[KLUG Advocacy] A User Leaving

[Adam Williams]

>Being somewhat busy at work, I wanted to get involved with this thread
>but I chose not to.  I wanted to state some gut feelings.
>Gut feeling 1)  CHANGE IS F'n HARD for the average user.  If you move
>one menu or put some field 1 over, users will female dog and moan till
>they are blue in the face and then wake up the next morning still
>moaning from yesterdays use of this different feces.  I may be
>exaggerating a little here, but my point is that even small changes are
>hard for the average user.  I guess after implementing a business
>software switch at our company I cannot put enough swear words in this
>and get my emotional feeling across so I won't even try.  

I'm with you.  Just close your eyes picture them all toiling away in a
concentration camp somewhere.  It really is theraputic.

>Just bear with
>me on my not wanting to fill the page with intercoursing this and feces
>that and person born without the benefit of wedlock who have certain
>tendencies toward the desire for no changes what-so-ever.

But they do want change.  They want it to get easier, and faster,
etc...  but without changing.

>Gut feeling 2)  I can empathize with this guy very much.  I made 1 point
>about how in gnome v. kde, kde will give you an hourglass until the app
>starts letting you know that you can set and relax for a second and WW3
>attack came right in my face about how ignorant f'ks like me should not
>talk about these things and that what I was talking about was

You have my empathy there.  You wouldn't believe how many flaming hot
e-mails I've gotten about my LDAP presentation.  RFC822 doesn't define
how LDAP represents mail aliases, it only defines the concept of a mail
alias!  So I'm an idiot!  Of course, the attribute being discussed is
"rfc822mailmember",  so doesn't it make sense to mention what rfc822
is?  Sheeesh.  I've gotten this one more than once.  There is no
shortage of jerks, a-holes, and mine-is-bigger-than-yours types out on
the Internet.  But I haven't really noticed a greater proportion of this
inanity in the Open Source community as there is in just about
everything else.  But on the Internet they can get away with being a bit
more rank than they would be face-to-face.

>soooooooooooo0oooo0oooooooooooo easy to configure and I would like to
>think of myself as no dummy and I still have not found how to configure
>this.  I tried for 1/2 hour and trying any more just is not worth it to
>me to continue trying.  

I mentioned in one of my e-mails that it was turned off in the default
builds due to some problem.  He must build everything himself.

>My email was even an invited reply for examples of differences.  

Sorry about that. :)

>Until the distributions make the kind of studies out of
>how to create the "Fischer Price" version of RH or the upcoming United
>Linux distro like M$ made XP out of NT/2000.  The open source desktop
>will not be able to obtain the mindset of the consuming public from the
>devil.  

I think for an end user the Linux desktop need to be pre-installed.  If
not my the OEMS then by a LUG, friend, etc...  XP, and kindred, come
pre-installed with all the icons just-so.  Installing something like NT
from CD onto bare metal IS a frustrating experience but one most users
are spared.  Unfortunately it will be very hard for Linux to duplicate
that effect.

>Most people would rather reboot and complain about the support
>they cannot get from megacorp than deal with bombastic bastards, insults
>of their intelligence and flame wars they know and care little about.

True.
 
>They also just want to get what they want done done.  I am sure though
>that A Fischer Price version is not advisable and not what the community
>wants.
 
I'm not convinced a version need to be dumbed down to be easier to use. 
But to do simple things one should never see the complexity.  One
doesn't have to use regedit ever time one installs software,  but thank
goodness it is there.  I think the same concept applies.

>Gut Feeling 3)  Most open source software comes from people tinkering
>and doing it for the love of it.  'Most' open source programmers want to
>get what they want done, done.  They are not so interested in making
>understandable graphical configuration tools for complicated services
>and configurations.  Only programmers that get salaries for performing
>such tasks do so (for the most part). 

I think this is true in part.  But 1.) I don't think the end-user
desktop should really have complicated configurations and services and
2.)  The GNOME and KDE control panels are light years beyond where we
were not to long ago.  

There is an issue of developement critical mass.  Back in the day making
an X application was hard, and every program reinvented the wheel.  With
things like XML parsers, gconf, glade, etc... it is MUCH easier to make
a GUI app on UNIX.  This is why you see a `slow explosion' of end-user
applications under KDE and GNOME.  Things like bonobo, gstreamer, etc...
are creating yet another higher level of components that will accelerate
things even more.  If you make is easy developers will do it.  But this
is just now really building up steam.  This is the reason I rankle at
people calling GNOME (or KDE) bloatware.  It is the exact opposite of
bloat.

>This morning I mounted a couple
>of iso files created from catting a cdrom device to a file by editing
>fstab with a loopback configuration.  Now I did not do it by
>understanding every item I put into fstab, I found a machine at work
>where I did it before, copied and pasted and wala, mounted iso file.  I
>am not going to create a nautilus style gui today and submit it to parse
>and edit an fstab.  I was lucky I had an fstab I could ssh into to see
>to save me time.  Try doing that today without doing things that Sally
>Someuser would not do.  Try getting Sally Someuser to evan understand
>that this kind of thing is possible.  

Right.  So that places it a bit outside this argument.  Simple things
should be simple to do.  But complex things aren't going to be simple.

>Mounting iso files is one of the
>many things that make OSS powerful.  No graphical admin tool for this
>yet and definitely no read your mind kind of wizard.  I'm not advocating
>read your mind wizards, just saying that would be required to satisfy
>most.

I'm all for wizards,  no harm so long as I don't have to use them. 
Ximian had a project called "Ximian Setup Tools" but I'm not certain
what became of it.  They've been silent for awhile.

>I tried with relative lack of success to get my Father to use Linux for
>a desktop and failed.  Maybe I did not spend enough time next to him
>solving issues or maybe I think from my perspective a little to much.  I
>like Xterm and someone like my Dad does not really want to type at a
>command line.  One time it took me ALOT longer than I wanted just to get
>him to Xterm over the phone and tell me what his ip address was so that
>I could proceed with sshing into his computer.  I failed at this attempt
>and I still love my Father and he just wants to get done what he wants
>to get done.  I guess I have come to peace with losing this battle now.

I've setup a couple of Linux boxes for `home' users with pretty good
success.  But this was all with very recent software.  Open Office (at
least for these) covers about 98% of what they want to do.

>Gut Feeling 4)  Linux on the desktop has it's best chance in the
>corporate setting where CFO's like the savings and IT personnel are
>there every day to solve issues that would have to be solved at the
>command line.  Let's face it.  Sally Someuser does not want to type at a
>command line.  Linux REQUIRES at least a little understanding of some
>problems to solve them and a certain percentage of those solutions will
>require the "antiquated" command line approach that I use 9 (or more)
>hours per day.

I'd agree, but with the caveat that I think the same applies to that
other OS.  I find myself spending most of my fix-it time at the command
line either way.  In fact XP and 2000 send me back to the command line
with things like "ipconfig", "nbtstat", and "arp" that aren't available
in any GUI.  At least 9x had "winipcfg"

>Gut Feeling 5)  Our best chance at the home Sally Someuser desktop is
>with the user who uses it at work and finds they like it.  Then they may
>be willing to put up with the insults to their intelligence to save the
>money they would have spent on M$.  

Agree.  In the very least I think people the most interested in Linux
are those who have met it somewhere.

>Wait, let me reconsider that.  Most
>people would not put up with that kind of treatment at alllllllllll in
>any way shape or form, not even to save $500 from a 500lb used car
>salesman who has the car of their dreams.  They just won't take it.

True.  And they shouldn't have to put up with it (unless it's about GW
giving away national awards to his favorite political hacks and a token
liberal, then all bets are off.  One of the most disgusting pompous
self-congratulating smarmy slabs of meat to ever walk upright.  Sorry,
he just came on the TV and I temporarily lost control of myself).

>Let me end my rant by saying I hope you guys know I'm on your side.  I
>like Linux on my desktop and my wife likes her win98 and my dad likes
>linux on a server he does not use and 2000,98,XP on his other three
>machines.  I'm OK with that.  The community will not succeed at some of
>these conversions unless mindsets change dramatically.  I truly can
>empathize with the writers position.

I have a little less empathy for the writer because I think he made some
tragic (and reasonably obvious) mistakes.  I could easily list "10
things wrong with Linux" and hardly a one of his problems would show
up.  Package management for one example.  That is one problem the
community has solved, at least better than one any other platform (be it
Mac, AIX, or Win32).  And it was something he cited.

But I think some of his problems are indicative of the number 1 problem
I would list: "You can't find the bloody documentation!"

If I imitate GW and congratulate myself for purely self serving effort,
I think I can say I've created some reasonably useful documentation on
things like problem solving, LDAP, Kerberos V, cvs, etc...   All done
because the documentation I could find is hopelessly scattered across
the Internet.  I've since found quite a few site that cover basically
the same stuff I did.  But finding the information.....

It really isn't (for the most part) that the information isn't out there
but that there is no mechanism for finding it.  And I've had exchanges
with other "authors" who feel the same level and type of frustration. 
The community as a whole places a pretty low value of documentation. 
The LDP only will accept documents in a certain format,  which excludes
99% of the documentation out there.  Refusing documentation in a format
like PDF blows my mind.  Even if they gave it a lower status and were
still willing to link to it it would help.  We have sites like rpmfind
and freshmeat for finding software but nothing equivalent for
documentation.    And freshmeat brings up a peeve, and I think a good
example of the attitude.   I reference alot of LDAP related projects in
my LDAP presentation and provide a fair amount of additional
documentation on some of them in the presentation.  I've been in contact
with project members of just about every project I mention to see if it
is OK or I have represented them fairly/accurately.  All have been
pretty happy to get the publicity/recognition.   I post a "comment" of
the project on freshmeat with an URL to the presentation (as it has
content concerning the project).  Several authors have thanked me for
the comment as they don't want to write documentation.  But freshmeat
sends me a crabby note about "spamming" the message boards.  How it is
SPAM I don't know,  I guess it must be because I'm not complaining about
how it won't compile on slackware and must suck (the gist of most
freshmeat comments).  It just stands as a great example of how the upper
crust of the community view documentation; all those newbies should
muddle for years on dumb terminals just like they did. 

But again this is a problem a LUG can solve for someone,  as "we" know
where to find the documentation, or at least how to find it.  Until
someone makes "crispveggies.net" to serve as a master index of Open
Source documentation.