My kid hates Linux (ZDNet)

There are major differences with x86_64 that show up everywhere, not just for math.  The
calling convention on the 64-bit system are far cleaner.  More arguments can be passed to
functions in the registers (6, I think) than on 32-bit systems, where the extra arguments have
to be placed on the stack.  Stack management function on x86 are not free; they take one or a
few cycles during every function call and function return.  This can add up.

x86_64 also allows more and better ways of addressing data that can save an explicit load to
register.

These are not theoretical improvements.  Lots of programs run much better on x86_64 than on
plain old x86.

The 64-bit systems do still have the problem of larger pointers which can crowd the cache, but
some programmers find ways around this.  BEA, for example, uses short heap pointers in their
JVM, which gives them all the speedups of the x86_64 programming model (described above)
without paying the cost of 64-bit pointers.

Counterpoint, RAM is pretty cheap these days. Just Add More.

Of course, you do come across motherboard limitations occasionally.

If 64 bit pointers are really that big a deal, how come the EDA guys don't use 4GB memory
pools with 32-bit offsets?  That's way you get the speed and huge memory space of 64 bits with
the space efficiency of 32 bit.  Seems like a win-win.

It's been quite a while since I've done EDA (some VLSI layout and simulation back in 2003).  I
remember some seriously crufty software produced by vendors who would do anything to avoid an
update.  Some of the tools I used were written *and compiled* pre-1998!  It was a nightmare
trying to get that junk to run.  I eventually got the toolchain working and then I never let
anybody touch that box again.  Not so much as a security update or a package upgrade lest it
break anything.

So...  If the EDA industry is indeed pushing back against 64 bit, there might be more to it
than just pointer size inflating the working set.  :)

Fair enough; your particular daily-use EDA app (proprietary?  you've never actually mentioned
what it is) performs worse.  You use what best supports your needs, after all.

However, my daily-use apps perform significantly better under 64-bit.  That 11% improvement
with dcraw was the only one I could recreate the benchmarks on immediately, as it's trivial to
recompile.

My main daily use app (GCC cross-compiler building a multi-million line codebase) runs
considerably faster under x86_64.  However, I no longer have an identical 32-bit system for
comparison any longer, so I can't supply benchmarks without blowing half a day on it.  (The
64-bit gnome desktop *feels* faster too, but that's obviously subjective)

One of the folks I work with has also raved about the improvements he saw using the 64-bit
versions of the particular FPGA synthesizer tools.  

Not to mention the speedup one gets by not needing bounce buffers (and other games) for I/O.

I suppose it's true that if you use an artificial 2.5 GB dataset and impose a 2 GB memory
limit, 32 bit would be faster.

In the real world, why wouldn't you just spend $50 for a 2GB memory upgrade?  Then the 64 bit
box would fly.  If you're not convinced, let's try this exercise again with a hypothetical 3.2
GB data set.  :)

In my experience, modern 64 bit boxes with stuffed with lots of ram are really cheap and
really damn fast.  I can't think of any reason to deploy 32 bit for servers/HPC these days.

That's very interesting.  Has anybody tried working around this?

One solution would be convert hot 64-bit ptr fields to 32-bit offsets pointing into a single
memory pool.  I'm not familiar enough with the networking code to know how traumatic this
would be.  (I'm definitely not saying do this everywhere; just where it really matters).

This topic might make a fairly fascinating paper.  :)

"Also, I just don't buy that there's no "partial ordering on freeness." Encouraging the use of
non-free software can lead to more non-free software and less free software, which is the real
danger."

The situation with not allowing ANY proprietary software is that people who are currently
using 
Windows or OS X will not switch to Linux, since they would consider it a step back in term of 
features and performance. What you (and many others) are advocating is in fact making sure
that 
millions of people keep on using their proprietary OS'es, when they could be using a free OS 
instead.

No, don't get me wrong: given the choice, I would much rather have a system that is 100% free 
software. But if I'm NOT given that choice, I would much rather have as much free software as 
possible, while still having an usable system. If I need to use proprietary drivers in order
to get 
WiFi (for example) going, what should I do? Use those drivers, or don't use WiFi? What if I
was a 
new user wanting to try out Linux? Yeah, good luck trying to turn me in to a Linux-user by 
saying things like "well, in Linux, you couldn't use your WiFi....". At that point I would
turn around 
and walk right back to Windows/Mac.

I'm not "encouraging the use of non-free software". I'm saying that if there is no free,
functional 
alternative, and the user needs that feature, then he should just use the proprietary drivers.

"I have encountered bugs in both
the vendor's software and the GNU software. I was able to fix the GNU programs with only a few
hours work, notify the maintainer, and move on with my job."

Since I'm not a coder, that particular benefit is meaningless to me. And it would be
meaningless 
to most of the people who would migrate from Windows or Mac. Why does everyone always 
assume that anyone running Linux is a coder? Telling users that "But you can fix and change
the 
code yourself!" only makes sense if the user has the skills and knowledge to do so. Most do
not.

If we ran only 100% free software, we would not have decent 3D-acceleration. We would 
probably not have WiFi. We would not have Flash. And those are just from the top of my head. 
And that would mean that for normal users, Linux-computer is nothing but a glorified "email & 
web"-machine. And even the "web"-part would be debatable, since so many websites use flash. 
Good luck trying to sell that kind of system to new users. They would just ask "why should I
run 
Linux for those tasks, since I could do them in Windows just as well?". And on this age of 
wireless-connectivity, they could do that wirelessly on WIndows, while being stuck to ethernet
on 
Linux.

If the goal is to make people use free software, how exactly does it help if we insist on
running 
crippled systems that give the impression that Linux is not good? How does it help if our
actions 
and comments are in fact making sure that millions of people prefer using Windows or Mac? We 
are pushing users to proprietary software, while we are talking about trying to get people use

free software. It just makes no sense.

P.S: Apologies for the spacing of my text. Safari does not like LWN it seems. Yes, I posted
this 
using a Mac, and there are reasons for that.

And non-purists forget that those BIOS cause large amounts of problems for your PC that the
kernel has to jump through massive amounts of hoops to fix or otherwise work around. 

Also they slow down startup and pretty much do everything they do badly. 

Actually the purists are quite well aware of this.  For example:
http://www.fsf.org/campaigns/free-bios.html  

I have heard of few isolated incidents of TIme Machine failing for some reason, but nothing
widespread. But that said: Linux at the moment doesn't have ANY kind of backuptool. Sure,
there are probably zillion apps in sourceforge or somewhere that can back up the system. But
what we need is a system that is integrated in the distro. back when Time machine was first
demoed, I wrote a feature-request to Ubuntu, saying that we need something similar. AFAIK,
that suggestion didn't go anywhere. No backup, no undo (that Time Machine handles beautifully
as well), no nothing. We are constantly reminded by the importance of backing up, yet we are
not provided with the tools to do it. Only Time Machine managed to make me actually back up,
since it's so easy. Plug in a drive, and forget it. No hassle, no teeth-grinding. Considering
the importance of user-data, it just boggles the mind that distributions have dropped the ball
on this so badly.

Like you said: the tools available in Linux might be more reliable. But does that reliability
matter one bit, if those tools go un-used? I would MUCH rather have unreliable backup-tool
that actually gets used, as opposed to worlds most reliable backup-tool that goes un-used.
Hell, I could see my in-laws (who use OS X) using Time Machine just fine, but asking them to
mess around with rsync or some esoteric Linux-tool that does the same thing, seems
far-fetched. And the process of restoring files in Time Machine is just so intuitive that
anyone can do it. I feel that that level of ease of use would not manifest in Linux. Linux and
it's apps tend to be "traditional". File-restore in Linux would propably present the user a
list with bunch of files, asking the user to pick the one he wants. In Time Machine you have a
timeline with graphical representation of the app you are restoring data for. Fluff, maybe.
But it's so pretty, intuitive and cool that it actually amkes the app fun to use. Imagine
that, a backup-tool that is _fun to use_?

The thing with Linux and OS X is that they both let the user get their job done. But whereas
Linux simply lets the user do his job, OS X goes the extra mile in satisfying the user. The
mundane tasks that are mundane in Linux or Windows, are made less mundane in OS X through use
of clean UI's, animations, polish and simplicity. That's what I mean with Linux-apps being
"utilitarian". They present the user the things he needs to do the thing he's doing, but
nothing extra. In Linux, backup-tool would present the user with a list of files, whereas in
OS X, it gives the user a timeline of app-windows floating is space. Functionally they both
manage the same task, but one of them is polished and covered with candy, whereas the other
is.... utilitarian.

To quote Vader: "Search your feelings, and you know this to be true".

Now, I might write some kind of "article" or comparison where I compare OS X and Linux (KDE or
Gnome, I'm not sure yet) with screenshots and all, depending on my resources and time.
Hopefully that comparison will be constructive and help make Linux better. But we shall see.

"What about Amarok and K3B, to name a couple well regarded Linux/FLOSS 
apps? "

Um, K3b is an app for CD-burning. It was not the first app of it's kind, nor will it be the
last. Amarok? It wasn't first of it's kind either, there were plenty of music-jukeboxes before
Amarok.

Sure, both of those are good apps (although compared to something like iTunes, Amarok seems so
overtly complex and confusing that I can't understand it when I look at the screenshots), but
they aren't really anything groundbreaking. you might say that they do some things better than
some other app does, but that doesn't make them groundbreaking.

And, FWIW, iTunes does support other music-players besides iPods. But even if it didn't, it
would be irrelevant for me, since I happen to own an iPod.

"The point isn't that Linux/FLOSS, or other OS for that matter, has the 
advantage here (tho arguably MS does in pure numbers terms simply by sheer 
market share), but that each has its innovatively strong products without 
reasonable compare on the other OSs."

What are those innovative apps in Linux? Don't get me wrong, there are loads of good apps on
Linux, I'm talking about _innovative_ apps. Something that makes you think "damn, why hasn't
this been thought of before?". Why didn't Linux get something like Expose before OS X did?
Time Machine? Delicious Library? Scrivener? Hell, even PhotoBooth?

"FWIW, I personally choose freedom."

I would like to do so as well. But if I want to edit and organize the RAW-pictures I take with
my camera, I have yet to find a better app than Aperture for that task. And that's just one
example. Are we really free, if we can't really do anything?

I guess the difference here is that before, computers were my hobby. Nowadays, I want to
actually do something with my computers. I don't use computers for the sake of using them, I
use them to accomplish something. And it just happens that for the things I want to do, I
can't really use Linux. I might get a secondary system that runs Linux, just so I can keep
tabs on it's progress though.

"It is far from meaningless, since the Free Software community is bigger than "I"."

It's meaningless for me. How exactly does the ability to tinker help me, since I do not have
the skills needed for that tinkering? That benefit might be great for coders, but not everyone
is a coder.

"It helps to think instead of "I" as "someone, somewhere, who is technically able, enjoys
software freedom, and wants to help me get those freedoms as well"."

What freedoms are we talking about here? Freedom to modify the source? Like I said: since I'm
not a coder, that particular benefit is meaningless to me. I HAVE had ideas that I thought
were really great. I lack the skill to bring those ideas to reality, so I have spent
considerable amount of time telling the various developers those ideas (and I could do that in
OS X as well, FWIW). Just about every single time those ideas have been shot down.

"Chances are yes, and far be it from me to discourage that by saying "That's nice, but I'll be
using a proprietary replacement in the meantime. So call me when it's ready and I'll complain
online about how crappy it is compared to its proprietary counterpart. Now, chop-chop, get to
it!"

I'm not telling anyone to jump when I tell them to jump. What I AM saying is that "I want to
do certain things with my computer, and for my needs, Linux is not suitable". What do you
suggest I should do? Give up my photography-hobby (for example) while developers work on an
Aperture-replacement? Make do with apps that have maybe 10% of the functionality I can find in
those evil proprietary apps? And for what? For a benefit (open source) that does not directly
benefit me?

"Now I realize that's overstating the case a little, but if we, as a community, are not there
to support our own initiatives, who is ?"

I sure as hell support them. But does that mean that I should live with an app that does
fraction of what some other app does? Yes, Linux is wonderful for many users. Hell, I have
personally converted several people. But fact remains that for some other people, it's not
suitable, and the strict adherece to "only free software!" is going to make that situation
even worse.

"Having a lot of users is nice, but a 100 online articles about how a free system doesn't do
something quite as stellar as a proprietary counterpart isn't worth a single bug-report."

Why should we consider Linux to be the "also ran"? You just said that it's utterly meaningless
if Linux-app isn't better than it's proprietary competitor. That kinda smells like we expect
the Linux-version to be inferior. Why couldn't it be superior? Why shouldn't we strive for
that superiority? Like it or not, proprietary software is the competitor. It's the yardstick
Linux will be measured on. people thinking of switching will think "I can do X, Y and Z with
Windows/Mac, how about Linux?". We should be able to tell them that "you can do all of that,
plus you can do A, B and C". But instead of that, we too often tell them that "Well, you can
do those things, but....".

Linux does not exist in a bubble. If Windows or Mac do the things users want to do, while
Linux does not, there's no chance for Linux. People will keep on using proprietary software.
Is THAT what we want?

Now, don't get me wrong. I love free software. If it were up to me, all software would be free
software. But that doesn't mean that I will insist on using only free software, even when
superior alternative is available.