[KLUG Advocacy] RE: Adoption of Wine WAS: Novell 'opening' SuSe

[Adam Tauno Williams]

> > Adam Tauno Williams wrote:
> > > spade.  WINE helps USERS run Microsoft and Windows applications.
> > Yes, it helps run those applications, it helps run those applications
> > POORLY.
> I know it's never going to reach "native" speed, but if more people
> supported/used Wine, could it get better? 

It isn't about speed - on a multi-jillahertz processor with gigs of ram
the applications cruise right along.  The problem is hey behave even
more erratically under WINE then they do under Windows.  From screen
artifacting to wierdness in file open/save.

>  Microsoft is NEVER going to
> release Office for Linux but that doesn't mean others wouldn't make it
> more "WINE compatible".
> > > Admin's are not users and are more concerned with IT than apps. 
> > I generally disagree with this statement.   I may spend more time in
> > "Office" (whatever version) doing officey things then many of 
> > my users.
> So, what do most of your users do with their computer (other than email
> and paperweight)?

They do work! :)  Posting parts to workorders, looking at equipment
history, sourcing purchase orders from vendors, etc... All that boring
stuff that computers are on their desks for.  People often forget that
for most works 'office' applications are really just a secondary benny -
I've heard managers argue that even having Excel or Word probably lowers
a workers productivity.  And what I've seen users take the time to
create in Excel, at least for spreadsheets, I'm inclined to agree.  They
surrender to the compulsion to record and analyze all sorts of bits of
information - entering it into an XLS by hand - when the information
could have just been provided to them by the business system.

> > > Most of
> > > what I've seen is that Linux Sys Admin's hate Microsoft and Windows.
> > Well it wastes at least a couple hours out of every day.  Lets call a
> > spade a spade,  there just isn't many things Windows or their products
> > do well, or at least not consistently.
> I'm not defending Microsoft's OS'.  Matter of fact, that is why I'm
> looking at Linux.  However, from a user's perspective, their app's,
> Office, in particular, are tops.  As I see it, their biggest problem in
> the app area is elimination of really good features...  

I think Office is not the problem, OOo is a very capable competitor.  My
concern, and where I see the need for something like WINE, is with
vertical applications for which there is no Open Source equivalent.

> > > borders on a religion thing.  If a Linux admin could, they would
> > > completely eliminate Windows and its applications programs.  
> > Sure,  the users, and me, would be more productive.   Just 
> > spent an hour trying to figure out why IE wouldn't let one user login
> to 
> <snip>
> > It is NOT a religious thing - crap is crap, no 
> > matter who sells
> > it.  M$ should stop wasting my user's time;  IE is like the 
> Microsoft considers IE as part of the OS, not as a separate app...  I

I don't care what they consider it  - the thing is horrible.

> understand that most other people disagree with this designation but
> their mind-set is significant.  I'm not sure why but they have big
> problems when it comes to the OS components.

Me either.  Just take Mozilla and add Active-X support - everyone would
be happy.  You aren't making money off the browser anyway.

> But, this ignores my point, which is about apps...  
> Let's talk about Microsoft's apps: the Office family, Visio, OneNote,
> Encarta, Project, Streets&Trips (Mappoint), Money, Age of Empires,

For Mappoint or Encarta I agree - all or most others have an Open Source
(if not superior) equivalent.

> HALO...  Ok, the last two are games but it's a rare PC user who plays no
> games.

They don't in the office - they would get fired.

>   Also, let's not limit it to Microsoft, what about the 1000's of
> other companies developing for Windows.  Let's talk about the Quicken
> family of products, Peachtree's products, AutoCAD family, Mathsoft, ACT,

Sure, I agree.   But I don't think WINE really helps.  If you are going
to run any of the above applications the additional cost of Windows
licenses are almost irrelevant.  And WINE will run all the above
applications, but badly, with even less stability then if you just ran
Windows.

> Your use of the M$ moniker is a prime example of the "religion" thing.

No, that is an angst thing.  After spending more time trying to figure
out how *^@&^$& much I needed to pay them than it took to install the
product.  I've had to employ 'certified licensing advisors'.  Absurd.

> Microsoft is no more or less concerned with money than any other major
> computer company.  The only difference between IBM, HP, Sun, Oracle,
> Ashton-Tate, CA, Corel, Symantec, Novel and all the rest is that
> Microsoft found a way to win it's competitions.  That's business.
> >From an investor's/finance position, I think Microsoft is a lot less
> greedy than most other companies out there.  If you think they are bad,
> take a look at 70s IBM and see what they would do if they were currently
> in Microsoft's position.

Sure.
 
> > > So, WINE
> > > doesn't help Linux (all that much) because the 
> > powers-that-be don't want
> > > it to.  If WINE were really promoted by admin's, then you would have
> > > Windows apps running on Linux boxes and that would be anathema.
> > I use applications in WINE everyday all day.  But is sucks,  it isn't
> > anything like using a native application.  Everything 
> > "works", printing, etc...  actually is quite impressive how well it
> works.  But 
> > from a user perspective it stinks.  The app looks different,
> cut-n-paste works
> > different, and a myriad other little things.  File save/open 
> > can be just downright weird.  And I only do this because no equivalent
> native
> > application is available.
> So, tell the developers about it.  Isn't that what the Open Source
> paradigm is about?  Oh, wait, OSI is about a person being able to
> program useful features for THEMSELVES.  In the end, this is really
> about everybody being a programmer at some level.  Which comes back to
> one of my points, Linux maintains a developer/IT bias.  WINE bucks this
> trend, it TRIES to put the user in control, at least of the apps they
> are running. 

I don't bother because I don't actually believe these things can be
fixed.  An application well-integrated into its environment (evolution
for instance) makes many assumptions,  the result of making those
assumptions is the ease of use we all enjoy in an environment like GNOME
(or Windows!).   There are just things about one environment that do not
map 1:1 to those in another environment, so those applications start to
act weird (drive letters are one, how network access and 'mounts' work
is another, or the entire concept of mounting, etc...)  You can make
attempts to paper these over, but in the end that paper will rip.  Samba
struggles in some cases to run print drivers for this same reason,  you
can have VERY brilliant people doing their level best - but you just
can't recreate one environment on top of another 100%.  You need to
emulate even the BUGS - because the applications were probably built
expecting those.

> > > Users don't care one iota what OS they are running.
> > Nope, but they care allot when things don't work.
> Yes, they do.  Could Wine work?  

No.

> If so, what has to happen to make it work?  
> By the way, are you using just Wine or Crossover Office?  

WINE. If I have to pay for something why not just pay for Windows [which guarantees 95% compatibility with Windows :) ]

> Would Wine be better if it got the support of some of the other major
> apps, say SAMBA?  

Sure, but Samba gets awesome support because it currently presents a
working solution - that solution is just constantly refined (of course
they still receive a torrent of flak about stuff that doesn't currently
work, or is 'too hard').  WINE is very rough, and only the adventurous
venture there.  You often need to apply separate hacks to get different
applications to work - that is not the sign of a project approaching
maturity.  From the number of 'frustration posts' shot out onto the
Samba lists (a mature project with EXCELLENT documentation) I think the
failure/attrition rate of that same pool of users attacking WINE (a
rapidly shifting project with DREADFUL documentation) would be
dramatically higher.

> I think Samba is the paragon for Linux backend/Windows
> frontend.  

Without a doubt.

> Why can't Wine work just as well as Samba? 

No, it faces a much larger problem domain, and one that moves even more
rapidly.  Personally, I'm very impressed that WINE has even managed what
they have so far.  I can run IE on LINUX and it sort-kinda-works,
nothing I would ever present to a user, but I can test that IE only
sites are "up".  Then I usually have to xkill the blasted thing.

>  This is a true
> question, I really don't know if there is something that inherently
> prevents Wine from working well.  

Yes:  far to broad of a problem,  the problem moves beneath it, a very
small pool of potential users (only users who NEED to run vertical Win32
applications NEED to run WINE), and viable MORE RELIABLE alternatives
from Terminal Services to VMware.  I run Windows 2000 and Office 2000 in
VMware.  I paid for all three - but it always works - no glitches, no
weird errors [more than normal anyway].

> I think the real answer is that IT supported Samba.  It got support and
> it's quality reflects this.
> > > and graphics, Linux doesn't hold a candle to the 
> > applications available
> > > on Windows.
> > I wouldn't go as far as "hold a candle".  If it wasn't for IE specific
> > web site 90% of my users could easily do everything they need 
> > without a  drop of Windows.  For very vertical application you are correct.
> As you said, you spend more time doing "officey things" than your users.
> I suspect that's because they are using some "vertical application".

Yep,  only those vertical applications are almost all encapsulated as IE
only websites.  The true Win32 applications that are vertical (we have
several) are available on a Windows 2000 Terminal Server.  Mostly
because managing deployment of applications on Win32 (especially with a
large WAN involved) is another nightmare all together.  See the 'more
reliable alternative' mentioned above.

> The vertical applications are the killer.  Finance has it's accounting
> packages; engineering has its CAD stuff; marketing & sales has
> ACT/contact management; Manufacturing may be using CAM; Maintenance is
> using Excel as a database because they can no longer get dBase.  

Yep.  I see this all the time.  But I don't believe WINE is a solution.
Most (all?) companies that I see this in desperately need to take a step
back and look at their infrastructure - or start to look at I.T. *AS*
"infrastructure".

> They
> could use a couple of access wizards but that needs a PC geek in the
> department and IT hates these types of apps because they think they will
> end up with them in there laps.  

And they will.  That is why we absolutely fight the deployment of every
new thing unless the user can demonstrate they've done a real analysis
of the problem.  That is pretty darn rare.

> And for the home users, the "vertical apps" are even more important.  My
> family might use Word once every couple of weeks. I use Excel because I
> like it and know what it can do but no one else touches it.  Database,
> what's that?  The killer apps at home are: Quicken (in particular, it's
> connectivity to banks), email, Photoshop, Nero, and games (everyone has
> their own favorite) and Mastercook, in that order.  Linux has parity for
> email but it's Photoshop and Nero competitors are just not as user
> friendly.  I'm not a photo/graphics person myself.  GIMP might even be
> technically better than Photoshop but (my users say) it's not as
> friendly.  I know there are better CD writing tools but Nero just plain
> makes it simple.

I can't make any statements toward Nero or Photoshop since I've never used either.

> > But how does emulating those applications help get 
> > applications or users
> > onto LINUX?  If you are using a $500 office package and other Windows
> > apps - just run Windows,  it probably came with the computer anyway.
> > Emulation just adds another layer within which things can go wrong.
> > > Is it rational to want a completely Linux world?  Why?  
> > No, and that will never happen.  Neither did anyone in this thread
> > suggest it would (or should)
> To answer the original question... Because Windows OS sucks.  Wine might
> be another layer but does the stability of Linux solve enough problems
> to create a lower energy solution overall?

No.  It might if WINE was perfect,  but I strongly suspect that for aforementioned reasons that is an unattainable goal.

> > > Why not use the best of both worlds?  
> > I'm willing to use the best of all worlds.  I use Java apps, and PHP
> > apps, and .NET apps (I even write those!).  But as a platform 
> > that 'best
> > of worlds' will never contain Windows, for innumerable non-religious
> > boring technical facts.
> Do you mean just Windows OS or all of Windows apps?  

Apps can be ported, or just recreated.  See "Project", or "Inkscape", or
"GRAMPS", for projects arising to match specific deficiencies in the
application pool.  But this takes time.   But given that time, if it
were up to me - which it is not - I'd rather see effort towards creating
great GNOME apps than *attempting* to emulate a quite dissimiliar
platform in order to run expensive apps that may change tomorrow and
then stop break again.

> Again, this reiterates my point, IT people inherently think of the OS
> side.  In their dislike of Windows "the OS", they carry that over to
> Windows "the applications".  That is were the "religious" blindness
> comes in.

An application and its platfrom cannot be separated.  They are as tethered as a species and its planet.  Drop an antelope on Mars?

> > > The only real answer is that most Linux people
> > > completely refuse to believe that ANYTHING on the Microsoft side is
> > > better than what they have.  
> > Not true.  .NET is *AMAZING*;  it is simply put: what Java should have
> > been.  ODBC is great.
> Again, this is a IT/Developers viewpoint.  While they have a valid view,
> it's not the only one.   Try thinking from the user's standpoint.  Put
> yourself in the shoes of your accounting department.  
> One of the reasons why the PC was so successful in the 1980's is that it
> took control away from Big Iron IT and put it in the users hands. 

Sure,  I was an advocate of that.  I installed our first PC ethernet
drop - using my own money to buy the card and the cable (no small amount
in those days, even for a crappy NE2000 compatible).

> not that IT is inherently bad but, quite frankly, IT really doesn't have
> the handle on what's best for the users.  

Horse @*^&@!  Fire your IT department.

> Now comes the real conflict.
> What's best for the users usually means the most work for IT.  

Again, "Horse @*^&@!".  Most of what we do is specd by "users".  But we
determine how to get from "A" (where we are) to "B" (what the user
wants) - that is our job,  we have to know "A", "B", and "C" (the pool
of technologies and techniques available to us).  "C" is what the IT
department adds, because the user doesn't know about or should even care
about "C".  If a user says "I need Goldmine/ACT, etc..."  then the user
is WRONG!  The user needs to be trained to say "I need to track
attributes x1, x2, x3, and x4 about a customer and produce correlations
of data this way: z1 and z2".   Then IT can say - "Ah!  We already have
x1 and x3 in application/database y7,  we'll have to develop a way to
automatically record x2 into application/database y3, and create an
interface so you can enter x4.  Now do you need those reports presenting
z1 and z2 on a periodic (scheduled) basis or on-demand?  Just
internally, or directly available to the customer via the extranet?"  If
you've approached I.T. as infrastructure then getting this kind of
responsiveness is much easier then if you've allowed the EVOLUTION
(rather than the DESIGN) of myriad little data silos all over the
company/organization (where I can almost guarantee the same data is
getting recorded over and over in varying levels of integrity).

> understand why IT would dislike and, therefor, resist this.  But can IT
> understand why the users would fight for the other side?

Dumb users fight us.  Smart users *LOVE* us.  And not infrequently
'dumb' users transform into smart ones after seeing the process a few
times.  I.T. can save users hours and hours of time - if the will
participate in the process.

> > > Sorry guys, that is completely irrational.
> > And why no one in this thread suggested it.
> True, it was not suggested in this thread but it is an underlying
> thought in most of this group and the Linux community at large.

I personally don't perceive that in KLUG - in the community at large?  Who knows.  It certainly exists in some segments.  

PLEASE don't assign I.T. attitudes in general by the swill posted by unemployed nut-jobs on slashdot. ;)


> > They question in this sub-thread related to Apple;
> Right.  The thread was drifting, hence my change of subject lines....  I
> took a tangent from the VMWare/Wine sub-topic.

Always acceptable.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
Url : http://www.kalamazoolinux.org/pipermail/advocacy/attachments/20050810/f3085696/attachment.bin