Monday, 30 August 2004
[Last Week]
[Monday] [Tuesday] [Wednesday] [Thursday]
[Friday] [Saturday] [Sunday]
[Next Week]
[Daynotes Journal Forums] [HardwareGuys.com
Forums] [TechnoMayhem.com Forums]
{Five Years
Ago Today]
08:20 -
The Wireless Networking chapter is due Wednesday, so I'll be working
hard to get that done. I don't think I'm going to make it, but I'm
going to try very hard to get it to my editor by the end of the week.
That means there won't be much posted here.
Boy, can I call 'em? In October 2001, Bill Gates promised that Longhorn
would ship in 2003. At that time, I posted the following prediction to
a mailing list I belonged to:
"2003?
Hah. Not bloody likely. I don't expect to see Longhorn client until
2008 and Longhorn server until 2009, if Microsoft is lucky. Of course,
they can ship anything they want and call it Longhorn, but for what
they've said will be in Longhorn, I don't see it happening much before
the end of the decade."
As it turns out, those words were prophetic. Microsoft announced Friday
that they've taken a meat axe to what they'd previously defined as
Longhorn. Their goal is now to get the client version of what they now
define as "Longhorn" out by "late 2006" (which of course really means
2007), with the server version to follow later.
And the new "Longhorn" isn't Longhorn at all. Call it "Shorthorn". It's
basically XP Reloaded. The core of Longhorn was to be the new WinFS
filesystem, and that's gone entirely. Although Microsoft isn't giving
details about what other features are to be eliminated, it appears that
the second pier of Longhorn, the Avalon display technology, will also
be cut back dramatically. About the only thing left is the Indigo
middleware technology, and at this point we don't know for sure what
cuts, if any, that will suffer.
In all, it seems to me that Microsoft's newly-defined "Longhorn" is
destined to be too little and too late. Linux won't be standing still
for the next three years. Think back to where desktop Linux was three
years ago, and imagine where it will be three years from now.
Microsoft's desktop strategy is in deep, deep trouble.
[Top]
Tuesday, 31 August 2004
[Last Week]
[Monday] [Tuesday] [Wednesday]
[Thursday] [Friday] [Saturday]
[Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com
Forums] [TechnoMayhem.com Forums]
{Five Years
Ago Today]
9:35 - Two copies of Building the Perfect PC
arrived here yesterday. Even our editor doesn't have a copy yet, so I
assume they're hot off the press. Those of you who signed up for a
review copy should be getting it shortly.
The book is really impressive. I never cease to be amazed at what
O'Reilly's editing, layout and production folks do with the raw
manuscripts I send them. We're responsible for the content, certainly,
but it's the magicians at O'Reilly who turn it into an attractive book.
And this is a very attractive book. It's the first book we've done in
full four-color, and color makes a huge difference.
My quilting friends tell me that the custom when making a quilt is to
make at least one error. That derives from the Amish, who believe that
making a quilt perfect is an offense to God. And so they make one
intentional errror, dropping a stitch or whatever, in each quilt they
make. With books, that's not necessary, because I've never seen a
perfect book and never expect to. No matter how careful the editing and
proof reading, at least a few small error inevitably slip through. When
I scanned the book last night, I noticed only one. In discussing an
Antec 1080AMG case, we somehow dropped a character and referred to it
as a "108AMG" case. Oh, well.
Heads-down work continues on the new chapter. I did have a to-do
item on my calendar to migrate Barbara's primary desktop system to
Xandros Desktop today and tomorrow, but that'll have to wait until I
get the chapter complete and off to Brian Jepson at O'Reilly.
Any time there's a significant change in computing environment, the
possibility of mistakes is greater. Accordingly, I decided to implement
a more rigorous backup program in preparation for moving Barbara to
Linux. Previously, I'd been doing a full backup to tape every Sunday,
with interim daily backups (actually, several times a day) done to
remote volumes across the network. Periodically, I'd burn archive CDs
(lately, DVDs).
I wanted to start doing a bit more to secure our data, and also to
provide finer granularity in terms of how many versions I'd be able to
recover. So I'm now doing daily backups to DVD+RW discs from my primary
Xandros box. I have six daily DVD+RW discs, labeled Monday through
Saturday, five weekly discs, labeled Week 1 through Week 5, and twelve
monthly discs, labeled January through December. That gives me daily
granularity for most of a week, weekly granularity for a month, and
monthly granularity for a year, which should suffice. That's of our
active working data directories. The archive and deep archive
directories are backed up to DVD+R when I move files to them, and also
quarterly.
As to the software, I'm using K3b under Xandros to burn DVDs. It
works pretty much like Nero Burning ROM under Windows, but is actually
faster. Each evening after we walk the dogs for the final time and
Barbara heads back to bed, I simply drag our primary working directory
over to the appropriate DVD disc and drop it there. K3b burns the disc
in a few minutes and then ejects it. Not much work required, and I end
up with a nice snapshot of our working data burned to DVD, where I can
retrieve files if necessary without doing a restore operation.
[Top]
Wednesday, 1 September 2004
[Last Week]
[Monday] [Tuesday]
[Wednesday] [Thursday] [Friday] [Saturday]
[Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com
Forums] [TechnoMayhem.com Forums]
{Five
Years Ago Today]
08:35 -
Deadline day for the Wireless Networking chapter, and I'm nowhere near
done. I'm going to spend today cleaning up the chapter so that I can
send it it, although I'm not happy with it yet. I did at least get my
access point to the point where I could access it and shoot some screen
shots. Note to self: don't change the IP address of an AP and set it
aside for a month or two without writing down the new IP address
assigned to it.
Last night was one those nights. I went back to bed by about 11:00 and
read a mystery until about 11:30. When I turned out the light, I
figured I'd get to sleep quickly, as I usually do. But 1:00 a.m. came
and went with me still wide awake, as did 2:00 a.m. and then 3:00 a.m.
I think I got about four hours' sleep, and my tail is dragging this
morning.
And it looks as though we may need to replace the roof. Yesterday
morning, I heard hammering and banging at the house next door, which is
vacant. Then the doorbell rang, and it was a guy asking if they could
fill up their water cooler at our outdoor faucet. I told him sure and
then wandered over to talk to the guy in charge.
They were putting a new roof on the house next door, which surprised me
because the roof they were replacing was just installed a few months
ago. "Hail damage," the guy told me. "Looks like you have some, too." I
asked him how much they were charging to replace the roof next door,
and he told me about $4,700. That's an expense we sure don't need at
the moment, and I mentioned that to him. "No problem," says he, "your
insurance will pay for it." He said he'd climb up and take a look
before he left for the day, just to verify that there was damage.
So I called our insurance agent, John Phillips at State Farm. John said
to give him a call back today if the guy said there was damage and he'd
send out an adjuster. If the adjuster agrees that the roof needs
replaced, we'll have to pay a $500 deductible, which is a lot better
than $4,700.
It never occurred to me that hail would damage the roof. That sounds
stupid, I know, and I certainly realize that huge hail can do a lot of
damage, but it just never crossed my mind that the hail storm we had
some time back would have damaged the roof. Oh, well.
I just called John back, and their automated attendant says the office
is closed for Labor Day. That seems odd, so I'm hoping it just means
they aren't in the office yet.
[Top]
Thursday, 2 September 2004
[Last Week]
[Monday] [Tuesday]
[Wednesday] [Thursday] [Friday] [Saturday]
[Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com
Forums] [TechnoMayhem.com Forums]
{Five Years
Ago Today]
09:32 -
I sent off a chapter yesterday to my editor at O'Reilly, although it's
not the final first-draft chapter. I hate missing deadlines. I'll be
working the rest of the week and probably through the weekend to turn
out the actual first-draft chapter. As soon as I finish that, I need to
write a 1,000-word tie-in article for the O'Reilly web site, probably
"Building the Perfect Linux PC".
O'Reilly tells me that the book is shipping out of the warehouse by the
thousands, but some bookstores may not have it until mid-month. Amazon
was still showing it as not yet published the last time I checked, but
it should show up there and on the other on-line booksellers' pages in
the next couple days. Those of you who are getting review copies should
receive them before long, although I don't know exactly when.
I worked all day yesterday on three or four hours' sleep. Fortunately,
I slept well last night. Very well. I didn't even hear the garage door
going up and down when Barbara left for work. I woke up at 8:30.
Malcolm was curled up asleep under the window in the bedroom, and
Duncan was asleep on the floor next to me. Good dogs.
Today I'm doing more work on the Wireless Networking chapter.
[Top]
Friday, 3 September 2004
[Last Week]
[Monday] [Tuesday]
[Wednesday] [Thursday] [Friday]
[Saturday] [Sunday] [Next
Week]
[Daynotes Journal Forums] [HardwareGuys.com
Forums] [TechnoMayhem.com Forums]
{Five Years
Ago Today]
08:00 -
I get very tired of hearing the specious argument that all or most of
Microsoft's security problems exist because Microsoft's market share is
so large. Linux, so the argument goes, would be plagued with just as
many exploits and of equal severity if only it had market share
comparable to Windows.
To illustrate the bogosity of this claim, all one need do is look at
web servers, a segment where Microsoft is the minor player. In web
servers exposed to the Internet, Apache rules,
as it has done for many years. Apache has something like a 70% market
share. IIS is a distant second, at under 20%. So, by these people's
reasoning, Apache should have been the victim of many more and more
severe exploits than IIS.
But think back. How many severe exploits against Apache can you
remember? None? I can't, either. Now, think about how many severe
exploits you can remember against IIS, including at least a couple that
came close to bringing down the Internet itself. Do the names
download.ject, Code Red, and Nimda ring any bells? Yep, despite its
small market share, Microsoft IIS is the target of choice among people
who want to hack web servers.
How can that be? There are so many more servers running Apache that one
would expect it to be the target of choice, at least if one subscribes
to the faulty reasoning employed by these people. So, which do we
assume to be wrong, their reasoning or the indisputable facts? I think
the answer to that question is obvious.
As to Linux being vulnerable, let's think about that. Let's also think
about the Mac and OS X. Between them, Linux and OS X have a small but
significant market share. Certainly tens of millions of computers run
them. Tens of millions. And those tens of millions of computers are
pretty much unprotected. Many are behind firewalls, certainly, but few of them run virus scanners or malware detectors. In fact,
there aren't many Mac or Linux virus scanners available, and no malware
scanners I'm aware of. Why? Because viruses and malware aren't a
problem on Linux and OS X. There's no market for software designed to
solve non-existent problems.
So, we have tens of millions of wide-open systems. If you believe the
Windows-is-a-victim-of-its-own-popularity brigade, these Linux and Mac
systems are ripe for the plucking. After all, they can't be inherently
more secure than Windows, right? That means that any punk who wants to
make a name for himself could be the first cracker to create a
widespread exploit of Linux or OS X. Think of it! He'd be famous. Well,
famous in the only way that crackers can be famous. But he'd have done what no one else had done before.
Do you seriously think if it were as easy to exploit Linux or Mac
systems that it wouldn't have been done already? Give me a break. Look
at it this way. Would the average cracker rather be the very first
person to exploit 10 or 20 or 30 million Linux or Mac systems, or the
10,000th person to exploit 300 or 400 million Windows systems? I think
the answer to that question is obvious as well.
Obvious too is the bogosity of claims that Microsoft is a victim of its
own high market share. Microsoft would love you to believe that, and
spends a lot of time, money, and effort to make sure people continue to
believe it. But the reality is that Microsoft is a victim of its own
design philosophy, which has always been to favor features and ease of
use over security. Microsoft Software: Insecure by Design.
Now that security has become a major concern, Microsoft is falling all
over themselves to make Windows secure. It's hopeless. Windows and
Windows apps are unfixable, short of re-writing them from the ground
up. You can't secure a hopelessly broken app with patches and tweaks,
any more than you can replace the foundation of a house after it's been
built. Security is something you have to design into software, not
something you can add later.
What I tell you three times is true:
Linux is inherently more secure than Windows.
Linux IS inherently more secure than Windows.
LINUX IS INHERENTLY MORE SECURE THAN WINDOWS.
Sorry for the shouting. But please don't believe anyone who tells you something different.
Speaking of Windows security (an oxymoron if there ever was one),
Thomas C. Greene has an excellent article posted on The Register, WinXP SP2 = security placebo?
I don't always agree with Mr. Greene's opinions, but he's nailed this
one. In the article, Greene points out just how insecure a Windows XP
system patched with SP2 is, and recommends changes to better secure an
XP system. It's well worth the time to read.
12:06 - Here's why I don't trust anything CNN reports.
This article reports a slaughter among the children taken by Chechnyan
(read Islamic) terrorists. Any reasonable person would call these
bastards terrorists, kidnappers, murderers, and monsters, because
that's exactly what they are. And, to be fair, CNN does use the T-word
a few times in the article. But they generally try hard not to call a
terrorist a terrorist. Instead, they refer to these sons of bitches as
"militants" or "hostage-takers". Jesus. Could they have found less
emotionally-laden words to describe these scum? It makes my blood boil.
In attempting to appear unbiased, CNN in fact abandons their
responsibility to report fairly and accurately. Calling a spade a spade
isn't biased, it's accurate. Calling a terrorist who lurks concealed
among civilians and then rears his ugly head to slaughter innocents a
"militant" or a "guerilla" grants legitimacy to his actions.
Yet one more reason to eradicate the plague that is Islam. There are no
innocent Islamics. These outrages are part and parcel of their
so-called religion. It's time our leaders admit that fact and take
action accordingly.
[Top]
Saturday, 4 September 2004
[Last Week]
[Monday] [Tuesday]
[Wednesday] [Thursday] [Friday]
[Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com
Forums] [TechnoMayhem.com Forums]
{Five Years
Ago Today]
10:31 - Here's an amazing fact. Last month, about 10% of the traffic this site received was from Linux.
Now, granted, some of the traffic was from me, and my readers are much
more likely to run desktop Linux than are average users, but even so.
That's something like 60,000 hits and 12,000 page reads by Linux
clients.
That bodes well for the continuing growth of Linux as a mainstream
desktop OS. My readers are early adopters of things technological, and
early adopters are also opinion leaders. Early adopters are the tip of
the pyramid. As time passes each early adopter will be joined by
several, then many, others. It happens at a grass-roots level, spread
by word of mouth.
For example, I'm building a PC for our friends Paul and Mary. At one
point, I was on the phone with Paul, and said, "If you really want me
to, I'll install Windows on this system for you, but I think you'd be
better off running Xandros Linux and not having to worry about the
virus/worm/Trojan of the day." After some discussion of applications
compatibility and so on, Paul shouted to Mary, "Bob wants to install
Linux instead of Windows on this system. Is that okay with you?" Mary
shouted back, "Sure," and that was that.
Now, understand that neither Paul nor Mary has ever actually used
Linux. They saw it running once or twice at our house, but that's about
it. But they're both very bright people. Both have PhD's in Organic
Chemistry. Paul is a professor at Wake Forest University, and Mary is a
chemist for Dow. Both of them are used to taking on new challenges and
succeeding. Both are pretty confident of their own abilities, and both
face life with a can-do attitude. They're exactly the kind of people
who are excellent candidates for Linux.
So, the chain reaction continues. Brian Bilbrey showed me Xandros Linux
when Barbara and I were visiting the Bilbreys over the Memorial Day
weekend. By Independence Day, I had converted to using Xandros Linux
full time. Here it is Labor Day, and I'm about to make three more
converts to using Xandros Linux, Paul, Mary, and Barbara.
Then there are those who've tried Xandros based on what I've written
here, many of whom are now running Xandros either exclusively or at
least as a secondary OS. And, speaking of opinion leaders, I even
convinced Jerry Pournelle to install Xandros Linux on one of his
primary systems. He hasn't converted to Linux, yet, but the seed is
planted.
And each of those seeds is another locus of infection. Many of them
will infect other people with their enthusiasm for desktop Linux as an
alternative to Windows. It'll be a long process, but I have no doubt
that over the next few years Linux will prove itself to be the
deadliest threat to Microsoft's domination of the desktop.
At some point, we'll be able to celebrate Parity Day, when Linux has
desktop share equal to Windows. When will that be? I don't know, but my
guess is it'll be before the end of this decade. With Shorthorn,
Microsoft has granted desktop Linux the precious gift of time, and I
think that'll be Microsoft's undoing.
[Top]
Sunday, 5 September 2004
[Last Week]
[Monday] [Tuesday]
[Wednesday] [Thursday] [Friday]
[Saturday] [Sunday] [Next Week]
[Daynotes Journal Forums] [HardwareGuys.com
Forums] [TechnoMayhem.com Forums]
{Five Years
Ago Today]
[Top]
Copyright © 1998, 1999, 2000, 2001, 2002, 2003, 2004 by Robert Bruce Thompson. All Rights Reserved.
