spend about 70% of my time doing research and planning for my code. The reason that number isn't higher is because I'm damn meticulous when I code, so it takes me longer than the average person. I wish more people could say the same.
Since SE-Linux isn't ready yet (and is far from being mature), maybe the.de parliament should examine OpenBSD as an option. After all, it's got integrated crypto and it's a solid, stable OS.
Of course, paying people to make Linux secure and solid would be fine, too.:-)
Stephen King was an idiot - he thought a 70% visit/purchase ratio was too low (it's really amazingly high). King failed to take into account: people downloading but not actually reading the books, people re-downloading the books, people not wanting to pay for an unfinished work, his books are all have the same plot, many readers would have used the public library anyway, etc. Anyway, my point is that the Steven King fiasco was not a model example.
I totally agree. What the KDE team needs to do now is rip out the internals and simplify them. There's no reason why something like KDE couldn't run speedily on a P200 with 64MB of RAM if it was properly designed.
I'm not saying KDE sucks any more than anyone would say Linux 2.2's VM sucks. Both were good considerng the experience of the programmers and designers, but they both need(ed) revision. There will probably be wailing and gnashing of teeth, but when it's done, it will be worth it.
Check to make sure you don't have the "Extra Stereo" plugin (or any other effects plugins for that matter) enabled, as this will emphasize clipping and skipping.
Slashdot Mirror
Steve Gibson is an idiot. He was personally affected by a DDoS, so he's now against raw sockets for emotional reasons. By his logic, Linux shouldn't be given to desktop users unless raw sockets are removed
It's not really RedHat's fault. It's the fact that they rely on a bunch of self-prescribed "programmers" who don't have the discipline to put any thought into the code (includes planning and meticulous logic analysis). No moderately-experienced programmer should ever have buffer-overflow problems bigger than "off-by-one" mistakes. But in wanting the code to "do something already", input routines are written quicky and shoddily.
I'll quit ranting now before I get nasty. Time to get some sleep.
You're right, and if he was a white famous football player, he would have landed his rear-end in jail, too. Remember, the Rodney King incident had just happened, so there was an "us vs them" mentality that wouldn't have been there if race wasn't an issue.
How about using the OBVIOUS argument???
The O.J. Simpson trial was a race issue, not a money issue.
Fair enough.
Sort of exposes Microsoft's attempt at sophistication, no?
Also: "Keep It Simple, Stupid" works too.
You wouldn't have to do it that much. You say "it will take no less than X amount of time" to management the first few times, but when you finally deliver, management will realize that it was worth it, because the result will be excellent. The rationale that "you'll have to spend more time/effort/money to work around bad software than you will by having to wait" seems to work pretty well, at least for me.
I think the latest onset of buggy code has something to do with ESR's phrase "shut up and show me the code". Some confused people mess it up, and take it as, "shut up and code," which basically means "don't bother planning anything, just keep hacking until it works."
I spend about 70% of my time doing research and planning for my code. The reason that number isn't higher is because I'm damn meticulous when I code, so it takes me longer than the average person. I wish more people could say the same.
Of course, paying people to make Linux secure and solid would be fine, too. :-)
s/corporate/GPL/
Um... OpenBSD can run anything Linux can that's not Linux-specific.
BSD doesn't run on top of Linux, which doesn't run on top of HURD, and KDE, GNOME and XFree all run on top of the GNU toolset.
Actually, I believe its official name is Debian GNU/HURD.
Stephen King was an idiot - he thought a 70% visit/purchase ratio was too low (it's really amazingly high). King failed to take into account: people downloading but not actually reading the books, people re-downloading the books, people not wanting to pay for an unfinished work, his books are all have the same plot, many readers would have used the public library anyway, etc. Anyway, my point is that the Steven King fiasco was not a model example.
Of course it must be! We all know that Apache on Debian runs much slower than Apache on SuSE...
With IPv6 and/or IPv4-multicast, this wouldn't be much of a problem. Bug your ISP about getting IPv6!
What do you use to separate the numbers in a coordinate? (5,3)
I totally agree. What the KDE team needs to do now is rip out the internals and simplify them. There's no reason why something like KDE couldn't run speedily on a P200 with 64MB of RAM if it was properly designed.
I'm not saying KDE sucks any more than anyone would say Linux 2.2's VM sucks. Both were good considerng the experience of the programmers and designers, but they both need(ed) revision. There will probably be wailing and gnashing of teeth, but when it's done, it will be worth it.
Check your RAM. If it's not a RAM bug, it's probably an X server bug.
I've never had a non-hard-lock situation that the Magic SysRQ keys couldn't handle.
Some of them are. Others are completely software-based.
Check to make sure you don't have the "Extra Stereo" plugin (or any other effects plugins for that matter) enabled, as this will emphasize clipping and skipping.
Well using RPM directly on a non-RPM system is kinda dangerous (you can get packages overwriting each other's files).