If the system were to keep around the unedited original versions of each config file, then on update it could diff your edited version against that, creating a patch, which it could attempt to apply to the new version. (If the patch does not apply cleanly, of course, then you still have to hand-merge, but I suspect that in the majority of cases the lines the user has changed will not be the lines that have changed in the new version. For instance, in http.conf a lot of us only change a couple of basic things, like DocumentRoot and such, and for most upgrades the patch would apply cleanly, unless the app has made major changes to the config file (e.g., between Apache 1.x and 2.x). I have changed just one thing in wgetrc (to make it do passive ftp, since I'm behind a NAT gateway) -- that should apply cleanly almost no matter WHAT the wget people do to the next version of the config file, I would think.
I'm not entirely sure where the unedited originals should be kept; perhaps in a separate hierarchy of their own (e.g.,/etc-originals) or perhaps right beside the in-use config files, with a.orig suffix added or something.
Besides/etc, ~/.* should probably be handled much the same way.
> When Google first created its system, it worked well because the > internet wasn't as filled with people trying to manipulate the results.
No, that's the thing: it was and they were, but the techniques that had been developed to do so up to that point did not work so well with Google, because Google's type of ranking was different from the search engines that came before.
I think it's amazing that it took as long as it did for them to work out the systems that they have. Even now, Google is *still* harder for them to manipulate than AltaVista was before Google came around. (Okay, so it's easy to manipulate the results for a strange combination of obscure words that nobody normally puts together (e.g., nigritude ultramarine), but that's mostly because there are no naturally-good results for such searches. You just *try* to get your blog to come up first for "generic viagra".)
However, Google can't afford to be complacent; they need to actively work on their ranking algorithms and improve/adjust them continually, to prevent the SEO people from ever fully catching up.
> But every now and then a bully miscalculates, as we saw with SCO versus IBM.
That was more than just miscalculation as to the victim's resources. SCO knew very well what kind of resources IBM could bring to bear. There was something deeper going on in their corporate psyche than mere miscalculation.
> I worry about it all the time. My users constantly volunteer their passwords > when I don't ask for them.
You're lucky: your users know their passwords. If I tell my users that they need a password for something, they tell me they don't have a password, don't want a password, and that I have to fix it so they don't need one.
> You could still bring the system up with a boot floppy (or Knoppix, or > something else) and replace the administrator credential with one of your > choosing.
Ah, that's why we need encrypted filesystems; you enter the private key for the filesystem at boot time, and the system uses it whenever it accesses the disk, but it's only ever stored in RAM, not on disk, so after any reboot it has to be entered again. Then the attacker uses a hardware keyboard input recording device...
> Back in my day, we didn't even have buttons. We had to move the cursor
Move the cursor? You young whippersnappers have it easy. Time was, we didn't have a cursor, and if we wanted to insert a word, we had to retype the whole line -- and we liked it, because it was better than using punchcards. Why, before white-out was invented, we had to retype the whole page to make a change, but did we complain? No, we were happy we had carbon paper, so we didn't have to retype it twice! My great grandpappy carved his own quill pen, and he was just happy he didn't have to make his own India ink...
Re:Google: Fix the top post reply method
on
Gmail Goes Public
·
· Score: 1
> Where I've worked, people leave the entire thread at the bottom of an email > in case you later CC someone so that they can know what you're talking about.
You work with people who don't want to bother to think about what is relevent and what is not. Under no circumstances is keeping a lengthy thread around necessary just for someone to know what you're talking about, and even if someone later does really need to see the whole conversation, that's what archives are for.
> On usenet, on the other hand, interlacing threads works well because you > can always check previous articles in the thread.
That used to be the case, but these days too many people are using online readers, so that as soon as something expires off the server they cannot easily go back and read it -- and with Google groups' interface rapidly going into the privy of late, they may not be able to find it at all. With email, though, unless you've got some kind of draconian storage limit from hell as corporate policy, somebody's always going to have the whole thread around and can dig it up and pass it along, if need be -- normally, however, a three-sentence synopsis is a better to bring someone new into a conversation in the middle, especially in this decadent era of twelve-second attention spans when the probability of a coworker reading the entire thread you send them is next to nil anyhow.
It may be that the lower-limit on usernames is so that spammer-style brute-force attacks on the namespace can be better thwarted. Although, it doesn't do much about dictionary attacks, so maybe I'm misguessing.
Re:Google: Fix the top post reply method
on
Gmail Goes Public
·
· Score: 5, Insightful
> Bottom posting is for grizzled usenet hippies.
Bottom-posting (quoting the whole message and then putting your reply at the bottom) and top-posting (quoting the whole original message below your reply) are both cretinous and bad. The correct way to quote is interleaved, i.e., you quote a relevant excerpt, reply to it, then if necessary quote another relevant excerpt, reply to it, and so forth.
Gnus gets this right: it quotes the whole message (depending on how you have it set up) (except the signature (if it can tell where the signature starts)), but if you go to any point in the message and start typing, it breaks there and rewraps the quoted portions above and below, and your reply gets inserted at the proper place, unquoted, as a separate paragraph. Any parts of the quoted message you don't need to reply to, you're supposed to delete before sending. Gnus warns you if you try to send a message that's mostly quoted material and very little original response (though it'll let you do it if you insist).
But I don't suppose it's reasonable to hold a webmail interface to the standard of functionality set by Gnus.
Two colors are employed: lime green, and magenta. A script alternates the text and the background between these two colors, so that at any given point in time, the text and the background are never the same color, but both the background and the text flash back and forth between the two colors. The rate of switching the colors may be static or may vary. A plugin also may be used to add further animation to the page.
I guess maybe it's the combination of saturation and brightness together in the same color. A lot of sites use high-brightness colors (notably, evil blinding #FFFFFF has been all the rage for going on a decade now, for no good reason I can fathom), and then again there are high-saturation colors that are also popular (e.g., #0000FF or #FF0000 for links), but putting the two together is just pretty nasty. The phrase "Angry Fruit Salad" comes to mind.
> Well, I have been in the pilot project from the very beginning and there > are builds of OpenSolaris up and running. We have the source and are working > on a PowerPC port to the Open Desktop Workstation
That's all well and good, but it doesn't address the parent's point.
> Some of us want an OS that can run on 128 simultaneous processors as well as > one or four or twelve all with the same kernel. Not a cluster. One big > computer.
This does. I think Linux does do umpteen processors these days (though I could be getting confused; I have no personal experience with such large systems), but you're barking up the right general tree: the parent was probably thinking in terms of the desktop (as, indeed, I have a tendency to do myself), and as he notes, an open-source Solaris may indeed be too little too late for the desktop. Server space is quite another matter, though. Solaris is well respected there, especially for the relatively high-end stuff.
Solaris also has the best *name* of any operating system, ever:-)
I plan to experiment with Solaris a bit in VMWare, but in my case it's mostly because I'm the sort of person who experiments with sundry OSes just for the sake of experimenting. The high-end kind of scenerio you're talking about has no direct and immediate relevance to my life.
In that age range, your job doesn't have much appeal; no white-collar job does unless it has serious celebrity status (e.g., US President). The minute you start talking about sitting at a desk, there'll be a mass exodus to the booth where one of the other kids' parents is talking about their job at McDonald's. Now *that* job is cool.
If it were a junior high or high school career fair, that would be different; those kids (well, some of them) want to do white collar work when they get out of school. But in elementary school, they want to be firemen, astronauts, athletes, President,... or sell french fries. Software engineering is BORING to them, almost as boring as being a professional student.
Now, if you were a professional video game reviewer, that would be different.
The new improved command-line interface springs to mind. Compared to that, WinFS is small potatoes.
Oh, you meant for non-geeks? In that case, it's all about the new improved MSN icon on the desktop, and the nifty new wallpaper, and that sort of thing.
> For that kind of bang/buck, manufacturers might want to start bundling > Linux with Windows in a dual-boot configuration.
If they did that, they would have to pay full price for the OEM Windows install, instead of the special bulk price they get by agreeing to the contract with Microsoft that, among other things, prohibits them from doing that.
That provision was originally put in the contract when IBM and OS/2 were the big competitors Microsoft was worried about. It was subsequently used to prevent BeOS from being bundled in just the fashion you describe.
> Also Perl prgrammers tend to put more than one statment on a line by > convention.
As long as you don't overdo it *too* much, that can actually improve readability rather significantly, by allowing an entire function to fit on the screen at once that otherwise wouldn't, which makes a huge difference in how easy it is to read. Even in longer functions (though Perl functions longer than one screenful are IMO usually a symptom of bad design), fitting more context on the screen at once is a Good Thing(TM).
All I want is the features back that were dropped between 1.4 and 2.0. Most notably, in terms of the panel. The 1.2/1.4 panel is *significantly* more functional than the 2.x one has yet managed to become. Most notably, in 1.x I can have a tiny always-on-top clock panel, which I can drag to anyplace I want it. (I keep it just to the left of where the minimize/shade/etc buttons on a maximized window are, so that it covers up an empty section of titlebar.)
When is Gnome 2 going to get these features back?
Actually, the panel is the only part of Gnome that I really care about. I don't use the default window manager anyhow, and I *certainly* don't use that Nautilus junk. (Haven't needed a GUI file manager since tab completion was invented, and I don't need shortcut icons on the desktop either, because panel drawers are better, and have freed me from the need to obsessively minimize everything all the time; I haven't seen my wallpaper in days, and I don't miss it.) I don't use the web browser, because I have Firefox. I don't use Gnome Office, because I have OpenOffice. Really, the panel is the key feature I need from Gnome. (And it's the panel -- and its extremely useful drawers feature -- that keeps me from switching to KDE or something else.)
In summary, the panel is really important, so, please, please, can we have the 1.x panel features back? Until we get those, new Gnome releases are of no interest to me.
> But what if a large portion of the mountain were ejected
That's pretty much what happened in 1980. If you look at a picture of the mountain today, you'll see that the crater is fairly sizeable. It wasn't nearly that big before 1980. The immediate area had some very impressive mudslides and stuff as a result, and a whole lot of layers of sediment were deposited, and all kinds of interesting stuff. It's a fascinating eruption to study, because we have a lot more information about it than most of the other large ones in recorded history (e.g., Krakatoa), due to its location and recentness.
Some kinds of fixing are fairly non-controversial. Most notably, grammar and spelling fixes (barring international spelling differences) don't generally raise any significant ire, unless the article in question is in the middle of an edit war anyway. You can go around correcting comma splices and stuff like that to your heart's content, and never get in an edit war.
A Master's in Computer Science is seldom required, granted. So, get a degree in something else. Your BS in CS will go with lots of things. What else interests you besides programming? Security? Computer science seems to be a good combination with security, so why not get a degree related to that? Or is science more your thing? What kind of science? Biology? You could end up programming bio-science stuff, like DNA sequencing or who knows what. Physics? There's *lots* of physics stuff to do that involves computer science -- aircraft design, for instance. Meteorology goes well with computer programming too. Or maybe science isn't your thing. Maybe you'd rather get your Master's in Library Science, and get a job writing and supporting library automation systems... or you could go for math (which, incidentally, is almost as much fun as programming) and end up writing actuarial software or something like Mathematic or whatnot, or doing computer-based pure math research for a university, or just wind up as a professor. (Being a professor doesn't pay as well as some things, but the working conditions are okay, and the benefits are decent (e.g., decent hours, summers off, two weeks for Christmas, a week around Easter time... and your kids get *serious* tuition discounts if they have to pay at all...) so it's not altogether a bad way to go.) Or you could sell your soul to Catbert and get an MBA. There are tons of options. What field do you like?
For *you* it's not too small, and maybe for the people you've convinced. Not everyone has the same eyes. Their use of a lower resolution does not hurt anyone. Leave them be.
Slashdot Mirror
If the system were to keep around the unedited original versions of each config file, then on update it could diff your edited version against that, creating a patch, which it could attempt to apply to the new version. (If the patch does not apply cleanly, of course, then you still have to hand-merge, but I suspect that in the majority of cases the lines the user has changed will not be the lines that have changed in the new version. For instance, in http.conf a lot of us only change a couple of basic things, like DocumentRoot and such, and for most upgrades the patch would apply cleanly, unless the app has made major changes to the config file (e.g., between Apache 1.x and 2.x). I have changed just one thing in wgetrc (to make it do passive ftp, since I'm behind a NAT gateway) -- that should apply cleanly almost no matter WHAT the wget people do to the next version of the config file, I would think.
/etc-originals) or perhaps right beside the in-use config files, with a .orig suffix added or something.
/etc, ~/.* should probably be handled much the same way.
I'm not entirely sure where the unedited originals should be kept; perhaps in a separate hierarchy of their own (e.g.,
Besides
I want to be the first result on Google for the word "of". Who's with me? ...
We'll just use these simple techniques he's outlined, buy a few links,
> When Google first created its system, it worked well because the
> internet wasn't as filled with people trying to manipulate the results.
No, that's the thing: it was and they were, but the techniques that had been developed to do so up to that point did not work so well with Google, because
Google's type of ranking was different from the search engines that came before.
I think it's amazing that it took as long as it did for them to work out the
systems that they have. Even now, Google is *still* harder for them to
manipulate than AltaVista was before Google came around. (Okay, so it's
easy to manipulate the results for a strange combination of obscure words
that nobody normally puts together (e.g., nigritude ultramarine), but that's
mostly because there are no naturally-good results for such searches. You
just *try* to get your blog to come up first for "generic viagra".)
However, Google can't afford to be complacent; they need to actively work on
their ranking algorithms and improve/adjust them continually, to prevent the
SEO people from ever fully catching up.
> But every now and then a bully miscalculates, as we saw with SCO versus IBM.
That was more than just miscalculation as to the victim's resources. SCO knew very well what kind of resources IBM could bring to bear. There was something deeper going on in their corporate psyche than mere miscalculation.
> I worry about it all the time. My users constantly volunteer their passwords
> when I don't ask for them.
You're lucky: your users know their passwords. If I tell my users that they
need a password for something, they tell me they don't have a password, don't
want a password, and that I have to fix it so they don't need one.
> You could still bring the system up with a boot floppy (or Knoppix, or
> something else) and replace the administrator credential with one of your
> choosing.
Ah, that's why we need encrypted filesystems; you enter the private key for
the filesystem at boot time, and the system uses it whenever it accesses the
disk, but it's only ever stored in RAM, not on disk, so after any reboot it
has to be entered again. Then the attacker uses a hardware keyboard input
recording device...
> Back in my day, we didn't even have buttons. We had to move the cursor
Move the cursor? You young whippersnappers have it easy. Time was, we didn't
have a cursor, and if we wanted to insert a word, we had to retype the whole
line -- and we liked it, because it was better than using punchcards. Why,
before white-out was invented, we had to retype the whole page to make a
change, but did we complain? No, we were happy we had carbon paper, so we
didn't have to retype it twice! My great grandpappy carved his own quill pen,
and he was just happy he didn't have to make his own India ink...
> Where I've worked, people leave the entire thread at the bottom of an email
> in case you later CC someone so that they can know what you're talking about.
You work with people who don't want to bother to think about what is relevent
and what is not. Under no circumstances is keeping a lengthy thread around
necessary just for someone to know what you're talking about, and even if
someone later does really need to see the whole conversation, that's what
archives are for.
> On usenet, on the other hand, interlacing threads works well because you
> can always check previous articles in the thread.
That used to be the case, but these days too many people are using online
readers, so that as soon as something expires off the server they cannot
easily go back and read it -- and with Google groups' interface rapidly
going into the privy of late, they may not be able to find it at all.
With email, though, unless you've got some kind of draconian storage limit
from hell as corporate policy, somebody's always going to have the whole
thread around and can dig it up and pass it along, if need be -- normally,
however, a three-sentence synopsis is a better to bring someone new into a
conversation in the middle, especially in this decadent era of twelve-second
attention spans when the probability of a coworker reading the entire thread
you send them is next to nil anyhow.
It may be that the lower-limit on usernames is so that spammer-style brute-force
attacks on the namespace can be better thwarted. Although, it doesn't do much
about dictionary attacks, so maybe I'm misguessing.
> Bottom posting is for grizzled usenet hippies.
Bottom-posting (quoting the whole message and then putting your reply at the bottom) and top-posting (quoting the whole original message below your reply) are both cretinous and bad. The correct way to quote is interleaved, i.e., you quote a relevant excerpt, reply to it, then if necessary quote another relevant excerpt, reply to it, and so forth.
Gnus gets this right: it quotes the whole message (depending on how you have it set up) (except the signature (if it can tell where the signature starts)), but if you go to any point in the message and start typing, it breaks there and rewraps the quoted portions above and below, and your reply gets inserted at the proper place, unquoted, as a separate paragraph. Any parts of the quoted message you don't need to reply to, you're supposed to delete before sending. Gnus warns you if you try to send a message that's mostly quoted material and very little original response (though it'll let you do it if you insist).
But I don't suppose it's reasonable to hold a webmail interface to the standard of functionality set by Gnus.
Method for creating an eyesore:
Two colors are employed: lime green, and magenta. A script alternates the
text and the background between these two colors, so that at any given point
in time, the text and the background are never the same color, but both the
background and the text flash back and forth between the two colors. The
rate of switching the colors may be static or may vary. A plugin also may
be used to add further animation to the page.
I guess maybe it's the combination of saturation and brightness together in
the same color. A lot of sites use high-brightness colors (notably, evil
blinding #FFFFFF has been all the rage for going on a decade now, for no
good reason I can fathom), and then again there are high-saturation colors
that are also popular (e.g., #0000FF or #FF0000 for links), but putting the
two together is just pretty nasty. The phrase "Angry Fruit Salad" comes to
mind.
Those colour schemes are similar, in the sense that both of them make heavy
use of garish, clashing, high-saturation colours that DON'T GO TOGETHER.
> Well, I have been in the pilot project from the very beginning and there
:-)
> are builds of OpenSolaris up and running. We have the source and are working
> on a PowerPC port to the Open Desktop Workstation
That's all well and good, but it doesn't address the parent's point.
> Some of us want an OS that can run on 128 simultaneous processors as well as
> one or four or twelve all with the same kernel. Not a cluster. One big
> computer.
This does. I think Linux does do umpteen processors these days (though I
could be getting confused; I have no personal experience with such large
systems), but you're barking up the right general tree: the parent was
probably thinking in terms of the desktop (as, indeed, I have a tendency to
do myself), and as he notes, an open-source Solaris may indeed be too little
too late for the desktop. Server space is quite another matter, though.
Solaris is well respected there, especially for the relatively high-end stuff.
Solaris also has the best *name* of any operating system, ever
I plan to experiment with Solaris a bit in VMWare, but in my case it's mostly
because I'm the sort of person who experiments with sundry OSes just for the
sake of experimenting. The high-end kind of scenerio you're talking about
has no direct and immediate relevance to my life.
In that age range, your job doesn't have much appeal; no white-collar job
... or sell french fries. Software engineering is BORING
does unless it has serious celebrity status (e.g., US President). The minute
you start talking about sitting at a desk, there'll be a mass exodus to the
booth where one of the other kids' parents is talking about their job at
McDonald's. Now *that* job is cool.
If it were a junior high or high school career fair, that would be different;
those kids (well, some of them) want to do white collar work when they get
out of school. But in elementary school, they want to be firemen, astronauts,
athletes, President,
to them, almost as boring as being a professional student.
Now, if you were a professional video game reviewer, that would be different.
> What will be our reason to buy longhorn?
The new improved command-line interface springs to mind. Compared to that,
WinFS is small potatoes.
Oh, you meant for non-geeks? In that case, it's all about the new improved
MSN icon on the desktop, and the nifty new wallpaper, and that sort of thing.
> For that kind of bang/buck, manufacturers might want to start bundling
> Linux with Windows in a dual-boot configuration.
If they did that, they would have to pay full price for the OEM Windows install,
instead of the special bulk price they get by agreeing to the contract with
Microsoft that, among other things, prohibits them from doing that.
That provision was originally put in the contract when IBM and OS/2 were the
big competitors Microsoft was worried about. It was subsequently used to
prevent BeOS from being bundled in just the fashion you describe.
> Also Perl prgrammers tend to put more than one statment on a line by
> convention.
As long as you don't overdo it *too* much, that can actually improve readability
rather significantly, by allowing an entire function to fit on the screen at
once that otherwise wouldn't, which makes a huge difference in how easy it is
to read. Even in longer functions (though Perl functions longer than one
screenful are IMO usually a symptom of bad design), fitting more context on
the screen at once is a Good Thing(TM).
It can, of course, be overdone.
All I want is the features back that were dropped between 1.4 and 2.0. Most
notably, in terms of the panel. The 1.2/1.4 panel is *significantly* more
functional than the 2.x one has yet managed to become. Most notably, in 1.x
I can have a tiny always-on-top clock panel, which I can drag to anyplace I
want it. (I keep it just to the left of where the minimize/shade/etc buttons
on a maximized window are, so that it covers up an empty section of titlebar.)
When is Gnome 2 going to get these features back?
Actually, the panel is the only part of Gnome that I really care about. I
don't use the default window manager anyhow, and I *certainly* don't use that
Nautilus junk. (Haven't needed a GUI file manager since tab completion was
invented, and I don't need shortcut icons on the desktop either, because panel
drawers are better, and have freed me from the need to obsessively minimize
everything all the time; I haven't seen my wallpaper in days, and I don't
miss it.) I don't use the web browser, because I have Firefox. I don't
use Gnome Office, because I have OpenOffice. Really, the panel is the key
feature I need from Gnome. (And it's the panel -- and its extremely useful
drawers feature -- that keeps me from switching to KDE or something else.)
In summary, the panel is really important, so, please, please, can we have
the 1.x panel features back? Until we get those, new Gnome releases are of
no interest to me.
> Global warming causes Volcanoes. Just like it caused the Tsunami in
> Indonesia, the Kennedy assasination and male pattern baldness.
No, it's not global warming that causes baldness. It's the hole in the ozone.
HTH.HAND.
See for instance the before and after shots here.
> But what if a large portion of the mountain were ejected
That's pretty much what happened in 1980. If you look at a picture of the
mountain today, you'll see that the crater is fairly sizeable. It wasn't
nearly that big before 1980. The immediate area had some very impressive
mudslides and stuff as a result, and a whole lot of layers of sediment
were deposited, and all kinds of interesting stuff. It's a fascinating
eruption to study, because we have a lot more information about it than
most of the other large ones in recorded history (e.g., Krakatoa), due
to its location and recentness.
Some kinds of fixing are fairly non-controversial. Most notably, grammar and
spelling fixes (barring international spelling differences) don't generally
raise any significant ire, unless the article in question is in the middle of
an edit war anyway. You can go around correcting comma splices and stuff
like that to your heart's content, and never get in an edit war.
A Master's in Computer Science is seldom required, granted. So, get a degree
in something else. Your BS in CS will go with lots of things. What else
interests you besides programming? Security? Computer science seems to be
a good combination with security, so why not get a degree related to that?
Or is science more your thing? What kind of science? Biology? You could
end up programming bio-science stuff, like DNA sequencing or who knows what.
Physics? There's *lots* of physics stuff to do that involves computer
science -- aircraft design, for instance. Meteorology goes well with
computer programming too. Or maybe science isn't your thing. Maybe you'd
rather get your Master's in Library Science, and get a job writing and
supporting library automation systems... or you could go for math (which,
incidentally, is almost as much fun as programming) and end up writing
actuarial software or something like Mathematic or whatnot, or doing
computer-based pure math research for a university, or just wind up as a
professor. (Being a professor doesn't pay as well as some things, but the
working conditions are okay, and the benefits are decent (e.g., decent
hours, summers off, two weeks for Christmas, a week around Easter time...
and your kids get *serious* tuition discounts if they have to pay at all...)
so it's not altogether a bad way to go.) Or you could sell your soul to
Catbert and get an MBA. There are tons of options. What field do you like?
> But it's not too small.
For *you* it's not too small, and maybe for the people you've convinced.
Not everyone has the same eyes. Their use of a lower resolution does not
hurt anyone. Leave them be.