An interesting idea for you techies

[Rob Wehrli]

Jason wrote:


I read your entire message.  Your half-hearted attempts in your reply to
suggest that you gave "module shit" reasonable credit where due is
complete BS.  Your "WinBlows" and other noxious remarks including the
one about people who do not have "enough smarts" are complete rubbish. 
What do you say to the neuro-surgeon who uses a Windows 98 PC to check
his Email and then he replies in all capital letters?  You don't have
the smarts?  There are many, many people who use a Linux system who do
not feel the need to recompile their kernel, in fact, many of them do
not even waste the "physics" space by installing the compiler.  Does
this mean that a gaggle of lawyers do not have the smarts because they
use Linux and WordPerfect to sue your butt off?  The point in all of
this that I've been trying to make is to MODERATE yourself. 
Ahhhummmmm.  Take a *chill pill* and learn how to get along rather than
point the finger.  Take a look at the following link for some basic
advice.

http://www2.linuxjournal.com/articles/culture/0017.html


> Uhhh... Did you read to the bottom of my message before pressing
> reply? I have used modules for exactly that, and I believe I mentioned
> that possible use of them in the message you replied to.
> If you had read the entirety of my message, you would have noticed
> that physically removing and reinstalling hardware devices was also

Your message actually never says anything about removing/reinstalling
hardware, rather, "reset, in a way, certain parts of the kernel" and
"pull devices off or put devices on your SCSI bus" assuming hot-swapping
drives.  This is a fairly unlikely choice for the greatest percentage of
typical Linux users when you compare all of the reasons why you'd like
to use modules effectively...so long as you don't "get covered in module
shit."

> one use of modules that I acknoledged as a Good Thing, in fact, I even
> went so far as to include examples of how it could be used with
> desktop systems as well... but you didnt get that far before you
> replied.

Your rather esoteric or ill-fated uses for modules as you described in
SCSI and IDE "examples" used in your earlier message said:

> One thing that *is* good about modules, however, is the fact that you
> can "reset", in a way, certain parts of the kernel, allowing you to
> get away with things not normally advisable. For example, if you boot

Here you imply that modules allow you to get away with things that are
not "normally advisable"; Normally advisable by whom?  By You?  Your
advice *sounds* like more of the tongue of a cynic who hates Microsoft
and feels that if I've got anything better to do with my time than wait
for a kernel to compile that I am of not "enough smarts."  OK, so
modules are now the bastard child that allows you to steal the
neighbor's car for a quick trip to the store for more beer before
needing to kick the dog and beat the kids yet again.  Certainly not the
kind of thing real Linux users would want to use or embrace for any
reason, especially since they're "module shit."

> from IDE, you can unload the SCSI module, and then pull devices off or
> put devices on your SCSI bus, then simply reload the module and avoid
> a reboot. The truly daring can boot from SCSI and do the same with IDE

The "truly daring" sounds more like, "those who wear tights and a cape
and purposely jump from roof to roof to see if they *really* can not
fly."  A fool's paradise?

> (warning, plug the cable in straight and power up the drive first,
> some drive will attempt to suck voltage THRU the IDE cable if not
> powered up, generally causing a reboot or worse, motherboard failure).

...because you'll probably hose the entire thing if you screw it
up...stooooopid.

> They can also be extremely useful for someone attempting to get a
> given driver to work with an improper piece of hardware by modifying
> the source to the driver.

So, now, like the previous bastard child, modules are now good if you're
trying to take an "improper" device and fix that piece of junk using the
source.

> > What if you're debugging a new device driver, huh, hotshot?  A complete
> > kernel recompilation sounds a bit foolish now, huh?  Not just a whole
> 
> Uhhh... Did you read to the bottom of my message before pressing
> reply? I have used modules for exactly that, and I believe I mentioned
> that possible use of them in the message you replied to.

You can see just above that you never once mentioned "debugging
hardware" rather "attempting to get a given driver to work with an
improper piece of hardware."  The operative word here is "improper"
hardware.  The use of improper suggests that I'm trying to stick a
MacIntosh card into a PC, not that it isn't "fully debugged."  I'm not
trying to play a word game here by overly focusing on the words you're
using to make some point, rather, that you're spewing forth as if your
gospel knowledge of Linux is reasonable advice.  You're plainly calling
modules "shit" and now you're supporting their use and back tracking? 
It is clear to me that you do not very well understand modules and their
benefit, even if you have a very rudimentary basis in their use and
operation within a given Linux system...otherwise, you would not be so
easily to call them "bad names."  Grow up.  It is OK to not know
everything about something.  Like the article in the link above states,
if there is something you do not know, don't try to BS your way through
it.  Find the areas where you do have something to CONTRIBUTE and add
then.  Any asshole can be a cynic.  It takes a real person to open up
and become like Linux.  Open.  Not hiding.  Not spewing forth, but there
for the reading and taking if desired.  Don't be part of the "Jihad";
Linux is not a religious war.  Don't bad mouth Windows.  You probably
don't know enough about it to accurately bad mouth it, either.  I've
developed Windows applications since version 2.0.  I like Windows.  I
like Linux.  I like SunOS.  I like NextStep.  I like MacOS.  I like
Windows CE.  I like VxWorks.  I like uCOS/II.  I like a lot of operating
systems and environments.  There are many good things in each of them. 
None, including Linux, are "perfect."  They all attempt to meet the
needs of a very diverse set of users' needs for computing applications
and appliances.  Don't join the crowd of Linux users who feel that they
have to bad mouth everything that is GNL  (snicker, Not Linux?).

Let me try to make a peace offering:  I'll start by a little personal
revelation.  My group of Hawaii Linux Users joined with another
"competing" group of Linux users to join forces and have a more unified
approach to offering group services.  We joined to become one large
group rather than two small groups.  My group was largely heavily
technically involved.  I owned an ISP in the early 1990s (before the
web!) and it was about 1994 that I started promoting Linux to other
ISPs.  While it was clear then that Linux did not have the necessary
services to be considered capable of truly high-availability, especially
with some of the networking problems it faced under heavier loads, I
promoted its tertiary features.  By 1996, I'd organized a multinational
Linux Usersgroup to Linux Usersgroup (InterLUG) video teleconference
using Linux-powered CUSEEME reflectors.  The basic idea was to have high
bandwidth Linux servers from around the world (T1 or higher) set up and
configure a reflector and those who were able, provide a feed.  I
organized a "receivership" of this activity at Portland State University
in Portland, OR.  As I was friends with the PLUG president, David
Mandel, he invited me to participate at his LUG meeting.  So, a fellow
HLUGer and I, on our own nickels, travelled to Portland from Hawaii to
show-n-tell our video teleconference *and* our streaming video on demand
server powered by a Linux-based VDO server.  The basis of the entire
presentation was Linux Services for Linux ISPs.  On my advice, David
invited Phil Hughes, publisher of Linux Journal.  At the meeting, Phil
heckled us non-stop.  He kept saying that "if I wanted to see a Windows
demonstration, I wouldn't have gone to a Linux meeting" and things of
that sort because we were showing the client applications running on
Windows and MacIntosh hardware.  The entire backend was being powered by
Linux, but it didn't matter to Phil at all.  All he could do was bitch
about Windows this and Windows that like a little whining baby.  He
tossed around freebies from the magazine to the crowd, and they tended
not to notice his rudeness as much.  He never gave us a single bit of
credit for having 9 countries of Linux Users talking and "interacting"
to/with each other via CUSEEME all within Linux.

Back in Hawaii, my team of technical "empowered" met each month with the
"other group's" seemingly technically incapable.  Their biggest concerns
were things like how to do a client-side PPP connection to the
Internet.  For those of us who had banks of modems regularly taking
incoming PPP connections from Mac and Windows clients, the single
workstation PPP connection seemed trivial.  We were busily designing
distributed, multiuser databases for sharing data common to the 44
countries in the Pacific Disaster Management Information Network, where
earlier (1995) I'd installed bunches of Linux machines to form the
network infrastructure.  The people in my "high tech" group simply
wanted to use Linux to make some bread.  We'd been making our bread from
Windows, DOS, Novell, SunOS, BSDi, DEC and HP UNIX and a host of others,
and now we were keen to do it using our "preferred" OS.  The "other
group" had absolutely no interest in using Linux to make a profit.  They
were almost some form of "Linux Purist" albeit, without much technical
skill.  I tried to remind them that it costs a lot to live in Hawaii,
but they simply wanted to "explore" the world of Linux in some fashion
of personal academic pursuit.

It was then that one of the "other group" members challenged that Linux
was better than Windows.  I snapped a quick look at him as if he were
drooling from the side of his head.  This was probably late 1996. 
"Better than Windows?"  I asked, incredously.  "What do you mean
better?  You like Linux better than Windows?"  It's just better.  He
told me.  "Name one thing that it is better at doing?"  I asked him. 
After that, the details are a bit fuzzy because so many things that he
said I quickly negated in argument with the entire group present. 
Eventually, it came down to a set of two final things that he absolutely
felt that Linux was "better" at than NT (I rarely use "Windows" in my
actual discussions, as I felt that NT was the only reasonably solid
platform for development of Windows applications...so, for the record, I
said NT rather than Windows in my arguments.):

Of course, neither of the two things were "inherently" Linux, rather two
"applications" that ran on a Linux-based system.  (Read the GNU/Linux
arguments of RHS, if you care to.)

Applix Words as as good as Word (He finally modified it to "Just as good
as Word" from "better than Word")

CGI applications were easier to write than under Windows.

The latter example suggested a multitude of applications needed to use
any useful CGI app, but I allowed him the benefit of the doubt by
saying, OK, so, let's take a common use for a Word processing
application and a common need for a CGI application and let's see once
and for all if Linux is better than Windows NT.  So, I asked that we
define what better really means.  Quicker...was the resounding response
from all of the users.  Quicker development time or quicker execution
time?  Both were important, with execution time being their preference. 
I told them that if they were the business owner who spent $12,000 a day
on development staffing that two more days of development time was
darned important when the difference might be only a few milliseconds in
application execution speed and that the networking overhead of the CGI
app might very well negate that performance benefit.  Well, they were
definitely not in the profit and loss mindset, so, it was more important
for them to get the last ounce of performance rather than the bigger
bang for the buck.  You can probably see why I was somewhat frustrated
at the union of the two groups.

So, the next step was to devise a suitable application:  They meandered
for a very long while offering superficial ideas that would not interest
anyone outside of a very limited group, certainly not a "common" need. 
They also offered impractical ideas that ment that testing them would
produce improbable result sets.  Things like:  How about an application
that allows someone to send a fax to their dry cleaner to check to see
if their dry cleaning is done.  Well, it is probably a common task for
that individual, but it probably isn't a common thing for a whole lot of
individuals using a Word Processing application.  Not only that, they're
answer for it was to use it to print a document out and then walk over
to the fax machine to fax it.  Not exactly screaming paperless office
and information age.  They countered with piping the output to a fax
peripheral and other, then-current means of automating the process. 
Finally, I stepped in and offered a suggestion.

How about a mailing lable program that bangs against a database of
customers to send targeted flyers out to the specific customer types? 
That at least sounded like something that a real business might want to
do from time to time.  And, I continued, we'll use the CGI side of it to
capture the customer data to the database, simulating an order entry
system or simple "flyer request" mechanism.  The merging of the two
projects into a combined project made sense to everyone, since it was
really a time when everyone wanted to maximize the interactivity of
data.

I completed the development, testing and demonstration of the NT
application in 4 hours using Windows NT 3.51, MS Word 6.0 and O'Reilly
and Associates' WebSite Professional 1.0. I used MS Access for the
database, ODBC and JDBC-ODBC bridge for connectivity and Java JDK 1.1.3
(I believe) for the development language.  I wrote a Word macro to open
the Access database and query it along with printing the lables to an
Avery standard label layout and printing the correct number of flyers
from the Word document choices.  As you can imagine, a lot of my work
was pointing and clicking, dragging and dropping.  Subsequently, I spent
about two weeks of spare time writing a similar Linux application.  The
"execution" times were as visibly similar as I could tell by looking at
them.  The only obvious difference was the number of manual steps needed
at each layer for the Linux side of it to complete.  Overall, the Linux
"application" was substantially slower because of more manual input
needed whereas the NT side of it was mostly automated.  Was more
automation possible on the Linux side?  Certainly, but I'd already spent
more than 20 times the labor in getting to THAT point.  While I'm
equally certain that I didn't convince anyone that Windows is "better"
than Linux (it wasn't my goal to do so), I don't think that anyone
appreciated the fact that Windows is very well done, too.

To put all of this time-consuming blabber to rest and summarize some
kind of point for you here:  You are only limited by your ability to
perceive.  Don't let your own perceptions of the world confuse you into
believing that they are the only way to perceive the world.  Join forces
with those whom even share completely dissimilar views so that you may
learn from them and their perception of the world.  A good harmony is
really sweet.  A screeching set of fingernails on the blackboard is raw
noise.

Take Care.

Rob!