[Leaplist] good read

[Steve Litt]

On Friday 14 March 2008 06:23, Fred Moore wrote:
> I don't know how many have come accross this article.. but it has about
> 10 really valid points that hinder Linux.. I think almost all of us are
> gilty at times.. Fred
>
> http://www.intranetjournal.com/articles/200803/pij_03_10_08a.html

I read that article and find only one really accurate point that he made -- 
the point that I'm going to call it FUD after reading it.

I could refute most of the 10 points, but I have a book to write today, so let 
me just refute his most aggregious, "Hurdle #4: Wireless is a Mess".

Well yeah, getting wireless to work on Linux is a mess, so I guess that's 
truthful. He then goes on to say "No, it is the distro developers lack of 
commitment to making darn sure that open source drivers receive precedence 
over those of the restricted variety."

Oh, I get it, he doesn't like ndiswrapper. His solution is, instead of getting 
an ndiswrapper solution out a couple weeks after a new onboard wnic is 
released, that notebook buyers wait months for someone to do the necessary 
reverse engineering to put out a "true, free software driver". Oh, and about 
that time the new onboard wnic will be old and superseded, and a new one 
comes out.

He says "While blaming vendors is getting old, they are part of the problem." 
PART of the problem? Let's see Matt Hartley make a driver for something with 
no specifications, no API, no known interface. Let's see him sit there, day 
after day, eyes blurring in front of a packet sniffer, reverse engineering 
every single little feature. Knowing it's likely that by the time he 
finishes, the product will be obsolete. If the vendor published the API, 
*even I* could write a driver (I'd probably need some serious guidance from 
Tony Becker, but I could do it). Without an API, it's blind, binary reverse 
engineering. Vendors' refusal to release an API is 95% of the problem, so 
it's disengenuous to say they're "part of the problem".

The author says "In the end, users are cheap and want to use what came with 
their notebooks -- very few exceptions here. Suggest that they go out and 
purchase from one of the no-name wireless vendors that do support Linux and 
the user will balk at the very idea."

The author needs an English lesson here; what does "one" mean as in "purchase 
from one?" Does he mean purchase an external wnic? Where ya gonna get it? How 
are you going to know it's any good? Or does he mean the whole laptop. I 
guess he's right -- I'm cheap because I'm unwilling to spend an extra 50% to 
100% on a computer guaranteed to run *at least one distro* of Linux. If it 
were 10% to 20%, it would be a different story, for me and for lots of 
people. But spending $2000.00 on a mainstream laptop in 2008 is just bad 
economics, and spending $1300.00 on a laptop with RAM and HD compatible with 
2006 applications is just plain short sighted.

Like I say, I could go on and on through most of his other 9 points, but I 
have a book to write. I'd like to say one last thing. If this guy is the same 
Matt Hartley as this:

http://www.computertroubleshoot.com/about.html

then this guy has no clue how to develop a driver, an app, or anything else. 
The guy's a columnist.

Now don't get me wrong -- some of my best friends are writers :-). But in my 
case, besides writing, I've originated several free software projects, one of 
which is quite popular (VimOutliner). I've at least scratched the surface of 
the difficulties in writing software with zero budget. The guy on the "About 
Matt Hartley" web page is portrayed as a talking head. Heck, Rob Enderle 
could have written the same article.

That's not so far fetched -- a Google search brought up many "what's wrong 
with Linux" articles written by a guy named "Matt Hartley". He speaks 
of "Linux" as if it's a business entity. He asks us to reorganize our 
priorities. If he really understood free software he'd understand that our 
priority is to scratch our own itch, and secondarily to attract more 
developers to help scratch it even farther. That's Capitalism, free software 
style. If he doesn't like it, he can pay me a salary to reverse engineer 
secret black box wnics.

Like I said, he's right about one thing: I'm calling his article FUD.

STeveT