There wouldn't be two of you for very long. I'm sure many of you readers have noticed evidence of chaos in individuals and society for some time now. Well, there's reason that we continually get perturbed off our favorite limit cycles with a positive Lorenz exponent.
Environmental influences.
Presumably your organic brain would be affected by a different set of environmental factors (free radicals, cosmic rays, getting hit by a truck, etc.) that would cause your neurons to begin firing along a different sequence than would be followed by the downloaded brain sitting in silicon (or whatever.)
This gives a whole new meaning to the term Service Pack...
I'd agree the best sysadmins know how every single part under the hood contributes or detracts from their overall system performance.
That said, however, many folks (like me) are too lazy or don't have enough time to read up on the nuances of performance. Honest, I've really been meaning to read that hdparm man page(sigh.)
Personally, I think one of the greatest strengths that Linux will offer in the future will rest on Dynamic Dummy Configuration that translates 5 or 10 simple-minded radio box checks and automatic hardware probing into 423 Right Answers(TM) for an actual build of the kernel from source. Granted, it will never be as perfect as a perfectly knowledgable sysadmin carefully weighing each and every answer, but the time value of that kind of sysadmin could probably be recouped by just buying another hunk-o-hardware.
Source access and compilation for the individual platform and usage profile gives the casual sysadmin (and most can't afford to be more than casual, given the demands made on them) more potential flexibility, adaptability and performance in the long run than dynamically loaded modules can do (even if they're done really well.)
It might conceivably get to the point where you could configure the same source distro to run either as the OS on the NIC alone, or for a cell-phone, or for a gamer's kiosk.
I really like the low CPU overhead of SCSI and was looking into what the latest and greatest setup would be for high performance low overhead disk I/O. Ultra160 looked like one of newest greatest things (along the same lines as You can't be too rich or too thin...)
However, disappointment soon set in as the only Ultra160 controller I found was only available for PowerPC hardware (that should have been a hint, I know, but that x86 stuff is so alluringly cheap...)
It looks to me as if an Ultra160 PCI controller, even if loaded up with enough 10K rpm disks to feed it on the bottom side, might max out a 66 MHz/32bit wide PCI bus on the top side.
Does FibreChannel hit this same bottleneck and do folks find it worthwhile over, say, U2W SCSI ?
The main reason Scott Adams can get so much mileage out of PHBs and their silly antics is 2-fold:
95% of the audience is managed by someone else.
11/12 managers are incompetent in one or more respects
and the basic reason for the preponderance of bad managers that are easy to poke fun at from the armchair is....
Good Management is a Difficult Job that Few are Qualified to Do
So few managers have what an earlier poster summarized as "brains" and "ears" that most organizations run hobbled.
Codified management techniques in Quality are great iff they help managers and the managed to grasp the fundamental precept that
they need to be able to communicate 2-ways (eg., these are my expectations, what are your expectations, come to me with problems and I will really try to do something about them by going to bat for you against powerful ogres that scare you and me both, contribute to a team and win, flimsy excuses 3 months after the damage is already done makes you lose, etc.) with their underlings,
not to be blinded by their own preconceptions of value (they're "like me" and therefore of greater value so I'll fulfill the Peter Principle and promote ubergeek to manager), and
to recognize where and how people can contribute most to the success of the business (you organize well, you cross t's and dot i's but not on time, you communicate well with others, you code great but can't get along with meatspace, etc.).
Well, if the Macinstosh ports of M$Ware are any indication, the *NIX versions will be just ever so slightly annoyingly incompatible or incapable with what the same application does running on M$-OS.
Is that not the main reason for the corporate IT world quashing out Macs in their stables (I means, other than the fun and amusement from annoying Mac Users)?
I recall seeing an announcement for the K6-IIIp 380 MHz a while back, but never saw much in the way of notebooks, except for a smattering of the "lower end" sumo size and weight variety...
I really was hoping for something nice from AMD in that arena.
Now that Celeron's and PII's are gobbling up the low end desktop away from the K6, the Athlon has but a short window to shine like the star it is before the magical marketing bean 1000 MHz Intel chips eat everyone's lunch. AMD has suffered so many loss making quarters that it's doubtful they can stand many more rounds in the ring with the 800 lb gorilla.
Alas, no one ever wants to read someone else's code because it's not up to their own high standards. Yes, my boy, we live in a world with different multiple high standards.
Same sentiments with double strength when it comes to
documenting
testing
debugging
maintaining
some else's code.
Anyway, I suggest an escalation of the contest from a mere editorial review (kind of like looking for gemstones in the barn). Some of those reviewers are damn good coders in their own right! Give them extra credit for submissions of design patterns patches (judged according to how much they show elegence, less lines of code, more extensible, less memory, fewer bugs in future) that could usefully be employed where they are not at present!
Deciding on appropriate prizes is difficult. I haven't seen the Netscape cheerleaders on TV lately, so a group photo with the celebs is probably out of the question. However, I'd say some recognition and acknowledgement is due those who contribute to Mozilla the most. A little semi-official defacement of the Netscape portal to include a Top Tech News Story headline like "Studly Programmers Tire of Success" might be in order.
My baby bell has been multiplexing my line with others so where I used to get 48kb and 50 kb connections on a V.90 modem I now never get more than 33kb. I'm guessing they don't want to pay for more copper.
Did my phone bill drop to 33/48 of the old price? I DON'T THINK SO.
The nice phone rep trying to be helpful told me that "there seems to be no problem with your line" and that, well, they were only obligated to provide 9600 b service under their agreement with our regulator$.
I get those same requests for bucks from MIT as well. In fact, changing addresses every few years is not sufficient to escape them - I think MIT must use the same powerful direct mail services as the large credit card companies in order to keep up to date with alums that might otherwise escape their mailing lists.
But all of this doesn't surprise me coming from what I learned was legally The MIT Corporation. Whether the DNS domain should have been mit.edu or mit.com is a tossup.
It's interesting they actually got in bed with MS, who, in the parlance of that excellent book, Hard Drive, has been known in the industry for date rape when it comes to new technologies and ideas (after the visit to a sweet young software boutique to explore a partnering relationship no phone calls are returned the next morning, or else you get an offer you can't refuse.) MIT is traditionally pretty careful about extracting the maximum monetary benefit from the work of its faculty and students. It is unlike them to give in that much to anyone. It must figure the bird in the hand is worth it.
While MS is capable of stronger arm twisting than other corporations, note that Project Athena was funded by IBM and DEC and, no, most of those computers did not run MVS and VMS. While I haven't been there in years, I get the feeling the dorms are wired with Ethernet and more than a few Linux machines.
The real motivation for MS with its MIT partnership is to attach a large pipe fitting at the mouth of the Course VI (EE/CS) graduating class to get new talent. The old talent at MS probably gets burned out managing the complexity of 1e7 lines of code that is deliberately twisted into a Gordian Knot with inextricable connections and fragile tendrils leading back to revisions n-1, n-2,... in order to impede competitors. Once those stock options become vested the reasons for staying in Redmond and putting in 80+ hour weeks begin to go away (FYIFV). Then the corporate management finds it's time to find new gelflings to write Word 2002.
Despite my lifestyle I still have two hands, fortunately.
Nevertheless, once I saw this device I seriously was thinking of getting one just for the improved efficiency of input, removing big clunky keyboard, and freeing up one hand for "other important tasks."
Slashdot Mirror
There wouldn't be two of you for very long. I'm sure many of you readers have noticed evidence of chaos in individuals and society for some time now. Well, there's reason that we continually get perturbed off our favorite limit cycles with a positive Lorenz exponent.
Environmental influences.
Presumably your organic brain would be affected by a different set of environmental factors (free radicals, cosmic rays, getting hit by a truck, etc.) that would cause your neurons to begin firing along a different sequence than would be followed by the downloaded brain sitting in silicon (or whatever.)
This gives a whole new meaning to the term Service Pack...
I'd agree the best sysadmins know how every single part under the hood contributes or detracts from their overall system performance.
That said, however, many folks (like me) are too lazy or don't have enough time to read up on the nuances of performance. Honest, I've really been meaning to read that hdparm man page(sigh.)
Personally, I think one of the greatest strengths that Linux will offer in the future will rest on Dynamic Dummy Configuration that translates 5 or 10 simple-minded radio box checks and automatic hardware probing into 423 Right Answers(TM) for an actual build of the kernel from source. Granted, it will never be as perfect as a perfectly knowledgable sysadmin carefully weighing each and every answer, but the time value of that kind of sysadmin could probably be recouped by just buying another hunk-o-hardware.Source access and compilation for the individual platform and usage profile gives the casual sysadmin (and most can't afford to be more than casual, given the demands made on them) more potential flexibility, adaptability and performance in the long run than dynamically loaded modules can do (even if they're done really well.)
It might conceivably get to the point where you could configure the same source distro to run either as the OS on the NIC alone, or for a cell-phone, or for a gamer's kiosk.
Then, I may just be full of 5hit.
I really like the low CPU overhead of SCSI and was looking into what the latest and greatest setup would be for high performance low overhead disk I/O. Ultra160 looked like one of newest greatest things (along the same lines as You can't be too rich or too thin...)
However, disappointment soon set in as the only Ultra160 controller I found was only available for PowerPC hardware (that should have been a hint, I know, but that x86 stuff is so alluringly cheap...)
It looks to me as if an Ultra160 PCI controller, even if loaded up with enough 10K rpm disks to feed it on the bottom side, might max out a 66 MHz/32bit wide PCI bus on the top side.
Does FibreChannel hit this same bottleneck and do folks find it worthwhile over, say, U2W SCSI ?
GUI is to command line as TV is to books.
Quite so.
The main reason Scott Adams can get so much mileage out of PHBs and their silly antics is 2-fold:
- 95% of the audience is managed by someone else.
- 11/12 managers are incompetent in one or more respects
and the basic reason for the preponderance of bad managers that are easy to poke fun at from the armchair is....Good Management is a Difficult Job that Few are Qualified to Do
So few managers have what an earlier poster summarized as "brains" and "ears" that most organizations run hobbled.
Codified management techniques in Quality are great iff they help managers and the managed to grasp the fundamental precept that
It's just that easy and just that hard.
Well, if the Macinstosh ports of M$Ware are any indication, the *NIX versions will be just ever so slightly annoyingly incompatible or incapable with what the same application does running on M$-OS.
Is that not the main reason for the corporate IT world quashing out Macs in their stables (I means, other than the fun and amusement from annoying Mac Users)?
Quite so!
I recall seeing an announcement for the K6-IIIp 380 MHz a while back, but never saw much in the way of notebooks, except for a smattering of the "lower end" sumo size and weight variety...
I really was hoping for something nice from AMD in that arena.
Now that Celeron's and PII's are gobbling up the low end desktop away from the K6, the Athlon has but a short window to shine like the star it is before the magical marketing bean 1000 MHz Intel chips eat everyone's lunch. AMD has suffered so many loss making quarters that it's doubtful they can stand many more rounds in the ring with the 800 lb gorilla.
Alas, no one ever wants to read someone else's code because it's not up to their own high standards. Yes, my boy, we live in a world with different multiple high standards.
Same sentiments with double strength when it comes to
- documenting
- testing
- debugging
- maintaining
some else's code.Anyway, I suggest an escalation of the contest from a mere editorial review (kind of like looking for gemstones in the barn). Some of those reviewers are damn good coders in their own right! Give them extra credit for submissions of design patterns patches (judged according to how much they show elegence, less lines of code, more extensible, less memory, fewer bugs in future) that could usefully be employed where they are not at present!
Deciding on appropriate prizes is difficult. I haven't seen the Netscape cheerleaders on TV lately, so a group photo with the celebs is probably out of the question. However, I'd say some recognition and acknowledgement is due those who contribute to Mozilla the most. A little semi-official defacement of the Netscape portal to include a Top Tech News Story headline like "Studly Programmers Tire of Success" might be in order.
Not correct.
My baby bell has been multiplexing my line with others so where I used to get 48kb and 50 kb connections on a V.90 modem I now never get more than 33kb. I'm guessing they don't want to pay for more copper.
Did my phone bill drop to 33/48 of the old price? I DON'T THINK SO.
The nice phone rep trying to be helpful told me that "there seems to be no problem with your line" and that, well, they were only obligated to provide 9600 b service under their agreement with our regulator$.
Call me spoiled rotten...
I get those same requests for bucks from MIT as well. In fact, changing addresses every few years is not sufficient to escape them - I think MIT must use the same powerful direct mail services as the large credit card companies in order to keep up to date with alums that might otherwise escape their mailing lists.
But all of this doesn't surprise me coming from what I learned was legally The MIT Corporation. Whether the DNS domain should have been mit.edu or mit.com is a tossup.
It's interesting they actually got in bed with MS, who, in the parlance of that excellent book, Hard Drive, has been known in the industry for date rape when it comes to new technologies and ideas (after the visit to a sweet young software boutique to explore a partnering relationship no phone calls are returned the next morning, or else you get an offer you can't refuse.) MIT is traditionally pretty careful about extracting the maximum monetary benefit from the work of its faculty and students. It is unlike them to give in that much to anyone. It must figure the bird in the hand is worth it.
While MS is capable of stronger arm twisting than other corporations, note that Project Athena was funded by IBM and DEC and, no, most of those computers did not run MVS and VMS. While I haven't been there in years, I get the feeling the dorms are wired with Ethernet and more than a few Linux machines.
The real motivation for MS with its MIT partnership is to attach a large pipe fitting at the mouth of the Course VI (EE/CS) graduating class to get new talent. The old talent at MS probably gets burned out managing the complexity of 1e7 lines of code that is deliberately twisted into a Gordian Knot with inextricable connections and fragile tendrils leading back to revisions n-1, n-2, ... in order to impede competitors. Once those stock options become vested the reasons for staying in Redmond and putting in 80+ hour weeks begin to go away (FYIFV). Then the corporate management finds it's time to find new gelflings to write Word 2002.
Despite my lifestyle I still have two hands, fortunately.
Nevertheless, once I saw this device I seriously was thinking of getting one just for the improved efficiency of input, removing big clunky keyboard, and freeing up one hand for "other important tasks."
twiddler