I claim patent on using 'make' since I've been using it to parallel start and stop MMO game services since 2004 -- based on experience with Veritas Cluster Server.:P
This is the same muddled thinking that validated the Iraq war vote. There is what is the right thing to do, and then there's pandering and hoping it'll work out.
As you've been saying to all of us that our singular vote counts, so does yours. It may not count in the overall trend, but at least a stand was made not to be swayed by the crowd. You have become the sheep.
This is what happens when you're on the Linux Treadmill and have to suck in your software through the Redhat/Fedora/RPM straw.
A better solution is a generalized framework that allows you to download direct from the Source with only minor patches during compilation to install for your OS system.
Ports is such a framework to avoid this vendor lock-in and inability to get the latest releases and bug fixes. Why do you need to be dependent on some small specialized group for back ports that will die at any time?
Not only is Ports community supported so it's free, but it's OPEN-SOURCE so it can't die and usually no tweaks are necessary for the latest vendor source release. You can't even get this close with SRPMS.
UNIX, BSD, Linux are all open-source, but ONLY Linux/RedHat/Fedora chooses to avoid the benefits of open-source.
You could have smartly have just left it alone as just libc and its standardized API that has worked for years, and move all the specialized fast moving APIs out into its own library.
I can install multiple version of gtk/qt whenever I want without being forced to choose particular version of Gnome/KDE, etc.
Even when I upgrade the OS, it leaves compatibility libs around so old apps can still link to it while everybody else moves on.
As an example, I can upgrade from FreeBSD 6 to FreeBSD 7 without recompiling any of the packages I've already had installed! Isn't this supposed to be impossible?
The only time I have to say recompile anything is if I upgrade to a new PHP, then ONLY the PHP modules and extensions need updating. Not Apache, not SSL, not mod_perl, nada. I can upgrade bdb from 4.2 to 4.4 and the only thing I need to rebuild is ruby-bdb.
The point I'm making is so many of you have been locked into this box you can't see that there's a world that doesn't need this.
The Linux distributions got all of you by the balls with the precompiling of packages and products that are dependant on the OS release.
It also seems many of the Linux world savor this without realizing how you got cold-cocked. Ask the really important question of "WHY does Gnome 3.xx.x REQUIRE FC/RHEL/LinuxN 4.xx, and also WHY can't the same Gnome 3.xx.xx run on FC/RHEL/LinuxN 7.xx?" Did some API change between OS releases that broke everything? Why are you using an OS where every release is an 'API' breakage??
Ask yourself, does every release of World of Warcraft/Microsoft Office/Dawn of War require you to have a SPECIFIC Windows XP/2000/Vista version release (beyond security patches) with a complicated web of dependencies? Why can't I install multiple versions of gcc/OpenOffice/Python/etc in the way the original vendor didn't limit me?
Your middle-man got you dependent on them for their drug 'releases' in order to fulfill your basic needs.
The base OS and the package system SHOULD NOT have any dependencies on each other and ideally should not even care which version of what you have installed beyond simple ABI limitations. It should allow side-by-side execution of multiple versions of programs/libs, because software is long-lived and everybody has different needs.
This is another reason Linux will never overshadow Windows simply because the community and the commercialization has chained the users to this never ending, single-pipeline upgrade treadmill in a single-release turn-key/embedded-like/PalmOS-like system fashion. Funny enough, raw Linux itself doesn't limit the users either.
So, if you can get away from the endless, locked-in, upgrade whipping, you can then focus directly on the app itself and which version serves your needs best instead of which OS has the version of the application you need. The decision between PHP4 or PHP5 requiring hours of paid, 'managed' care support contracts because the vendor doesn't "support" it.
Myself, I live on Ports which suffers no such middle-man, straw-sucking limitations beyond keeping the port framework up-to-date.
Currently deployed Windows Mobile is a working subset of the Win32 api used on PCs, uses the same 'standards compliant' protocols to communicate with other computers, and the only difference is it was compiled for the ARM.
With such a blistering example that the x86 instruction set is unnecessary to achieve platform portability, what advantage can the Atom bring the table? Out-of-order execution isn't really a big advantage especially when ARM already is already sampling such cpus already.
The current SoC embedded deployments are converging with PC's destiny, and what will really come to matter is the packaging prettiness (iPhone) and implementation of the system and ISA (unlike HTC's and Qualcomm's inability to ship working video drivers for the Kaiser/Tilt), than how fast and fancy the CPU is.
The last point is especially important when we want a subset of the desktop/laptop functionality with all the expected modern polish without destroying the technology we still can't improve - the battery.
Another way is to stop wasting time trying to fix human fallibility by increasing lanes, and instead go straight to a train/trolley system that minimizes the standing wave issue. Trains/trolleys fix it through predictable automation/highly-trained-drivers for those high-volume routes.
What about Wipeout and Cold Storage, Underworld, etc etc? It's one of the earliest games to start introducing people to music that's not Top40 radio crap. Who doesn't have the original full Wipeout soundtrack?
That's the fault of all the dumb ass wannabe programmers out there who are getting by as corporate-paid web programmers. Most only learned it up to 2.0 spec; maybe 3.0. For further progress, they just shunted into Flash and that's it! It's not like there's compilers, nor pressure, to comply with the standard.
It's all about the lazy man's philosophy that as long as it works, on to the next project. Until browsers start failing bad syntax, the web will stay with the basics, and new HTML revisions are just more leavings of white tower masturbation.
More likely, it'll end up being that slashed up and reimagined trash because Americans really don't get this robot thing. It's certainly not in their culture, nor technology they export.
It means Linus isn't omniscient, and shouldn't be pushing the ideology that "the whole and only point of a distribution is to make it easy to install." This is especially true when his premise is obsoletely based on Debian being a compile distribution, and right after he was talking about "may the best code win."
If he can contradict himself in one article, I wonder what else is he wrong about, and ALSO based on false premises because he's now too old and lazy to explore it himself.
He's displaying what I think are the classical traits of a guy who's been around so long he figures he knows everything about the subject, and doesn't need to look at what new or better ideas are out there.
Sounds like the tool is not the question, but people being too freaking lazy to learn how to work with tags, branches and merges. I'm not reading anything about how Git really changes how you use VCS. Branching isn't annoying, it's a required skill for proper VCS usage which I've found lacking in many.
Claiming Git is superior because you can substitute "local repositories" for proper tagging and branches is derivative, and isn't enough to convince me that it solves a real problem. I'v dealt with it in CVS every day for the last 7 years in corporate development, so I'm quite open to anything better, but the only improved developments have been atomic commits and changesets, which Perforce has it in spades, and not this local repository ability.
I can respect separate repositories in that it solves the "open, shared access to one repository" issue, and makes it easier to have loosely coupled development groups working on the same codebase, but that still only delays the inevitable task of merging and working out conflicts at the end. I think it's still better to have little experimental branches that improve a small feature-set and then quickly merges back into trunk.
What else does Git do that really simplifies and improves the VCS experience?
Slashdot Mirror
ifconfig cel0 down
wtf
RTS means you send units out to some location and let them sort it out while one plans the bigger picture of the war -- not the individual battles.
RTT is where you send units out to some location and micromanage each battle out. Most of the "RTS" games are really only RTT.
Also, Herzog Zwei predates Dune II and better qualifies as one of the first RTT games.
We welcome our Git brothers for reinventing the wheel again versus CVS's established http://www.cvsup.org/ for mirroring and distribution.
The more the merrier.
I claim patent on using 'make' since I've been using it to parallel start and stop MMO game services since 2004 -- based on experience with Veritas Cluster Server. :P
Didn't help last time when it was Bush vs Kerry vs Cthulhu.
Maybe voting Kucinich, or even Ron Paul, might be better for some real change.
This is the same muddled thinking that validated the Iraq war vote. There is what is the right thing to do, and then there's pandering and hoping it'll work out.
As you've been saying to all of us that our singular vote counts, so does yours. It may not count in the overall trend, but at least a stand was made not to be swayed by the crowd. You have become the sheep.
People just don't want to admit that Tesla's 'death ray', or the Wardenclyffe event actually did something so remotely away.
This is what happens when you're on the Linux Treadmill and have to suck in your software through the Redhat/Fedora/RPM straw.
A better solution is a generalized framework that allows you to download direct from the Source with only minor patches during compilation to install for your OS system.
Ports is such a framework to avoid this vendor lock-in and inability to get the latest releases and bug fixes. Why do you need to be dependent on some small specialized group for back ports that will die at any time?
Not only is Ports community supported so it's free, but it's OPEN-SOURCE so it can't die and usually no tweaks are necessary for the latest vendor source release. You can't even get this close with SRPMS.
UNIX, BSD, Linux are all open-source, but ONLY Linux/RedHat/Fedora chooses to avoid the benefits of open-source.
Then glibc is a poor place to put all your eggs.
You could have smartly have just left it alone as just libc and its standardized API that has worked for years, and move all the specialized fast moving APIs out into its own library.
I can install multiple version of gtk/qt whenever I want without being forced to choose particular version of Gnome/KDE, etc.
Even when I upgrade the OS, it leaves compatibility libs around so old apps can still link to it while everybody else moves on.
As an example, I can upgrade from FreeBSD 6 to FreeBSD 7 without recompiling any of the packages I've already had installed! Isn't this supposed to be impossible?
The only time I have to say recompile anything is if I upgrade to a new PHP, then ONLY the PHP modules and extensions need updating. Not Apache, not SSL, not mod_perl, nada. I can upgrade bdb from 4.2 to 4.4 and the only thing I need to rebuild is ruby-bdb.
The point I'm making is so many of you have been locked into this box you can't see that there's a world that doesn't need this.
The Linux distributions got all of you by the balls with the precompiling of packages and products that are dependant on the OS release.
It also seems many of the Linux world savor this without realizing how you got cold-cocked. Ask the really important question of "WHY does Gnome 3.xx.x REQUIRE FC/RHEL/LinuxN 4.xx, and also WHY can't the same Gnome 3.xx.xx run on FC/RHEL/LinuxN 7.xx?" Did some API change between OS releases that broke everything? Why are you using an OS where every release is an 'API' breakage??
Ask yourself, does every release of World of Warcraft/Microsoft Office/Dawn of War require you to have a SPECIFIC Windows XP/2000/Vista version release (beyond security patches) with a complicated web of dependencies? Why can't I install multiple versions of gcc/OpenOffice/Python/etc in the way the original vendor didn't limit me?
Your middle-man got you dependent on them for their drug 'releases' in order to fulfill your basic needs.
The base OS and the package system SHOULD NOT have any dependencies on each other and ideally should not even care which version of what you have installed beyond simple ABI limitations. It should allow side-by-side execution of multiple versions of programs/libs, because software is long-lived and everybody has different needs.
This is another reason Linux will never overshadow Windows simply because the community and the commercialization has chained the users to this never ending, single-pipeline upgrade treadmill in a single-release turn-key/embedded-like/PalmOS-like system fashion. Funny enough, raw Linux itself doesn't limit the users either.
So, if you can get away from the endless, locked-in, upgrade whipping, you can then focus directly on the app itself and which version serves your needs best instead of which OS has the version of the application you need. The decision between PHP4 or PHP5 requiring hours of paid, 'managed' care support contracts because the vendor doesn't "support" it.
Myself, I live on Ports which suffers no such middle-man, straw-sucking limitations beyond keeping the port framework up-to-date.
What is 'Windows compatibility'?
Currently deployed Windows Mobile is a working subset of the Win32 api used on PCs, uses the same 'standards compliant' protocols to communicate with other computers, and the only difference is it was compiled for the ARM.
With such a blistering example that the x86 instruction set is unnecessary to achieve platform portability, what advantage can the Atom bring the table? Out-of-order execution isn't really a big advantage especially when ARM already is already sampling such cpus already.
The current SoC embedded deployments are converging with PC's destiny, and what will really come to matter is the packaging prettiness (iPhone) and implementation of the system and ISA (unlike HTC's and Qualcomm's inability to ship working video drivers for the Kaiser/Tilt), than how fast and fancy the CPU is.
The last point is especially important when we want a subset of the desktop/laptop functionality with all the expected modern polish without destroying the technology we still can't improve - the battery.
Apparently, this guy hasn't heard of Carbon (aka clib and BSD) where a lot of OS X stuff still gets programmed.
Ya, but obviously forks, nail clippers, and any significant amount of liquid in one place is more dangerous.
No need to fear what's in that one pound of battery, or a few people putting together a few separated ounces of IED fluids into the toilets.
Another way is to stop wasting time trying to fix human fallibility by increasing lanes, and instead go straight to a train/trolley system that minimizes the standing wave issue. Trains/trolleys fix it through predictable automation/highly-trained-drivers for those high-volume routes.
And we waited for Linux to become a real-ish unix for over 10 years. ;)
What about Wipeout and Cold Storage, Underworld, etc etc? It's one of the earliest games to start introducing people to music that's not Top40 radio crap. Who doesn't have the original full Wipeout soundtrack?
The follow ups have been eh.. however.
That's the fault of all the dumb ass wannabe programmers out there who are getting by as corporate-paid web programmers. Most only learned it up to 2.0 spec; maybe 3.0. For further progress, they just shunted into Flash and that's it! It's not like there's compilers, nor pressure, to comply with the standard.
It's all about the lazy man's philosophy that as long as it works, on to the next project. Until browsers start failing bad syntax, the web will stay with the basics, and new HTML revisions are just more leavings of white tower masturbation.
Ya, it's called Second Life now.
You forgot:
Final Cut: Extreme!
And he still saves more lives per minute than your rant will in your lifetime.
Except, it's a one syllable diphthong. Your KEE-yo-to looks like 3 instead of 2.
:)
Correctly, it's KEyo-to (2 syllables), or REyuu (one) and sounds more like liuu but with the L pronounced halfway like an R.
You know how they make fun of Asians that say rice like lice. Say RLiuu rike that.
More likely, it'll end up being that slashed up and reimagined trash because Americans really don't get this robot thing. It's certainly not in their culture, nor technology they export.
:P
Live action Clash of the Bionoids, yay!
It means Linus isn't omniscient, and shouldn't be pushing the ideology that "the whole and only point of a distribution is to make it easy to install." This is especially true when his premise is obsoletely based on Debian being a compile distribution, and right after he was talking about "may the best code win."
If he can contradict himself in one article, I wonder what else is he wrong about, and ALSO based on false premises because he's now too old and lazy to explore it himself.
He's displaying what I think are the classical traits of a guy who's been around so long he figures he knows everything about the subject, and doesn't need to look at what new or better ideas are out there.
Sounds like the tool is not the question, but people being too freaking lazy to learn how to work with tags, branches and merges. I'm not reading anything about how Git really changes how you use VCS. Branching isn't annoying, it's a required skill for proper VCS usage which I've found lacking in many.
Claiming Git is superior because you can substitute "local repositories" for proper tagging and branches is derivative, and isn't enough to convince me that it solves a real problem. I'v dealt with it in CVS every day for the last 7 years in corporate development, so I'm quite open to anything better, but the only improved developments have been atomic commits and changesets, which Perforce has it in spades, and not this local repository ability.
I can respect separate repositories in that it solves the "open, shared access to one repository" issue, and makes it easier to have loosely coupled development groups working on the same codebase, but that still only delays the inevitable task of merging and working out conflicts at the end. I think it's still better to have little experimental branches that improve a small feature-set and then quickly merges back into trunk.
What else does Git do that really simplifies and improves the VCS experience?
weird interaction with noscript.
How do you kill what which has no life?
If you kill a sand worm, it will only shatter into many sand trout to form other worms.