I'll even go out on a limb and suggest that the closer Gnome moves in the KDE's direction, the more usable it gets, and frankly, I don't have a problem with that.
Personally, I think they're both heading in the
wrong direction. Both are a long way from where
they should be, and neither are remotely well
designed or thought out.
Gnome2 boots about 30% faster than KDE3. They're both patheticly slow in that department, though
Agreed, which is one of the reasons I use neither.
Since fvwm2 does everything I could possibly want
(including many things I can't do with either
GNOME or KDE), and is 20 times faster, why would
I consider switching? Eye candy does not a good
desktop experience make...
Now, the problem I have is that I own three of these, Acorn Electrons. Can I simultaneously use an Electron and Positron, or am I going to end up annihilating something?
I think you're safe. I too had (actually, still
have) an Electron. I also had the Micro Power
Positron
game. So far, I haven't annihilated anything
(well, apart from a few alien invaders anyway).
Just type: echo "alias ls='rm -rf'" >> ~/.bash_profile
At University, if someone left themselves logged
into a terminal, we'd place a file named "*"
in their home directory. Much hilarity ensued
when they typed rm * to remove it.
You know you never have written an email to this adress with this subject.
Would you really click on this attachment??
It goes like this. The mail hits our company
yesterday morning at 10:58. By 11:00 I've sent a
company wide mail out telling people that it's a
virus that's slipped past our scanner, and not to
open it. At 11:02 I get apologetic messages from
those who had already done so -- "I thought it was
someone sending me something", "It was just a zip file", "I didn't know".
Yes you did, you morons!
I've told you enough
times! You will never teach people not to do this.
People are stupid.
Re:How does FLAC compares to others?
on
Phish Moves To FLAC
·
· Score: 5, Informative
I wonder how FLAC compares to other compression methods (namely mp3 and ogg) in terms of quality and size...
FLAC is lossless, which means it is CD quality.
Literally. It will be a bit-for-bit perfect
representation of what you'd get on the CD. As part
of the tradeoff, you get larger filesizes. FLAC will
typically give 2:1 compression, compared to the 10:1
you're likely to achieve with MP3 or Ogg Vorbis, so
your files will be around 5 times larger.
Also, Ogg is a container format, not a
compression method. Ogg Vorbis is their flagship
lossy audio compression scheme. Note, however,
that FLAC is migrating to Ogg, so in future,
FLAC files will come with a.ogg extension.
fast put through rates,
high bandwidth through the router,
Myth. My ISP runs solely on Linux routers. No,
they're not P90s, they're 2U dual Xeon rackmount
servers. But they
still PCs, they're cheaper than Ciscos, they
don't come with all the limitations of IOS,
and they can handle all the bandwidth you'd want
to throw at them. Note that this is a
national backbone ISP,
not some regional setup with 12 customers. PCs
are more than capable of routing high volume
traffic.
more than a hand full of network ports
How many do you need? Stick in a couple of quad
ethernet cards, and you've already got more
ports than most Cisco routers.
The one area I would agree with you is for paired
devices using HSRP or VRRP. It wouldn't surprise
me if that's available for Linux, but if so, it's
not in widespread use yet.
That's as well as the usual stuff like travel, reading, clubbing and gig-going that any normal person does.
Bizarre though it seems to me, most people don't
count gig-going as normal behaviour. OK, I'll admit
I go to more
than most, but none of the people I work with
go to any at all, and they think it a bit odd that
I do. Ho hum. Other than going to gigs, I don't
tend to do much non-techie stuff. I've got
an interest in motorsports, and enjoy Karting
(though I don't do it as often as I'd like), and
drag racing (so far
only as a spectator, though
I hope to be running a car at some point).
Of course you can remove transparency, reduce the color depth from the original 16 colors, etc. to make it smaller - heck, I can make it 1.395K at 2 colors!
Yes, but PNG is smaller even aat the same colour
depth. I make the title 2603 bytes (that's 25%
smaller) as a PNG. The patent topic image comes
out 1.7% bigger as PNG -- i.e., to all intents
and purposes, they're the same. Note that the fact
that he's using transparency at all shows Taco
doesn't know what he's doing with digital images.
They're antialiased to only work with a white
background anyway, so there's no benefit in them
having the transparency at all. If you remove the
transparency, then PNG starts to become
significantly smaller.
I got 49 bytes (in gimp) from the gif, and 101 from the PNG (and no, pngcrush couldn't help).
You can get a 1 pixel invisible image down to
68 bytes by making it grayscale. That avoids the
overhead of the PLTE and tRNS chunks. The reason
it can't get down to the size of a GIF (which by
using a 1 bit palette can in
this case can be 43 bytes, not 49) is that
PNG has a mandatory file signature, and
IHDR and IEND chunks that add 33 bytes to start
with. The IDAT chunk then has 12 bytes of overhead
plus the data. If PNG supported uncompressed images,
then theoretically, we could have
2 bytes of data (1
byte for the pixel colour + 1 for the transparency),
which would get you a 47 byte image. Since PNG
mandates deflate compression, though, the overhead
actually increases the data to 11 bytes, and
thus takes the overall image size up to 68 bytes.
For pretty much any non-contrived example like
this, though, PNG will be smaller than GIF.
But, this the U.S. government, not a bank. It's my payroll, not an account.
In which case, it's even worse. The system was
built using taxpayers' money (including yours),
in such a way that it discriminates against you.
You'll probably find that they have a legal
obligation to fix their broken HTML (and if they
try claiming it's not broken, point them at the W3C
validator -- had it been able to pass that, it would
have worked in Mozilla). They've had to go out of
their way to design a site that only works in IE.
Not everyone is a wannabe Linux hacker; those with business goals rather than tech goals need to make business decisions, not tech decisions.
True enough. But you're overlooking the fact that
government bodies never have to make business
decisions. They have to make public interest
decisions, and the public interest is not best
served by throwing public money away on closed
soruce software. I agree with another poster,
though, that mandating open source isn't
necessarily the answer (although in many cases,
it will be). Mandating open data
formats certainly is, as is insisting on an
irrevocable royalty free license to software that would convert to other file
formats, should the chosen software prove
unsuitable at a later time.
Since it's based on KHTML from Konqueror, in many respects it's more compliant than Mozilla.
*Cough!* While I applaud Apple's choice of KHTML,
and think it shows some promise, it's way
behind Mozilla. I tried using Konqueror, but it
was a) slow[1], and b) way too good at screwing up
half the sites I visit (and create). In particular,
it's support for CSS sucks, where Gecko shines. Yes, it's getting
better, and I have no doubt that in time, they'll
fix all the bugs. But for the moment, I'll stick
with Mozilla.
[1] To be fair, this is more likely to be a
problem with KDE/Qt than with the KHTML rendering
engine.
Re:fvwm allowed me to make my perfect linux deskto
on
fvwm Turns Ten
·
· Score: 1
Hmm... it's seems like you don't neen even FVWM then, just run the naked X server!
Aahhhh, the memories. The first time I got X running
on my Linux box was with TinyX, a slimmed down X11
that came on a single floppy. You basically
got the X server, an xterm, but no window manager.
Of course, in those misguided days, I wanted to
run mwm, which was only available for Linux at
vast expense (lesstif hadn't evolved that far then).
Re:OK, most are ugly, but fvwm is the fug-ugliest
on
fvwm Turns Ten
·
· Score: 2, Insightful
just because many X11 WMs are poster children for bad taste and color blindness, doesn't mean fvwm should be kept on life support. Pull the plug already and let this brain damaged embarrassment to the *nix community die.
Care to suggest a suitable replacement? I've been
using fvwm for pretty much all of its 10 year
lifespan. In that time, I've tried a number of
alternatives, but keep returning to fvwm because
not one of the others has all the features that I
need, and fvwm does. I'm certainly not going to
ditch it in favour of a lesser wm. If another
viable alternative presents itself, then I'll
take a look. But I haven't yet found one.
My sites are typically 25% IE, 15% Moz, 50% misc robot traffic, and Opera, Safari, etc. in the lower ends. Sure, IE is at the top, but not my much. I think, overall, that IE has a little over 60% of the share.
Then your sites are not respresentative of the
Internet as a whole. My sites, along with the rest
of the net have IE up at around 95%. The fact that
50% of your hits are robot traffic just shows that
the sites are too low volume to gather any
meaningful stats. I'm getting around 40 million
hits per month, and 95%+ of those are IE.
Mozilla is the next closest at 1.25% combined
(mostly Windows, partly Unix adn partly Mac).
I often wonder why the heck credit card purchases don?t require a PIN at the very least.
Ask and ye shall receive. In the UK, all customer
present credit
card purchases will
require a PIN within 18 months.
It's being trialled in Northampton at the moment,
before being rolled out to the country as a whole.
PINs are already required in many countries in
continental Europe.
I've always suspected F5 and Cisco's LocalDirector division put out FUD on round-robin DNS to drive people to their products.
They solve different problems. Round robin DNS
works reasonably well for solving non misson critical
scalability (with a few
caveats, but we'll ignore those for now). But what
it doesn't address is resilience. If a machine
goes down, RR DNS will still send traffic to it.
A load balancer won't. That said, Cisco did a
great job of persuading anyone to buy
LocalDirectors. They're absolutely useless for
pretty much everything. It wasn't until they
bought out Arrowpoint that they acquired a
reasonable load balancing product, and even then,
they have a number of flaws and there are better
alternatives available.
I can say that 60% bandwidth for P2P sounds about right, or maybe even somewhat conservative.
Agreed. From speaking to my ISP, around 85% of
bandwidth on his network was P2P at one point.
Re:Confused
on
OSI vs SCO
·
· Score: 2, Informative
why is Linux way off to the side disconnected from everything else when a largish part is composed of BSD tools and another largish part is derived from Unix?
Not true. Linux is almost completely written from
scratch. There are a few ported BSD drivers, but
the core OS is a completely new work that shares
no code with "genuine" Unix. Of course, Linux
distributions included BSD derived code, but the
suit here regards the kernel, not userland, and
the kernel wasn't derived from anything...
Re:What are the odds that Ogg will replace mp3?
on
Ogg Now An RFC
·
· Score: 2, Interesting
What are the odds that Ogg will replace mp3 as the standard format for music?
It already has, at least for me. All of my CDs are ripped to Ogg Vorbis, primarily because I know
that I'll be able to play them in perpetuity,
thanks to the licensing issues.
For the general population, the sad answer is
that it probably never will. The lack of portable
players is often cited as the barrier to widespread
adoption. But while I'm sure it's a factor, I
don't think it would matter anyway. After all, the world
is still using GIFs, despite the widespread
availability of superior alternatives. The only exception would be
if Fraunhofer go nuts with their licensing demands.
I suspect on of the reasons PNG hasn't displaced
GIF entirely has been that Unisys asked for
sensible amounts when licensing.
The current most theoretically efficient method discovered is what's known as the "Metropolis II" layout after the algorithm used to design it
Obviously the optimal layout depends on the type
of data you're entering. So what's ideal for a
journalist may well be very different to the
ideal programmer's layout. The Maltron layout
was designed by statistical analysis of a
large amount of prose text (since most data
entry is prose text), and is pretty close to
optimal for that purpose.
Well some of us have gotten over the GeeWiz factor of running your own servers on the internet.
Nope. How anyone can live without the ability to ssh
home and check up on stuff is beyond me. It's not a
gee wiz novelty, it's an absolute necessity.
And having some ports blocked (which I dont have any on TWC) would be an extra benefit for me so my firewall has less stuff to block.
Cough. Less stuff to block? What for? If you
assume that upstream filtering is working, you're
screwed anyway, because when it fails, you won't
know, and The Bad Guys(tm) can still get in. A
firewall *has* to be paranoid and block everything, regardless of whether it's being
blocked upstream as well.
And if you're worried about performance,
then what are you running your firewall on? A
wristwatch? Mine's a P75, and it can *easily*
handle all the traffic that's thrown at it. Even
when my DSL line is completely maxed out, I don't
see any performance issues on the firewall.
Slashdot Mirror
Personally, I think they're both heading in the wrong direction. Both are a long way from where they should be, and neither are remotely well designed or thought out.
Gnome2 boots about 30% faster than KDE3. They're both patheticly slow in that department, though
Agreed, which is one of the reasons I use neither. Since fvwm2 does everything I could possibly want (including many things I can't do with either GNOME or KDE), and is 20 times faster, why would I consider switching? Eye candy does not a good desktop experience make...
I think you're safe. I too had (actually, still have) an Electron. I also had the Micro Power Positron game. So far, I haven't annihilated anything (well, apart from a few alien invaders anyway).
At University, if someone left themselves logged into a terminal, we'd place a file named "*" in their home directory. Much hilarity ensued when they typed rm * to remove it.
It goes like this. The mail hits our company yesterday morning at 10:58. By 11:00 I've sent a company wide mail out telling people that it's a virus that's slipped past our scanner, and not to open it. At 11:02 I get apologetic messages from those who had already done so -- "I thought it was someone sending me something", "It was just a zip file", "I didn't know". Yes you did, you morons! I've told you enough times! You will never teach people not to do this. People are stupid.
FLAC is lossless, which means it is CD quality. Literally. It will be a bit-for-bit perfect representation of what you'd get on the CD. As part of the tradeoff, you get larger filesizes. FLAC will typically give 2:1 compression, compared to the 10:1 you're likely to achieve with MP3 or Ogg Vorbis, so your files will be around 5 times larger.
Also, Ogg is a container format, not a compression method. Ogg Vorbis is their flagship lossy audio compression scheme. Note, however, that FLAC is migrating to Ogg, so in future, FLAC files will come with a .ogg extension.
Myth. My ISP runs solely on Linux routers. No, they're not P90s, they're 2U dual Xeon rackmount servers. But they still PCs, they're cheaper than Ciscos, they don't come with all the limitations of IOS, and they can handle all the bandwidth you'd want to throw at them. Note that this is a national backbone ISP, not some regional setup with 12 customers. PCs are more than capable of routing high volume traffic.
more than a hand full of network ports
How many do you need? Stick in a couple of quad ethernet cards, and you've already got more ports than most Cisco routers.
The one area I would agree with you is for paired devices using HSRP or VRRP. It wouldn't surprise me if that's available for Linux, but if so, it's not in widespread use yet.
Bizarre though it seems to me, most people don't count gig-going as normal behaviour. OK, I'll admit I go to more than most, but none of the people I work with go to any at all, and they think it a bit odd that I do. Ho hum. Other than going to gigs, I don't tend to do much non-techie stuff. I've got an interest in motorsports, and enjoy Karting (though I don't do it as often as I'd like), and drag racing (so far only as a spectator, though I hope to be running a car at some point).
Yes, but PNG is smaller even aat the same colour depth. I make the title 2603 bytes (that's 25% smaller) as a PNG. The patent topic image comes out 1.7% bigger as PNG -- i.e., to all intents and purposes, they're the same. Note that the fact that he's using transparency at all shows Taco doesn't know what he's doing with digital images. They're antialiased to only work with a white background anyway, so there's no benefit in them having the transparency at all. If you remove the transparency, then PNG starts to become significantly smaller.
You can get a 1 pixel invisible image down to 68 bytes by making it grayscale. That avoids the overhead of the PLTE and tRNS chunks. The reason it can't get down to the size of a GIF (which by using a 1 bit palette can in this case can be 43 bytes, not 49) is that PNG has a mandatory file signature, and IHDR and IEND chunks that add 33 bytes to start with. The IDAT chunk then has 12 bytes of overhead plus the data. If PNG supported uncompressed images, then theoretically, we could have 2 bytes of data (1 byte for the pixel colour + 1 for the transparency), which would get you a 47 byte image. Since PNG mandates deflate compression, though, the overhead actually increases the data to 11 bytes, and thus takes the overall image size up to 68 bytes. For pretty much any non-contrived example like this, though, PNG will be smaller than GIF.
In which case, it's even worse. The system was built using taxpayers' money (including yours), in such a way that it discriminates against you. You'll probably find that they have a legal obligation to fix their broken HTML (and if they try claiming it's not broken, point them at the W3C validator -- had it been able to pass that, it would have worked in Mozilla). They've had to go out of their way to design a site that only works in IE.
True enough. But you're overlooking the fact that government bodies never have to make business decisions. They have to make public interest decisions, and the public interest is not best served by throwing public money away on closed soruce software. I agree with another poster, though, that mandating open source isn't necessarily the answer (although in many cases, it will be). Mandating open data formats certainly is, as is insisting on an irrevocable royalty free license to software that would convert to other file formats, should the chosen software prove unsuitable at a later time.
*Cough!* While I applaud Apple's choice of KHTML, and think it shows some promise, it's way behind Mozilla. I tried using Konqueror, but it was a) slow[1], and b) way too good at screwing up half the sites I visit (and create). In particular, it's support for CSS sucks, where Gecko shines. Yes, it's getting better, and I have no doubt that in time, they'll fix all the bugs. But for the moment, I'll stick with Mozilla.
[1] To be fair, this is more likely to be a problem with KDE/Qt than with the KHTML rendering engine.
Aahhhh, the memories. The first time I got X running on my Linux box was with TinyX, a slimmed down X11 that came on a single floppy. You basically got the X server, an xterm, but no window manager. Of course, in those misguided days, I wanted to run mwm, which was only available for Linux at vast expense (lesstif hadn't evolved that far then).
Care to suggest a suitable replacement? I've been using fvwm for pretty much all of its 10 year lifespan. In that time, I've tried a number of alternatives, but keep returning to fvwm because not one of the others has all the features that I need, and fvwm does. I'm certainly not going to ditch it in favour of a lesser wm. If another viable alternative presents itself, then I'll take a look. But I haven't yet found one.
Then your sites are not respresentative of the Internet as a whole. My sites, along with the rest of the net have IE up at around 95%. The fact that 50% of your hits are robot traffic just shows that the sites are too low volume to gather any meaningful stats. I'm getting around 40 million hits per month, and 95%+ of those are IE. Mozilla is the next closest at 1.25% combined (mostly Windows, partly Unix adn partly Mac).
Errmmmm... no you're not! Simple arithmetic shows that 30,400 bits of data don't fit down a 760 kbit pipe...
I always do that when signing for a package, or initialling a receipt for cashback in a shop. They've never once checked, or asked for ID...
Ask and ye shall receive. In the UK, all customer present credit card purchases will require a PIN within 18 months. It's being trialled in Northampton at the moment, before being rolled out to the country as a whole. PINs are already required in many countries in continental Europe.
They solve different problems. Round robin DNS works reasonably well for solving non misson critical scalability (with a few caveats, but we'll ignore those for now). But what it doesn't address is resilience. If a machine goes down, RR DNS will still send traffic to it. A load balancer won't. That said, Cisco did a great job of persuading anyone to buy LocalDirectors. They're absolutely useless for pretty much everything. It wasn't until they bought out Arrowpoint that they acquired a reasonable load balancing product, and even then, they have a number of flaws and there are better alternatives available.
Agreed. From speaking to my ISP, around 85% of bandwidth on his network was P2P at one point.
Not true. Linux is almost completely written from scratch. There are a few ported BSD drivers, but the core OS is a completely new work that shares no code with "genuine" Unix. Of course, Linux distributions included BSD derived code, but the suit here regards the kernel, not userland, and the kernel wasn't derived from anything...
Your boss tells you to make a warhead? Wow! Even when I worked for the Ministry of Defence, that never happened to me...
Obviously the optimal layout depends on the type of data you're entering. So what's ideal for a journalist may well be very different to the ideal programmer's layout. The Maltron layout was designed by statistical analysis of a large amount of prose text (since most data entry is prose text), and is pretty close to optimal for that purpose.
Nope. How anyone can live without the ability to ssh home and check up on stuff is beyond me. It's not a gee wiz novelty, it's an absolute necessity.
And having some ports blocked (which I dont have any on TWC) would be an extra benefit for me so my firewall has less stuff to block.
Cough. Less stuff to block? What for? If you assume that upstream filtering is working, you're screwed anyway, because when it fails, you won't know, and The Bad Guys(tm) can still get in. A firewall *has* to be paranoid and block everything, regardless of whether it's being blocked upstream as well. And if you're worried about performance, then what are you running your firewall on? A wristwatch? Mine's a P75, and it can *easily* handle all the traffic that's thrown at it. Even when my DSL line is completely maxed out, I don't see any performance issues on the firewall.