still, a 30" screen... oh-h-h-h-h yeah baby, that's nice!
Huh? No it's not. It's a piece of crap. There's
no benefit in having a big screen just for the
sake of it. The value comes in being able to
display more information. But the quoted resolution
is easily achievable on a standard 21" CRT. The
larger screen size simply displays the same amount
of data at a lower resolution (around 100dpi). For
a much lower price, you can get a large (24") CRT
that will display significantly more information.
Can't understand who would be buying Apple's new
30" screen myself. Or not at that price, anyway.
For a display in a corporate foyer, then maybe
it'd be worth considering. But not for real uses
where people actually work with it.
No-one with a life has used imperial when shopping for years, if not decades.
Indeed. I can't remember the last time I bought
anything in Imperial units when I wasn't at a
bar. And even then, I tend to buy bottles of
bitch piss which come in metric units anyway.
But stuff you buy in shops is all metric, and
has been for a long time. I'm a bit odd in that
I think in metric. Apparently most of my
generation still think in Imperial units (despite
having only been taught matric units at school).
It shows that Microsoft actually cares about security.
Sadly no, it doesn't. It shows that Microsoft are
sufficiently concerned about the effect that a
reputation for bad security will have on sales in
the long term to actually do something to improve
the situation. But whatever the reasons, I'm
reasonably impressed with SP2. It's a huge
step forward for Microsoft, particularly in terms
of not just providing mechanism, but supplying
sensible defaults. There are a few things I'd
like to take a bit further than they've done.
I'd like IE to block *all* popups, not just those
triggered by JavaScript. I'd also like to be able
to block outbound traffic with the firewall, and
it only currently lets you filter incoming traffic. But it's a definite step in the right
direction.
Why shouldn't soldiers away from home have unlimited size email boxes
Soldiers away from home should have access to email
at all! That they do is a sad commentary on the
state of the military. Does anyone seriously
believe that the military doesn't have traitorous
elements? Given them access to an easy communication
mechanism is not a clever idea. Restricting
communication channels for those in sensitive areas is the
only sane thing to do. But that's too politically
unacceptable. Sigh...
Although many of the features have improved since then, the bulk of the Hotmail service is becoming increasingly unreliable for email that just "has to get there".
If it "just has to get there", you wouldn't be
using email in the first place. But even if you
are using email, why on earth would you
be using Hotmail? If it's that important, you
should be using your own SMTP server over which
you have control. Instead, you're relying on a third party,
that you're not paying, and with whom you have
no service level agreement. Not a smart move for
data you care about...
when Cart (or Indycar or whoever it is) competed in Montreal last year on the same circuit, their lap times were significantly less.
Yep, from memory, the F1 cars were running around 4
seconds per lap quicker. Given that F1 cars are around
3 seconds per lap faster this year than last, I'd
expect the difference to be even more pronounced
now.
Modern distros that use the latest versions of KDE and (especially) Gnome feel considerably heavier than before
See, this just comes across all wrong to me. I use
neither, as both are too bloated for my tastes. But
of the two, it's KDE and not GNOME that the slower
and bloatier. I'm curious as to how anyone can see
it the other way around. Certainly on all the
hardware I've tried, KDE is measurably slower.
As a completely unscientific test:
leto:~% time for i in 0 1 2 3 4 5 6 7 8 9 ; do konsole -e date; done
real 0m7.535s user 0m4.559s sys 0m0.762s leto:~% gnome-terminal -e date leto:~% time for i in 0 1 2 3 4 5 6 7 8 9 ; do gnome-terminal -e date; done
I have my doubts When free OS exist that require far less effort on your part? What exactly do your users need to get their job done?
More than can be provided under Linux at the moment.
Trust me, if I could have rolled out Linux desktops,
I would have done so long ago.
I'd rather have a KDE desktop that I can plug my camera and PDA into.
I'm sure you would. Equally, it's my job to ensure
that you can't:-) It's a vector for introducing
unauthorised and potentially harmful
files onto our corporate network. No thank you.
You must have some nasty DOS thing holding you back.
No, but there's a lot more to running a standard
office than just Word, Excel, mail and web
browsing. The call centre need integration with
the phone system, for example.
Various people need MS Project
or Visio. Finance need SAP. Marketing and analytics need SAS.
The creative team use Photoshop, Illustrator,
etc. Yes, a lot of people could get 90% of their
job done with a Unix desktop. But that remaining
10% is important, and the missing 10% is different
for each department.
As long as a program doesn't write to the registry than you can likely install it anywhere.
Not true. We have policies that prevent users from
creating any.exe,.com,.pif etc. files. That
way, even if a virus manages to get onto their
machines, it's limited in the amount of harm it
can do. We don't let them even see their C: drive,
either (amongst other restrictions). Draconian?
Yes, but it's the only sane approach for a
corporate network. With what we give them,
they can accomplish everything
they need to get their job done.
On the plus side, we remove the ability for them
to run Internet Explorer, and provide Firefox as
their standard browser. We're not evil... just
paranoid:-)
How is it so much better than GNU/Linux at the high end? Give specific examples because presently I have no way of verifying your claims.
It's certainly better, but the margin isn't as great
as it once was. Solaris still scales better to
reasonably large (50+) numbers of CPUs. Solaris also
(until recently) had better threading support. With
NPTL, though, Linux appears to be at the top of the
pile. Sun also claim that their TOE support in
Solaris 10 will give them better network throughput
for supporting huge amounts of bandwidth. Whether
this actually plays out in the real world remains
to be seen. I also haven't seen an equivalent of
things like IP multipathing[1] in Linux yet (although
they may be there -- I just haven't looked).
[1] Effectively redundant arrays of network cards,
with a highly available IP address, so if there's
a failure on one card (be it the card itself or
just a cable failure), the machine transparently
fails over to using one of the others. Tru64 also
had something like this, called NetRAIN.
Also note that if my signature had been forged, then that would have been clearly fraudulent. Because it was a number, it was considered authorised.
This is exactly the problem I have with Chip & PIN. I work for a credit card company, and I
raised objections about it. But they weren't
interested. The party line is that "it is proven
to have reduced credit card fraud in Europe, so the
UK banking industry has adopted it". Then they have
the nerve to try and sell it to the public as a good thing (which it is for the bank -- by
shifting the burden of proof from the bank to
the customer, they reduce their exposure, and
increase their profits).
MY experience is that people who use sendmail might as well just generate their configuration files using/dev/urandom. I guess MY real world experience proves YOU to be wrong, so now you're going to stop using sendmail, right?
Not at all. You didn't read what I wrote. I even
agreed that sendmail.cf leaves a lot to be desired.
You said that the only
way to get perl to look like line noise is to
deliberately obfuscate it. My real world
experience proves that to not be true, because
I have seen line noise perl that was
intended to be used rather than entered into an
obfuscated coding contest. I didn't say all perl
was like that, though, and I didn't
say it was a reason for you to stop using perl,
though, just as the cf syntax isn't going to
stop me using sendmail.
Now *you're* trolling. The only Perl programs that look like line noise
are the ones that are deliberately obfuscated
Nice try, but my real world experience proves you
to be wrong. Sure, it's perfectly possible to
write clean perl programs -- I've done so myself.
But I ban its use on our production systems simply
because it's just too easy to write line noise, and I
don't have the time to verify all the code that my
team write to ensure it's readable.
Please tell me something Sendmail does that Postfix doesn't.
Selectively route some mail by sender rather than
recipient. This is possible in Postfix, using
sender_based_routing=yes,
but only as an all or nothing affair. I need to be
able to send mail from certain senders via one
route, but all other mail using a different route,
based on recipient as normal. This isn't a contrived
example, either. We have a legitimate business
need to do that, and sendmail lets us do it.
People are familiar with it, so despite the fact that BIND and sendmail are absolute abominations, they get used.
Sigh. Y'know, I really should get used to sendmail
FUD on Slashdot, but here I am feeding the trolls
anyway. I use sendmail because it's better than the
alternatives, and it's far from an abomination.
I'm not going to claim the syntax looks good at
first glance, but then most perl programs look like
line noise too, yet the Slashdot crowd doesn't seem
to have a problem with that. When other MTAs can
match Sendmail's flexibility, then maybe I'll
consider switching. But not before.
Tom's been struggling financially for a while, and
even had to stop developing Arch because he didn't
have enough funds. Arch is the only open source
revision control system that is comparable to
BitKeeper. Subversion may be an improvement on CVS,
but it's nowhere near as comprehensive as Arch or
BK. Incidentally, even Larry McVoy admits that Arch
has the potential to be even better than BK. The
current difference is that BK is much more polished
and production ready.
Slashdot Mirror
Actually, it's JFIF -- the JPEG File Interchange Format.
So that when you do need to encrypt something, it doesn't stand out like a sore thumb, but rather it looks just like every other message you send.
Huh? No it's not. It's a piece of crap. There's no benefit in having a big screen just for the sake of it. The value comes in being able to display more information. But the quoted resolution is easily achievable on a standard 21" CRT. The larger screen size simply displays the same amount of data at a lower resolution (around 100dpi). For a much lower price, you can get a large (24") CRT that will display significantly more information. Can't understand who would be buying Apple's new 30" screen myself. Or not at that price, anyway. For a display in a corporate foyer, then maybe it'd be worth considering. But not for real uses where people actually work with it.
Indeed. I can't remember the last time I bought anything in Imperial units when I wasn't at a bar. And even then, I tend to buy bottles of bitch piss which come in metric units anyway. But stuff you buy in shops is all metric, and has been for a long time. I'm a bit odd in that I think in metric. Apparently most of my generation still think in Imperial units (despite having only been taught matric units at school).
Incidentally... Cruachan? As in Bloodstock?
Sadly no, it doesn't. It shows that Microsoft are sufficiently concerned about the effect that a reputation for bad security will have on sales in the long term to actually do something to improve the situation. But whatever the reasons, I'm reasonably impressed with SP2. It's a huge step forward for Microsoft, particularly in terms of not just providing mechanism, but supplying sensible defaults. There are a few things I'd like to take a bit further than they've done. I'd like IE to block *all* popups, not just those triggered by JavaScript. I'd also like to be able to block outbound traffic with the firewall, and it only currently lets you filter incoming traffic. But it's a definite step in the right direction.
Soldiers away from home should have access to email at all! That they do is a sad commentary on the state of the military. Does anyone seriously believe that the military doesn't have traitorous elements? Given them access to an easy communication mechanism is not a clever idea. Restricting communication channels for those in sensitive areas is the only sane thing to do. But that's too politically unacceptable. Sigh...
If it "just has to get there", you wouldn't be using email in the first place. But even if you are using email, why on earth would you be using Hotmail? If it's that important, you should be using your own SMTP server over which you have control. Instead, you're relying on a third party, that you're not paying, and with whom you have no service level agreement. Not a smart move for data you care about...
Yep, from memory, the F1 cars were running around 4 seconds per lap quicker. Given that F1 cars are around 3 seconds per lap faster this year than last, I'd expect the difference to be even more pronounced now.
alias man="LANG=C man"
I still use rxvt for exactly the reasons you describe. The above fixes man pages.
Ahhh. Not a real server[1] then :-)
[1] As in rack mounted, in a lights out machine room, no monitor, no keyboard, and a serial console as your friend.
No, not really. For 100 iterations:
I was somewhat surprised to see xterm beat rxvt. The latter does still have a smaller memory footprint, though.Actually, yes I did it for both, specifically to ensure that it was cached like you say. I just started one line too low when I cut and pasted.
It's a server -- why would you need X at all[1]? It shouldn't even be installed, let alone used.
[1] Apart from to install the abortion that is Oracle. One more thing pushing me twoards PostgreSQL...
See, this just comes across all wrong to me. I use neither, as both are too bloated for my tastes. But of the two, it's KDE and not GNOME that the slower and bloatier. I'm curious as to how anyone can see it the other way around. Certainly on all the hardware I've tried, KDE is measurably slower. As a completely unscientific test:
Seconded. "Prime obsession" is a great book.
More than can be provided under Linux at the moment. Trust me, if I could have rolled out Linux desktops, I would have done so long ago.
I'd rather have a KDE desktop that I can plug my camera and PDA into.
I'm sure you would. Equally, it's my job to ensure that you can't :-) It's a vector for introducing
unauthorised and potentially harmful
files onto our corporate network. No thank you.
You must have some nasty DOS thing holding you back.
No, but there's a lot more to running a standard office than just Word, Excel, mail and web browsing. The call centre need integration with the phone system, for example. Various people need MS Project or Visio. Finance need SAP. Marketing and analytics need SAS. The creative team use Photoshop, Illustrator, etc. Yes, a lot of people could get 90% of their job done with a Unix desktop. But that remaining 10% is important, and the missing 10% is different for each department.
Not true. We have policies that prevent users from creating any .exe, .com, .pif etc. files. That
way, even if a virus manages to get onto their
machines, it's limited in the amount of harm it
can do. We don't let them even see their C: drive,
either (amongst other restrictions). Draconian?
Yes, but it's the only sane approach for a
corporate network. With what we give them,
they can accomplish everything
they need to get their job done.
On the plus side, we remove the ability for them to run Internet Explorer, and provide Firefox as their standard browser. We're not evil... just paranoid :-)
It's certainly better, but the margin isn't as great as it once was. Solaris still scales better to reasonably large (50+) numbers of CPUs. Solaris also (until recently) had better threading support. With NPTL, though, Linux appears to be at the top of the pile. Sun also claim that their TOE support in Solaris 10 will give them better network throughput for supporting huge amounts of bandwidth. Whether this actually plays out in the real world remains to be seen. I also haven't seen an equivalent of things like IP multipathing[1] in Linux yet (although they may be there -- I just haven't looked).
[1] Effectively redundant arrays of network cards, with a highly available IP address, so if there's a failure on one card (be it the card itself or just a cable failure), the machine transparently fails over to using one of the others. Tru64 also had something like this, called NetRAIN.
This is exactly the problem I have with Chip & PIN. I work for a credit card company, and I raised objections about it. But they weren't interested. The party line is that "it is proven to have reduced credit card fraud in Europe, so the UK banking industry has adopted it". Then they have the nerve to try and sell it to the public as a good thing (which it is for the bank -- by shifting the burden of proof from the bank to the customer, they reduce their exposure, and increase their profits).
Are you sure? Those two are usually mutually exclusive...
Not at all. You didn't read what I wrote. I even agreed that sendmail.cf leaves a lot to be desired. You said that the only way to get perl to look like line noise is to deliberately obfuscate it. My real world experience proves that to not be true, because I have seen line noise perl that was intended to be used rather than entered into an obfuscated coding contest. I didn't say all perl was like that, though, and I didn't say it was a reason for you to stop using perl, though, just as the cf syntax isn't going to stop me using sendmail.
Nice try, but my real world experience proves you to be wrong. Sure, it's perfectly possible to write clean perl programs -- I've done so myself. But I ban its use on our production systems simply because it's just too easy to write line noise, and I don't have the time to verify all the code that my team write to ensure it's readable.
Selectively route some mail by sender rather than recipient. This is possible in Postfix, using sender_based_routing=yes, but only as an all or nothing affair. I need to be able to send mail from certain senders via one route, but all other mail using a different route, based on recipient as normal. This isn't a contrived example, either. We have a legitimate business need to do that, and sendmail lets us do it.
Sigh. Y'know, I really should get used to sendmail FUD on Slashdot, but here I am feeding the trolls anyway. I use sendmail because it's better than the alternatives, and it's far from an abomination. I'm not going to claim the syntax looks good at first glance, but then most perl programs look like line noise too, yet the Slashdot crowd doesn't seem to have a problem with that. When other MTAs can match Sendmail's flexibility, then maybe I'll consider switching. But not before.
Tom's been struggling financially for a while, and even had to stop developing Arch because he didn't have enough funds. Arch is the only open source revision control system that is comparable to BitKeeper. Subversion may be an improvement on CVS, but it's nowhere near as comprehensive as Arch or BK. Incidentally, even Larry McVoy admits that Arch has the potential to be even better than BK. The current difference is that BK is much more polished and production ready.