|
|
| Author |
Message |
| 25 new of 203 responses total. |
richard
|
|
response 100 of 203:
| 
Feb 2 16:27 UTC 2007 |
So unless there is unanimous consent among the entire human race, you
won't accept laws impacting you that are passed for the "greater good"?
|
sholmes
|
|
response 101 of 203:
|
Feb 2 16:31 UTC 2007 |
greater good reminded me of this definition of democracy:
"two wolves and a sheep voting on what to eat for dinner"
|
cough
|
|
response 102 of 203:
|
Feb 2 17:11 UTC 2007 |
thats stupid
|
remmers
|
|
response 103 of 203:
|
Feb 2 19:02 UTC 2007 |
If I may be so bold as to post something relevant to the topic of this
item (Microsoft Vista)...
ComputerWorld has an interesting article by Scot Finnie on some of the
things baked into Vista that he considers to be problems (DRM stuff and a
number of other things). You can read it here:
http://www.computerworld.com/action/article.do?
command=viewArticleBasic&articleId=9009961
(also at http://tinyurl.com/2xcbmy)
|
richard
|
|
response 104 of 203:
|
Feb 2 19:55 UTC 2007 |
bill gates really threw down the gauntlet when he declared Vista the
safest o/s system in history. Basically saying its hole-proof. I mean
I know he's got the best software engineers working on it, and he's
probably had them looking for holes in the beta version for some time
and maybe they finally got to the point where they couldn't find any.
But how can you possible write that much code and not leave some sort
of maze that leads right through it?
|
remmers
|
|
response 105 of 203:
|
Feb 2 21:20 UTC 2007 |
You can't. For an interesting exploit, see:
http://www.techtree.com/India/News/Vista_Speech_Recognition_has_Flaw/
551-78904-580.html (or http://tinyurl.com/2kbbkg).
|
twenex
|
|
response 106 of 203:
|
Feb 2 21:28 UTC 2007 |
Re: #100. No, I simply put my faith more in evolution than revolution.
Re: #103. How dare you?!
Re: #104. My, you do enjoy making us laugh. Most secure OS ever? That's what
he said that last time. Best engineers? They must spend most of their time
playing pocket billiards. They certainly can't be putting much effort into
creating decent code.
|
tod
|
|
response 107 of 203:
|
Feb 2 22:59 UTC 2007 |
re #104
I've sat in lectures from both James Whittaker and and Mike Howard. Both of
these gents have been on the development/security of Vista. Mike's pitch was
that they employed a file fuzzing for malformed data, threat modes and
blockbox testing, memory defenses and stricter services, block header
integrity checks, heap terminations on corruption, and rooted over at least
1.2 million annotations. Even through all of that, Mike knows and admits that
Vista will have bugs which will show up after shipping. He also admits though
that they will patch and fix whatever pops-up much more efficiently than is
done with any other flavor of OS you see on the market.
James Whittaker on the other hand..he's the guy who breaks the stuff before
it ships. He's a hell of a lot of fun and I'm envious of his job at
Microsoft. He discussed the origin of bad things, noticing the environments
of applications as well as their inputs and logic within. At the end of the
lecture, we explored an IE 7 bug in the internet options security tab where
you could set all sorts of restrictions against porn sites to protect your
kids at home only to have them create their own msrating.dll file in the
iexplorer.exe directory which is blank but bypasses your original settings.
>;)
|
vivekm1234
|
|
response 108 of 203:
|
Feb 3 15:30 UTC 2007 |
Re #106: Ahem! I think it was James Gosling that said that MS coders were
pretty good in some article ages back.
|
kingjon
|
|
response 109 of 203:
|
Feb 3 17:02 UTC 2007 |
Re #106, #108: It all depends on what the coders are putting their effort into.
Q.v.:
http://www.cs.auckland.ac.nz/~pgut001/pubs/vista_cost.html
"A Cost Analysis of Windows Vista Content Protection"
|
gull
|
|
response 110 of 203:
|
Feb 3 22:09 UTC 2007 |
Re resp:32: It's not just that alternatives weren't user-friendly.
Remember OS/2? It was actually pretty user-friendly, and for a while
had a loyal following at some businesses. The problem is the
application support just wasn't there. More new software got written
for Windows, so people bought Windows to run it. OS/2 made a last-ditch
attempt at salvation with a version that could run some Windows 3.1
applications, but it never caught on and the operating system disappeared.
Then there was BeOS. Very pretty, very user friendly. But the hardware
and software support wasn't there. Software companies didn't write apps
for it, hardware companies didn't write drivers for it, and it disappeared.
The fact is, hardware and software companies *like* having an OS
monopoly. It means fewer support issues for them. It also means
cheaper development, because they don't have to port their code to other
operating systems. They don't want to go back to the bad old days where
a major app might have to support half a dozen platforms. For that
reason the Windows monopoly is fairly self-sustaining even absent
Microsoft's dirty tricks.
Re resp:36: Linux has come a long way. I've done a few SuSE installs
recently and they've worked out of the box, just like Windows XP
(usually) does. Actually, I've had fewer driver issues with SuSE than I
have with XP. This represents a major effort on the part of the kernel
developers to support new hardware as it appears, often without any help
from the hardware manufacturers.
BUT, ease-of-installation is nearly irrelevent to Windows. Almost no
one actually ever installs it themselves. They buy a computer with it
pre-installed.
Re resp:93: Software support for Windows 2000 is going to dry up pretty
quickly now that it's been end-of-lifed. I'm already seeing new
hardware that doesn't have drivers for it, and Microsoft isn't going to
be providing security patches anymore.
Re resp:95: That's not what I've found. My experience is that Windows
2000 takes far longer to boot than Linux, but apps come up quicker once
the boot process is done. This is probably because a lot of Windows
apps pre-load during boot to get faster launch times. IE and MS Office
are in that category.
Now, it's true the Windows *desktop* appears quicker. But at first the
Windows desktop is unresponsive. You're wasting your time trying to
click on anything until that little hourglass goes away. By the time
all the apps pre-load, the virus checker loads and updates, etc., it's
taken almost five minutes for my Win2K box to stop thrashing its disk
and be responsive enough to be useable.
As an IT worker, what bugs me most about Windows is its black-box
nature. This gets worse with every release. It breaks in subtle ways
that are often impossible to fix without reinstalling the whole OS,
because the internals are walled off and largely unknown to everyone
outside of Microsoft. (In fact, it's so big and complex there probably
isn't any one person at Microsoft who understands it, either.) In many
cases you couldn't fix it if you did understand them, because files that
are open are untouchable, and most of the critical files are open
whenever the OS is loaded. Worst of all, it's non-deterministic -- you
can do the same procedure twice and get two different results. It's an
intensely frustrating operating system to work with.
I'm also often faced with a choice between security and functionality,
with Windows XP. Many Windows apps simply refuse to run without
administrative privilages. But if I give someone admin privilages and
walk away, I'll be coming back two months later and spending hours
removing the spyware from their computer.
|
twenex
|
|
response 111 of 203:
|
Feb 3 22:59 UTC 2007 |
I agree that hardware and software monopolies are attractive, but not in the
way they are implemented. It would be perfectly possible, for example, for
Linux to be a monopoly - and yet (unless MS manages to pwn it due to their
patent threats and deal with Novell) the effect would not be the same since
there are many companies involved in its production.
Just like we have now with the Intel (actually AMD) architecture, but lots
of PC companies.
|
keesan
|
|
response 112 of 203:
|
Feb 4 03:51 UTC 2007 |
My linux boots in 15 sec on a small drive.
|
tsty
|
|
response 113 of 203:
|
Feb 14 09:01 UTC 2007 |
you pay micro$chmidt lotss-0-bukxx to beta=test their nwe os.
hmmmmmmmmmmm
i think i;ll wait ... as usual
but the, i donlt have clients who demand to be on teh bleeding edge either!
in fact, one client of minme (fact!) wnast me to upgrade him to dos 6.22 (from
6.0) and to windoeze 3.11 (from 3.0).
his 386sx b0x works perfectly fro *him* ! and that makes us both happy.
pthbbbb!
|
easlern
|
|
response 114 of 203:
|
Feb 14 14:01 UTC 2007 |
What convinced him it was time to upgrade his space heater?
|
maus
|
|
response 115 of 203:
|
Feb 15 00:51 UTC 2007 |
Keesan, 15 seconds is mighty fast. What distribution and version, and on
what platform? What you say it boots, do you mean that it finishes the
initrd or that at 15 seconds, you get your dtlogin/xdm ?
|
keesan
|
|
response 116 of 203:
|
Feb 15 02:47 UTC 2007 |
I use loadlin to boot from a DOS directory with the kernel in it. I don't
get any dtlogin/xdm. I get vt1 (console, text). 15 sec is on a small hard
drive. If it has to check a larger one it can take longer. Basiclinux 3.40,
Slackware based. On a 486 or later. I could also boot with lilo directly
from power-on which might take a few seconds longer, and depends on the
computer . Some take that long to check their memory. Basiclinux does not
start any daemons. No random number generators. No X, no GUIs - add those
later after booting if you want them.
|
maus
|
|
response 117 of 203:
|
Feb 15 03:37 UTC 2007 |
Wow, that's spartan. I guess I am spoiled; I like to have at least X and
Motif or something like CDE or GNOME or something on my workstation,
and on servers obviously daemons will need to be run. Now that I think
about it, I don't think I even have any drives small enough to be
checked in 15 seconds.
|
keesan
|
|
response 118 of 203:
|
Feb 15 03:48 UTC 2007 |
This linux can be run off two floppy disks, or you can put it into an ext2
partition. I got several browsers and a few other things into 40MB, including
Opera and I think Abiword. I have been setting up friends for internet. No
crashes, quick boot, Opera for email, or webmail. No WIndows viruses. I
checked and no linux viruses either in 4 years, probably because I don't run
any servers. The 2-floppy version has X with swm, and xli, and a text browser
with xli for viewing images non-inline. Dialer, telnet, wget, ftpput, etc.
I set up friends to go right into X with a menu (icewm). 120MB was fine for
linux and also 60 DOS games with space to spare. We usually use 340 or 500MB
drives because they are faster. 16MB is best but 3MB is minimal.
|
maus
|
|
response 119 of 203:
|
Feb 15 04:21 UTC 2007 |
That certainly serves a different niche than what we need. We typically
look for something that works in a networked environment, is supported
(or at least supportable) and that works with modern hardware.
While they are not the speediest in the world, I've been fairly thrilled
by both SLED and RHEL; both are well behaved, well supported, reliable
and featureful.
Of course, on my workstation at the Dallas DC, I run OpenBSD 3.9 with
mwm as my working GUI environment.
|
keesan
|
|
response 120 of 203:
|
Feb 15 15:32 UTC 2007 |
My linux works in a networked environment. You add pieces as you need them.
I never heard of SLED or RHEL.
|
maus
|
|
response 121 of 203:
|
Feb 15 17:08 UTC 2007 |
SLED is SuSE Linux Enterprise Desktop
RHEL is Red Hat Enterprise Linux
What is the time commitment to tweaking it to useability? Are you able
to be notified of updates and automagically apply them?
up2date/yum/zmd/zen-updater are really nice and save me loads of time
and aggrivation.
|
keesan
|
|
response 122 of 203:
|
Feb 15 22:17 UTC 2007 |
There are no 'updates'. The author and the users compile things and offer
them to others for use. I compiled lynx and ghostcript and netpbm, all with
help, and packaged them. Anyone who feels like it adds whatever they like
and if it does not work, asks for help. xpdf did not work with libc5 X so
I figured out how to get pdftoppm to work instead, used with zgv. Someone
put together for me a 1-floppy version to use with my USB camera. I put
together an 8MB loop version to put ON my USB camera (on the memory card).
Someone else used the 1-floppy version to learn on then I compiled a special
kernel for them to run linux in 8MB ramdisk for use in Prague. They put
together mutt and sound packages for us. Another list member and I are going
to compile busybox against uclibc. He just compiled uclibc and made me an
account on his computer. This is not a turnkey sort of linux.
|
maus
|
|
response 123 of 203:
|
Feb 15 23:32 UTC 2007 |
That sounds like a lot of fun for a hobbyist who has more time than
anything else. At some point, when I have the time, I may check it out,
but I will say that it is ill suited for the use that I make of Linux
(or any other OS+Software stack). When I have machines that I or others
count on, I cannot rely on some guy in usenet; I have to have solutions
that are backed by a company with a financial incentive to do things
right, whether that is to keep software patched and automatically
available, or to provide well-integrated and thoroughly tested useful
software. If a security patch is not available because the user
community does not think the software is hip, my users are put at risk.
Patch management becomes extra important when you have a large number of
internet-facing computers; the one system that you forget to manually
patch can be the toe-hold for a malicious intruder or for rot and
entropy to set in.
|
nharmon
|
|
response 124 of 203:
|
Feb 16 00:51 UTC 2007 |
The only reason I've come across for running SLED or RHES is to get
support from Oracle (and probably other software vendors). Otherwise,
CentOS is great. :)
|
