In high school we were supposed to write an essay on some book we read -
I can't even remember what it was. At a loss for ideas, I recalled some
commentary on something else I read elsewhere, and wrote about
Apollonian/Dionysian conflict, primitive/cultured duality,
natural/artificial, making connections between the characters in the
book and the woodsman in Walden and Billy Budd in Billy
Budd. It was written quickly - an in-class quiz actually. It made
little sense to me but it sounded good. The teacher hailed it as
brilliant; she read it to the other classes, and
everyone in my grade knew about it. Two cute girls I barely knew, from one of those other classes,
came up to me and told me how great my essay was. I guess the lesson
is, for those of you still in high school, English teachers love
dualities and connections to characters in other books that are not the
topic of the essay.
After all, someone went to the effort to make a RAID 0 array using floppy disks!
At first I did a double take and figured it
would be some awkward, jury-rigged proof-of-concept
with ugly wires all over the place, but the obvious
googling came up with this:
http://ohlssonvox.8k.com/fdd_raid.htm with
great pictures.
It's beautifully executed on OS X and
very pretty to look at. Amazing!
Re:Maybe Yahoo is changing for a reason
on
Yahoo to Dump Google
·
· Score: 2, Interesting
A lot of commercial sites can be cut out by adding "site:org" to the
search. For a lot of things that will get you the
no-nonsense facts you used to get in
the old days.
Unfortunately it's a matter of time until all the sleazy huckster sites add a.org alias, and it's already happening. But for right now it kind of
works - take advantage while it lasts.
The problem is not so much the lack of caffiene in your system as it
is the toxins it leaves behind, and those need to be flushed out.
Not to diminish your advice, but precisely what chemicals are these
"toxins" you are referring to? Can you measure them with a blood assay?
Perhaps there are such things, but I see the phrase "flushing out
toxins" most often in the context of homeopathic remedies and such whose
main benefit is to enrich their promoters.
My guess would be simply that your brain has been accustomed to
expect caffeine and it takes a while for it to readjust. But then I'm
no expert.
These kind of things were possible in older systems
(e.g. Mac OS 9 and below froze every other application while you were
dragging a window around), but modern operating systems cannot work that
way anymore.
It is too bad they weren't designed differently, so that they could
work this way. With only a few realtime exceptions in a desktop machine
- MP3 players, CDROM burners (which I think are flawed this way; maybe
in the future RAM will be so cheap the necessary minimum data can be
pre-buffered locally before writing) - there is no theoretical reason
this has to be.
You don't want downloads to stop when they're in the
background, firewalls to stop working because they're never the
frontmost application etc...
I think that the way most timeouts are designed is fundamentally
flawed. A good timeout protocol will always have a way for the client
(or server) to say, "I'm busy and can't send more data now, but I'm
still alive, so don't disconnect me." Anytime there is a hard timeout,
the number must be selected according to human judgment, and there will
always be unanticipated situations where it is too short or too long.
But I guess it's too late to change this in many existing designs.
BTW my original post was about desktops, not servers, which have
different needs. But if a firewall is local to the desktop, then in
scheme I described the high-priority process will become "idle" as soon as
it's not waiting for CPU or disk, so if it's waiting for a packet from
the firewall, then the firewall would indeed start running long enough
to deliver that packet and un-idle the high-priority process.
Even if a process is at a low priority...if it comes on the CPU, it has
the potential to trash memory and slow the works down.
Exactly, this is the problem. I want the ability the force all
(non-critical, non-realtime) processes not associated with whatever
window is currently in focus to freeze completely, until the process
associated with the window in focus is idle, waiting for neither CPU
nor disk. Or until the user brings another window into focus, in which
case this behavior would immediately shift to the new window.
In other words I am suggesting a system option to do this: if the
process associated with the window that's in focus is not in the idle
state, make everything else freeze. Or even this: an option to freeze
everything else unless the process associated with the window that's in
focus has been idle for at least x seconds (user settable). This latter
would prevent swapping out while the user is typing.
It seems to me this would greatly enhance the user's perception of
speed on machines with limited memory, even if it means some background
tasks might end up taking longer because they lose out on some otherwise
idle CPU time.
The benefits of preemption are great enough that the kernel developers
decided to make the kernel itself preemptible. This allows a kernel
task such as disk I/O to be preempted, for instance, by a keyboard
event. This allows the system to be more responsive to the user's
demands. The 2.6 kernel is able to efficiently manage its own tasks in
addition to user processes.
One of the most time-wasting, counterproductive things for me on
Windows, and to a lesser extent on Linux, is that background tasks can
essentially take over the computer because of disk thrashing. Sometimes
(on Windows XP) I've seen it literally take minutes to bring a window to
the front and repaint it while the disk is thrashing doing
god-knows-what in the background. It is impossible to get any serious
work done when that is happening. Setting my process to highest priority
(on Windows) doesn't seem to help when disk thrashing is involved. I can understand how
it might take a few seconds to swap another
process out and bring mine in, but not
minutes.
Now I don't know exactly what's happening behind the scenes, but as
soon as I click on a window to bring it in focus, I want
everything on the computer to be devoted to doing just that until
its done, regardless of what disk blocks from other processes are in the
queue. My disk block requests should go in front of all of the others
unconditionally. In other words I should see essentially no more
latency than if the computer were completely idle in the background,
other than just the minimum delay required to bring it back into memory
if it's swapped out. For all I care just completely freeze all other
processes on the computer until my request is finished, or at least
give me the option to specify that it be done that way. My productivity
should override everything else on my own
desktop/workstation computer.
So, educate me about what I may be overlooking here, and will 2.6
help this? I mean for Linux of course, not Windows, which may be a lost cause (:
Does anyone remember those cheap transformerless adapters you could buy
around 1970? They replaced a 9-V transistor radio battery and
were the same size, with a cord coming out that plugged into 110V.
I opened one up, and there were 4 parts: a capacitor in series with
the AC line, a rectifier diode, a 9V zener diode, and an electrolytic
filter capacitor. I wonder if anyone got electrocuted by them, and
when they were (as I assume) banned.
More and more parents are going to be pressured into keeping 24 hour
tabs on their teenagers, due to fear of lawsuits if their kids get in
trouble as well as fear due to media-hyped crime stories. I see this as
a bad thing. Kids will grow up used to constant 24 hour surveillance,
fully prepared to become zombies in the Big Brother society of the
future where their every movement will be tracked.
I'm sorry, but an important part of growing up is getting at least a
taste of true freedom and yes, sometimes the risk that it entails. .
When I was a teenager I probably did a few things my parents wouldn't
have approved of, and I that was an important part of my experience.
I can't imagine imposing this on my own teenager, except (1) when he
actively wants it, if say he goes into a strange part of town, or (2) as
punishment if he gets into trouble - part of the punishment might be
that he would be monitored for the next two months or whatever.
If he
wants to be monitored all the time,
I hope they keep it low-budget at the beginning so it can survive. The
problem with a lot of new ventures (the.com bust providing a notable
collection of examples) is that they grossly overestimate the potential
market, spend lavishly on huge productions, then boom - the money's gone
and they disappear. On the other hand if they start small and grow
slowly as the market, if there is one, materializes, they can survive.
There's a huge amount of classic material out there, and how much can
the royalties on Feynman's lectures be anyway. There's lots and lots of
stuff I'd love to see (again in many cases) that already exists. I
don't care if it's black and white. Just let these (presumably cheap to
license) classic things run 24 hrs and see what happens. Anyway I wish
them luck.
I'd settle for an Uberdevice that replaces all my AC adapters. These
annoying things tend to hog 2 or 3 outlets due to their size, and it
seems such a waste to have reconvert the AC over and over again for
every device. They tend to become disassociated with their parent units
over time, and my house is littered with orphan AC adapters that I don't
know whether to throw away or not because I no longer know what they
were for. And then there are the units
with different milliamps but the same
kind of plug, that you don't know what goes where
(after a cleanup to untangle all the
wires)
without consulting the manuals, which of course have disappeared
(another subject). And let's not get into orphaned and missing
remotes...
Several years ago I worked on a 1-million+ line ERP app which was maybe 20% customer-modded - i.e.
very heavily.
All mods, i.e. differences from the original, were carefully documented
(in the form of extensive comments in the code, as well as keeping the
original code in a canonical commented-out format inside the source - in
a fashion that the original code could in principal be reproduced
algorithmically with a program if desired; in fact this was tested for
as part of the QA process).
With the help of diff3-like in-house tools, we kept the modified code
in sync with vendor updates in a relatively straightforward fashion.
Code clashes were automatically identified and carefully analyzed - it
was curious how the vendor often fixed the same bug we did, usually in a
slightly different way. Of course after a vendor update there was
extensive testing, but there were staff people dedicated to this.
My point is that with careful discipline it is possible to keep code
in sync. It's not necessarily cheap, but good software tools to help
with the process can make it surprisingly efficient.
In English we don't say "dollars hundred" but we
write "$100". Yet in our word processors we don't
type "1-0-0-$" (the way we say it), with the
word processor momentarily switching into
right-to-left mode when the "$" is struck
(although there may be some obscure autocorrect
setting for this that believe me, if I ever encountered it, I
would immediately turn it off). Perhaps it's because
I learned touch typing on a real typewriter, but
when I type I want the letters to come out in the
order I type them, period. Mentally,
when I think verbally "ten dollars", my fingers
"feel" the character sequence "$-1-0". It just
happens. As soon as some
wizard starts second-guessing me and changing the
order of things, it interrupts my flow and I
become less efficient, going into hunt-and-peck
mode with the arrow keys to go back to try to
undo what the wizard did, and when the wizard
insists on redoing what I just undid, spending the
time to find the setting for that wizard so I
can kill the damn thing.
[Balmer] questioned the notion that the open source's community approach
to fixing problems was superior to Microsoft's. "Why should code
submitted randomly by some hacker in China and distributed by some open
source project, why is that, by definition, better?"
That should have been, "terrorist hacker in China."
Perhaps combining these various pools of skill (which I do not
minimalize or trivialize for a moment) and supporting some real helpful
and Linux-promoting projects would be a better use of resources?
For instance, if Linux is to be a real competitor to Windows, how about
using these skills to build simple distros and simple methods of
installing and uninstalling apps on them that do not require arcane
command line utilities and other tasks. Just insert the CD or click on
the download and be done.
No, the skills needed to hack drivers are quite different, and a lot
rarer I believe, than those needed to "build simple distros". The
latter needs to be done, but it requires a different mindset. Leave the
driver hackers alone and let them keep hacking drivers - and thank god
for them.
(granparent poster:)
>
People make free drivers... Everyone wins when a device driver comes
out.
This is very true, the more hardware is supported by Linux the better.
However, we are not talking about device drivers here, we are talking
about embedding Linux into various devices.
There are lots of things you do as an academic exercise that end up
paying off in the end, both expectedly and unexpectedly. This is why
large corps. fund R&D, and it is why experimenting with
embedding Linux into various
devices for its own sake is to be encouraged.
(parent poster:)
Making different cards for different countries is going to be expensive.
So perhaps legal compliance can be handled in the driver! This is a
good idea, but people like myself want our drivers open so we can tinker
with them, use them in monitor mode, etc. The problem is, if we can do
these things, we can usually unlock the other foreign nation specific
features of the card, e.g. running your wifi net on channel 13 to escape
detection by your average scanner.
And this creates headaches for the vendor, because suddenly their FCC or
whatever compliance is called into question, and hence the legality of
their product. Not a nice place to be, business wise.
Sticky situation, and I'm not sure what the best solution is.
I understand what you're trying to say, but anyone can create FCC
havoc with a cheap transistor, a couple of capacitors, a resistor, a
battery, and a length of wire wrapped in a coil. Yet you can buy these
openly at any Radio Shack. So why should open drivers be any different?
What I hate, fear, and think is wrong, is a government mentality that says
they are different, thereby putting the manufacturers at risk of having
losing product certification if they release open drivers. This is not
the solution to the FCC problem. The solution is to crack down on
people actually generating disrupting EM interference (and it's usually
not so hard to detect and locate them when they do).
The danger I see in this kind of mentality is that it tends to
take a life of its own, like the mindset that
resulted in the DMCA, and eventually you may not
be able to buy a transistor at Radio Shack.
Perfect code is possible, but most customers are not willing to pay the
price.
Why should the customers pay the price (specifically in Microsoft's
case)? Isn't this MS's responsibility? Or are you advocating that MS
just sit on its $50B in the bank while its customers continue to suffer
damages from inumerable security problems? My guess is if they spent just
1/3 of that on getting rid of bugs, they would come pretty close to
perfection. I mean, that is an obscene amount of money.
I remember when Google provided an amazing
treasure trove of genuinely useful medical information.
But these days the spammers are really screwing
it up. For example, suppose you are in
serious need of medical information on "tramadol withdrawal" (exact
phrase). Essentially 100% of the 6000+ matches are, you guessed it,
spammer links to buy more tramadol. Useful medical information
has become almost impossible to find, unless it's some obscure thing there's no on-line
pharmacy pill for.
Don't bother going to college. In fact, if you're in school drop out
now regardless of your grade level. If you have kids, pull them out or
don't send them to school to begin with. There is after all, no value
in any of the information that will be learned.
You're not paying for the information when you go to school. You're
paying for a service - the service of having it taught to you. The
information itself can be free; just go to the library.
"Argument list too long" is kind of like the buffer overflow of the
scripting world. It is one of the worst Unix violators of "16. Rule of
Extensibility: Design for the future, because it will be here sooner
than you think."
I have been bitten by this several times over the years on various
Unices, where a script that has been running fine will one day break
when another file is added to a directory. Of course it can be fixed
with xargs etc. but if it is a complex, otherwise stable script written
by someone else, finding and fixing it can be time consuming and incurs
the risk of introducing bugs.
For example, suppose you have an otherwise stable, reliable a script
that processes say nightly invoices on a production machine, one file
per invoice. But the script writer forgets to use xargs. It runs
successfully for many years, then one day the company is "too"
successful - oops.
You can argue that scripts should be written right in the first place
to anticipate the problem, but the reality is that many aren't. And
there's no easy way to test for it other than tediously analyzing the
script line by line and making a judgment whether or not a particular
wildcard expansion will ever be a problem in the future.
I simply don't understand why, when the wildcard expansion limit is
reached, the underlying software can't just allocate more memory as
needed. I don't even care how slow or inefficient it is, I just don't
want it to blow up.
you left out the w's - no self-respecting ms program would even try to forbid a virus to get attached to itself.
Indeed, several years ago at least, when I last looked, MS Word required write access to
its DLL(s?) on Windows NT. I don't know when or if
this *ahem* "minor security issue" was
addressed.
Most of these are derived units. There are only 7 base units from
which all the others can be derived. Here is one possibility:
Units: metric Imperial
Length: meter furlong
Mass: kg stone
Time: second fortnight
Charge: coulomb franklin
Temperature: centigrade farenheit
Illuminance: lux foot-candle
Quantity: mole gross
And yes, the franklin is a precisely defined unit: there are exactly
2997.92 franklins in a nanocoulomb. Look it up sometime to improve you
education. While I'm not an expert at its history, I would imagine
it has something to do with a standardized kite connected to a Leiden jar.
Some of these are definitely more practical for everyday use. I
mean, if you just want 12 dozen pencils, quick: how many moles would
that be? Really, it depends on what your job is. For example, a
computer scientist would find the beowulf cluster to be a more practical
unit than a mole.
3 can be obtained by a simple whois. Reverse track the relay
servers, the website, do a whois on domain names of sending servers and
you have this condition met.
Ah, do I hear the wheels of progress already working on the next
Mozilla
extension? Imagine gleefully watching the California ones carefully
plucked out of your Junk folder, optionally with PDF files of the
necessary court papers ready to print and file.
This really happened: My girlfriend called me in a panic this morning because she dropped her
IBM(!) laptop about 3 feet, and it will no longer turn on. I haven't
seen it yet but it doesn't seem promising - the wireless PCMCIA card
(with the protruding antenna part) now "looks a little askew". So I
guess that's what it landed on, and it was probably crunched into the
computer, kind of like when your thigh bones crunch through your hip and
into your abdomen if you jump from an airplane into the ocean and land
on your heels. Oh well. IBM told her that for $35 they will tell her
how much it will cost to fix.
Wait until I tell her that this very same morning IBM announced laptop
airbags! OK, just for the hard drive, but still. I bet she'll get a
kick out of that. Well, actually, probably not, given her plight.
I posted
this reply to an earlier story
and will repeat the relevant information.
I've tried dozens of different kinds of pens over my lifetime, and the
one that I've settled on and now insist on is the inexpensive Pilot
EasyTouch Medium Point ball-point (the Fine Point is good too, but not
quite as smooth). It is the smoothest writing instrument I've found,
whether ball-point, roller-ball, gel, fountain pen, or whatever. And it
always just seems to work; it doesn't dry on me and require those
scribbles to get the ink flowing after several days of non-use, like
other ball-points. Strangely it doesn't seem to be a standard stock
item and I have to special order it from Staples. The blue color seems
slightly smoother than red or black, but that may be subjective.
Slashdot Mirror
In high school we were supposed to write an essay on some book we read - I can't even remember what it was. At a loss for ideas, I recalled some commentary on something else I read elsewhere, and wrote about Apollonian/Dionysian conflict, primitive/cultured duality, natural/artificial, making connections between the characters in the book and the woodsman in Walden and Billy Budd in Billy Budd. It was written quickly - an in-class quiz actually. It made little sense to me but it sounded good. The teacher hailed it as brilliant; she read it to the other classes, and everyone in my grade knew about it. Two cute girls I barely knew, from one of those other classes, came up to me and told me how great my essay was. I guess the lesson is, for those of you still in high school, English teachers love dualities and connections to characters in other books that are not the topic of the essay.
At first I did a double take and figured it would be some awkward, jury-rigged proof-of-concept with ugly wires all over the place, but the obvious googling came up with this: http://ohlssonvox.8k.com/fdd_raid.htm with great pictures. It's beautifully executed on OS X and very pretty to look at. Amazing!
A lot of commercial sites can be cut out by adding "site:org" to the search. For a lot of things that will get you the no-nonsense facts you used to get in the old days. Unfortunately it's a matter of time until all the sleazy huckster sites add a .org alias, and it's already happening. But for right now it kind of
works - take advantage while it lasts.
Not to diminish your advice, but precisely what chemicals are these "toxins" you are referring to? Can you measure them with a blood assay? Perhaps there are such things, but I see the phrase "flushing out toxins" most often in the context of homeopathic remedies and such whose main benefit is to enrich their promoters.
My guess would be simply that your brain has been accustomed to expect caffeine and it takes a while for it to readjust. But then I'm no expert.
These kind of things were possible in older systems (e.g. Mac OS 9 and below froze every other application while you were dragging a window around), but modern operating systems cannot work that way anymore.
It is too bad they weren't designed differently, so that they could work this way. With only a few realtime exceptions in a desktop machine - MP3 players, CDROM burners (which I think are flawed this way; maybe in the future RAM will be so cheap the necessary minimum data can be pre-buffered locally before writing) - there is no theoretical reason this has to be.
You don't want downloads to stop when they're in the background, firewalls to stop working because they're never the frontmost application etc...
I think that the way most timeouts are designed is fundamentally flawed. A good timeout protocol will always have a way for the client (or server) to say, "I'm busy and can't send more data now, but I'm still alive, so don't disconnect me." Anytime there is a hard timeout, the number must be selected according to human judgment, and there will always be unanticipated situations where it is too short or too long. But I guess it's too late to change this in many existing designs.
BTW my original post was about desktops, not servers, which have different needs. But if a firewall is local to the desktop, then in scheme I described the high-priority process will become "idle" as soon as it's not waiting for CPU or disk, so if it's waiting for a packet from the firewall, then the firewall would indeed start running long enough to deliver that packet and un-idle the high-priority process.
Exactly, this is the problem. I want the ability the force all (non-critical, non-realtime) processes not associated with whatever window is currently in focus to freeze completely, until the process associated with the window in focus is idle, waiting for neither CPU nor disk. Or until the user brings another window into focus, in which case this behavior would immediately shift to the new window.
In other words I am suggesting a system option to do this: if the process associated with the window that's in focus is not in the idle state, make everything else freeze. Or even this: an option to freeze everything else unless the process associated with the window that's in focus has been idle for at least x seconds (user settable). This latter would prevent swapping out while the user is typing.
It seems to me this would greatly enhance the user's perception of speed on machines with limited memory, even if it means some background tasks might end up taking longer because they lose out on some otherwise idle CPU time.
The benefits of preemption are great enough that the kernel developers decided to make the kernel itself preemptible. This allows a kernel task such as disk I/O to be preempted, for instance, by a keyboard event. This allows the system to be more responsive to the user's demands. The 2.6 kernel is able to efficiently manage its own tasks in addition to user processes.
One of the most time-wasting, counterproductive things for me on Windows, and to a lesser extent on Linux, is that background tasks can essentially take over the computer because of disk thrashing. Sometimes (on Windows XP) I've seen it literally take minutes to bring a window to the front and repaint it while the disk is thrashing doing god-knows-what in the background. It is impossible to get any serious work done when that is happening. Setting my process to highest priority (on Windows) doesn't seem to help when disk thrashing is involved. I can understand how it might take a few seconds to swap another process out and bring mine in, but not minutes.
Now I don't know exactly what's happening behind the scenes, but as soon as I click on a window to bring it in focus, I want everything on the computer to be devoted to doing just that until its done, regardless of what disk blocks from other processes are in the queue. My disk block requests should go in front of all of the others unconditionally. In other words I should see essentially no more latency than if the computer were completely idle in the background, other than just the minimum delay required to bring it back into memory if it's swapped out. For all I care just completely freeze all other processes on the computer until my request is finished, or at least give me the option to specify that it be done that way. My productivity should override everything else on my own desktop/workstation computer.
So, educate me about what I may be overlooking here, and will 2.6 help this? I mean for Linux of course, not Windows, which may be a lost cause (:
Here's a previous rant of mine on AC adapters.
I'm sorry, but an important part of growing up is getting at least a taste of true freedom and yes, sometimes the risk that it entails. . When I was a teenager I probably did a few things my parents wouldn't have approved of, and I that was an important part of my experience.
I can't imagine imposing this on my own teenager, except (1) when he actively wants it, if say he goes into a strange part of town, or (2) as punishment if he gets into trouble - part of the punishment might be that he would be monitored for the next two months or whatever. If he wants to be monitored all the time,
I hope they keep it low-budget at the beginning so it can survive. The problem with a lot of new ventures (the .com bust providing a notable
collection of examples) is that they grossly overestimate the potential
market, spend lavishly on huge productions, then boom - the money's gone
and they disappear. On the other hand if they start small and grow
slowly as the market, if there is one, materializes, they can survive.
There's a huge amount of classic material out there, and how much can
the royalties on Feynman's lectures be anyway. There's lots and lots of
stuff I'd love to see (again in many cases) that already exists. I
don't care if it's black and white. Just let these (presumably cheap to
license) classic things run 24 hrs and see what happens. Anyway I wish
them luck.
I'd settle for an Uberdevice that replaces all my AC adapters. These annoying things tend to hog 2 or 3 outlets due to their size, and it seems such a waste to have reconvert the AC over and over again for every device. They tend to become disassociated with their parent units over time, and my house is littered with orphan AC adapters that I don't know whether to throw away or not because I no longer know what they were for. And then there are the units with different milliamps but the same kind of plug, that you don't know what goes where (after a cleanup to untangle all the wires) without consulting the manuals, which of course have disappeared (another subject). And let's not get into orphaned and missing remotes...
Several years ago I worked on a 1-million+ line ERP app which was maybe 20% customer-modded - i.e. very heavily.
All mods, i.e. differences from the original, were carefully documented (in the form of extensive comments in the code, as well as keeping the original code in a canonical commented-out format inside the source - in a fashion that the original code could in principal be reproduced algorithmically with a program if desired; in fact this was tested for as part of the QA process).
With the help of diff3-like in-house tools, we kept the modified code in sync with vendor updates in a relatively straightforward fashion. Code clashes were automatically identified and carefully analyzed - it was curious how the vendor often fixed the same bug we did, usually in a slightly different way. Of course after a vendor update there was extensive testing, but there were staff people dedicated to this.
My point is that with careful discipline it is possible to keep code in sync. It's not necessarily cheap, but good software tools to help with the process can make it surprisingly efficient.
In English we don't say "dollars hundred" but we write "$100". Yet in our word processors we don't type "1-0-0-$" (the way we say it), with the word processor momentarily switching into right-to-left mode when the "$" is struck (although there may be some obscure autocorrect setting for this that believe me, if I ever encountered it, I would immediately turn it off). Perhaps it's because I learned touch typing on a real typewriter, but when I type I want the letters to come out in the order I type them, period. Mentally, when I think verbally "ten dollars", my fingers "feel" the character sequence "$-1-0". It just happens. As soon as some wizard starts second-guessing me and changing the order of things, it interrupts my flow and I become less efficient, going into hunt-and-peck mode with the arrow keys to go back to try to undo what the wizard did, and when the wizard insists on redoing what I just undid, spending the time to find the setting for that wizard so I can kill the damn thing.
That should have been, "terrorist hacker in China."
2. Bet half one way, half the other. Discard the half that loses.
3. n--
4. If n>0, goto step 2.
5. Profit!!!
Perhaps combining these various pools of skill (which I do not minimalize or trivialize for a moment) and supporting some real helpful and Linux-promoting projects would be a better use of resources?
For instance, if Linux is to be a real competitor to Windows, how about using these skills to build simple distros and simple methods of installing and uninstalling apps on them that do not require arcane command line utilities and other tasks. Just insert the CD or click on the download and be done.
No, the skills needed to hack drivers are quite different, and a lot rarer I believe, than those needed to "build simple distros". The latter needs to be done, but it requires a different mindset. Leave the driver hackers alone and let them keep hacking drivers - and thank god for them.
(granparent poster:)
> People make free drivers ... Everyone wins when a device driver comes
out.
This is very true, the more hardware is supported by Linux the better. However, we are not talking about device drivers here, we are talking about embedding Linux into various devices.
There are lots of things you do as an academic exercise that end up paying off in the end, both expectedly and unexpectedly. This is why large corps. fund R&D, and it is why experimenting with embedding Linux into various devices for its own sake is to be encouraged.
(parent poster:)
Making different cards for different countries is going to be expensive. So perhaps legal compliance can be handled in the driver! This is a good idea, but people like myself want our drivers open so we can tinker with them, use them in monitor mode, etc. The problem is, if we can do these things, we can usually unlock the other foreign nation specific features of the card, e.g. running your wifi net on channel 13 to escape detection by your average scanner.
And this creates headaches for the vendor, because suddenly their FCC or whatever compliance is called into question, and hence the legality of their product. Not a nice place to be, business wise.
Sticky situation, and I'm not sure what the best solution is.
I understand what you're trying to say, but anyone can create FCC havoc with a cheap transistor, a couple of capacitors, a resistor, a battery, and a length of wire wrapped in a coil. Yet you can buy these openly at any Radio Shack. So why should open drivers be any different? What I hate, fear, and think is wrong, is a government mentality that says they are different, thereby putting the manufacturers at risk of having losing product certification if they release open drivers. This is not the solution to the FCC problem. The solution is to crack down on people actually generating disrupting EM interference (and it's usually not so hard to detect and locate them when they do).
The danger I see in this kind of mentality is that it tends to take a life of its own, like the mindset that resulted in the DMCA, and eventually you may not be able to buy a transistor at Radio Shack.
Why should the customers pay the price (specifically in Microsoft's case)? Isn't this MS's responsibility? Or are you advocating that MS just sit on its $50B in the bank while its customers continue to suffer damages from inumerable security problems? My guess is if they spent just 1/3 of that on getting rid of bugs, they would come pretty close to perfection. I mean, that is an obscene amount of money.
I remember when Google provided an amazing treasure trove of genuinely useful medical information. But these days the spammers are really screwing it up. For example, suppose you are in serious need of medical information on "tramadol withdrawal" (exact phrase). Essentially 100% of the 6000+ matches are, you guessed it, spammer links to buy more tramadol. Useful medical information has become almost impossible to find, unless it's some obscure thing there's no on-line pharmacy pill for.
You're not paying for the information when you go to school. You're paying for a service - the service of having it taught to you. The information itself can be free; just go to the library.
I have been bitten by this several times over the years on various Unices, where a script that has been running fine will one day break when another file is added to a directory. Of course it can be fixed with xargs etc. but if it is a complex, otherwise stable script written by someone else, finding and fixing it can be time consuming and incurs the risk of introducing bugs.
For example, suppose you have an otherwise stable, reliable a script that processes say nightly invoices on a production machine, one file per invoice. But the script writer forgets to use xargs. It runs successfully for many years, then one day the company is "too" successful - oops.
You can argue that scripts should be written right in the first place to anticipate the problem, but the reality is that many aren't. And there's no easy way to test for it other than tediously analyzing the script line by line and making a judgment whether or not a particular wildcard expansion will ever be a problem in the future.
I simply don't understand why, when the wildcard expansion limit is reached, the underlying software can't just allocate more memory as needed. I don't even care how slow or inefficient it is, I just don't want it to blow up.
Indeed, several years ago at least, when I last looked, MS Word required write access to its DLL(s?) on Windows NT. I don't know when or if this *ahem* "minor security issue" was addressed.
Units: metric Imperial
Length: meter furlong
Mass: kg stone
Time: second fortnight
Charge: coulomb franklin
Temperature: centigrade farenheit
Illuminance: lux foot-candle
Quantity: mole gross
And yes, the franklin is a precisely defined unit: there are exactly 2997.92 franklins in a nanocoulomb. Look it up sometime to improve you education. While I'm not an expert at its history, I would imagine it has something to do with a standardized kite connected to a Leiden jar.
Some of these are definitely more practical for everyday use. I mean, if you just want 12 dozen pencils, quick: how many moles would that be? Really, it depends on what your job is. For example, a computer scientist would find the beowulf cluster to be a more practical unit than a mole.
Ah, do I hear the wheels of progress already working on the next Mozilla extension? Imagine gleefully watching the California ones carefully plucked out of your Junk folder, optionally with PDF files of the necessary court papers ready to print and file.
Wait until I tell her that this very same morning IBM announced laptop airbags! OK, just for the hard drive, but still. I bet she'll get a kick out of that. Well, actually, probably not, given her plight.
I've tried dozens of different kinds of pens over my lifetime, and the one that I've settled on and now insist on is the inexpensive Pilot EasyTouch Medium Point ball-point (the Fine Point is good too, but not quite as smooth). It is the smoothest writing instrument I've found, whether ball-point, roller-ball, gel, fountain pen, or whatever. And it always just seems to work; it doesn't dry on me and require those scribbles to get the ink flowing after several days of non-use, like other ball-points. Strangely it doesn't seem to be a standard stock item and I have to special order it from Staples. The blue color seems slightly smoother than red or black, but that may be subjective.
Somebody else here likes the EasyTouch also.