Linux-Advocacy Digest #785, Volume #29 Sat, 21 Oct 00 12:13:03 EDT
Contents:
Re: Pros and Cons of MS Windows Dominated World? ("David Brown")
Re: Claire Lynn (Bob Hauck)
Re: Pros and Cons of MS Windows Dominated World? (Bob Hauck)
Re: Is there a MS Word (or substitute) for Linux? (Bob Hauck)
----------------------------------------------------------------------------
From: "David Brown" <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy,comp.os.ms-windows.advocacy,comp.unix.advocacy
Subject: Re: Pros and Cons of MS Windows Dominated World?
Date: Sat, 21 Oct 2000 17:31:32 +0200
Mike Byrns wrote in message <[EMAIL PROTECTED]>...
>> What is "the Unix worm" ?
>
I didn't start using Unix until 1990, so I missed that one.
>
>> >And most people
>> >spend their time trying to break into Linux boxes because they are
easiest
>> >to get into out of the box.
>>
>> A poorly configured Linux machine is easier to break into than a
>> well-configured NT server. However, a well confingured Linux box is far
>> harder to hack than any NT/w2k box (other flavours of Windows cannot be
>> well-configured from a security perspective).
>
>That's just plain wrong, IMO. You claim a well configured Linux box is
harder
>to hack than _ANY_ Windows 2000 server? Care to post some proof of that or
are
>we expected to blindly believe the linux myth?
OK - the claim is exagerated and cannot be proven (logically, it would
require testing every single w2k server). We bounce back and forth between
extreme opinions, and hopefully compromise somewhere in the middle nearer
the unbiased truth than when we started.
My claim is based on what proffesionals use. For example, in major ISP and
web server provider near here, they use a mixture of NT4, w2k, Linux and
Solaris for their servers. NT4 (and newer ones running w2k) servers are
used for those customers that specifically need MS features (such as running
MS SQL server), Solaris is used for a few specialised applications, and all
the rest runs Linux. In particular, those that want ASP support always run
on Linux, and those that need the highest security run on the tightest Linux
boxes. These guys get paid a great deal of money to know about configuring
the systems and running secure servers, and money for licences or hardware
is not a major issue. When they want the safest systems, they choose Linux
every time.
>
>> >In fact, if you read about most hacks they have to do with Perl shopping
>> >cart applications.
>>
>> Perl is a language - Perl shopping cart applications are programs written
in
>> that language. If these have security holes, then that is the fault of
the
>> shopping cart programmers - the OS can do nothing to stop them. And
exactly
>> the same situation will apply if the Perl scripts are running on a NT box
>> (Perl is available for a wide range of systems - you cannot easily tell
if
>> the shopping cart is on Linux, NT, or anything else).
>
>Cool that means that all the VBScript exploits can be thrown out too!
Perl is fine for small systems, but it is not a good choice for very large
cgi applications like large shopping sites - it is too hard to write large
fool-proof systems. For small applications it works fine. VBScript,
however, makes it quick and easy to access vulnarabilities in the client
system.
>
>> I think the original author was suggesting things along the lines of the
>> fact that virtually every virus ever created has been for MS systems, and
>> now the vast majority take advantage of the inherint lack of security in
MS
>> office and IE. It is a simple matter to write a web page which asks the
>> dancing paperclip on a visitor's machine to delete some critical files.
>
>But that's all VBScript. Just like Perl. And anyone running IE5 got
notified
>of the patch and it offered to install it for them.
The VBScript is run on the client's machine, so the damage is done there.
With Perl, the script is run on the server and it is there that the damage
occurs. These are two totally different situations - the one requires that
the end user is an expert and keeps track of all the latest updates and
patches, the other requires that the server administrator is an expert.
>
>> >> #3. Closed Source.
>> >
>> >See #2. Open source means the source code is available for all hackers
to
>> >peruse. Scary.
>>
>> It has nothing to do with #2 - people cannot see the source of Perl
scripts
>> unless the server is mis-configured or the scripts are badly flawed.
>>
>> Open source means that people can find the flaws, and either fix them or
>> tell people about them so that others fix the flaws. The majority of
>> "security announcements" for open source products are fixes for potential
>> holes that are found and patched long before anyone has found a way to
>> exploit them.
>
>Same thing with Windows. Almost all of the "bugs" that Georgi Guninsky and
>others have found have no record of ever being used in a malicious exploit.
>Consider for a moment that Georgi Guninsky only targets _Windows_ systems
and is
>known to be a staunch UNIX advocate. I think it just goes to show just
what
>kind of guerilla tactics the Penguinistas are willing to stoop to. Doesn't
>really matter though. The systems admins in corporate America know what
>consititutes a real vulnerability and what's contrived just for the news.
Fair enough - a lot of the potential exploits in Windows are just that -
*potential*. But there are also a great many that are real (witness the fun
of Windows email viruses). And regarding Georgi Guninsky, when he finds a
potential hole he tells MS first, and only informs others if MS does not
deal with the problem suitably.
>
>> Closed source means that only MS sees the flaws - if they ever bother to
>> look. You only know there is a problem once someone has found
>
>Like all the Penguinista's on their smear campaigns.
I suppose to be fair on MS, they have usually found a whole pile of faults
that are already prepared for the next service pack before they launch
version 1 of a new OS or application.
>
>> and exploited
>
>Not necessarily. Most of these "flaws" have no record of ever being
exploited.
>
>> the flaw, and then you have to wait for MS to produce a fix (sometimes
>> taking a year to do so).
>
>Last one a remember took a whopping 12 hours :-)
Some are quick, but some are slow - it took about a year for a patch to
outlook that will stop it from running scripts without warning users of the
danger.
>
>> It also means that MS can put whatever they want
>> in the code, such as backdoor keys. Scary.
>
>Sure they can. You're being spied on right now since you are choosing to
use
>X-Newsreader: Microsoft Outlook Express 4.72.3110.5. If you're so scared
why
>aren't you using Linux?
I can't do everything with Linux, and for work I need NT. Outlook Express 4
is not too bad, actually, as long as the system does not have newer versions
of IE or MS Office. I don't really care if others can find out what news
program I use, but I do care if someone collects information about my
machine (such as the old registration wizard for Win95).
>
>> >> companies like Hewlett Packard and IBM are not very likely
>> >> to dump their brains into a Microsoft Kernel.
>
>Nor should they. They both sell UNIX for those machines. HP/UX and AIX.
They
>are not open source.
>
>> The GPL guarantees
>> >> everybody equal and unrestricted access to the code.
>
>That's such a bonus to the average Joe! Even to a developer, a lot of that
code
>is so sophmoric as to be, in fact, a lesson in how not to write C. The X
>Window system is a fine example of how not to design a high performance
GUI. I
There is no doubt about xfree86, but the same thing applies to Windows GUI
as well. A great deal of software started out well, then became overloaded
with features, patches, fixes and short-cuts. It is always faster and
easier to fix what you have than to start again. And just because Windows
GUI often feels faster than a typical X desktop on the same hardware, does
not mean that the Windows GUI is any better designed or implemented.
>think each Linux developer should be given a copy of Bruce McConnell's
"Code
>Complete" and be made to comply with it like the programmers at Microsoft
are
>made to.
Do you really think MS programmers write good, structured, modulerised code?
Why should they be any different from the vast majority of programmers round
the world? My access to MS source code has been as limited as anyone
elses - looking at example programs that come with various development
tools - but even for small bits of code, the stuff is a mess. I am not
saying that any particular open-source program is better or worse -
programmers vary everywhere. But you will not convince me that programmers
at MS are "made to comply" with anything, given that MS quality control has
passed code including passwords like "Netscape programmers are weenies" and
the concept virus.
>
>> >> Microsoft has 37,000 paid employee's versus Linux's 200,000
>> >> free lancing, free contributing programmer/analysts.
>
>I've never seen any of those linux numbers substantantiated? What is the
>basis? Even if there are a few thousand of these folks, unless they are
>unemployed, they can only spend a maximum of a few hours a day working on
>linux. That means that it takes about 3 to match just the time put in by
one
>full-time Carnegie-Mellon graduate from Microsoft. I've seen code from
>Microsoft and code from linux. Of the example's I've seen the Microsoft
code is
>consistantly better structurally, syntactically, algorithmically and
>stylistically.
There are probably over a thousand paid, full-time linux programmers around
the world (my estimate), employed by dedicated linux companies such as Red
Hat, and other companies that make heavy use of linux such as IBM and Corel.
There are a great deal more who are involved in writing applications for
linux (or porting from other unix versions), and a proportion of them will
make use of the source code and contribute changes or improvements. Also,
the modular structure of linux is such that a great many of the parts are
written by single people, or very small groups. Throwing large numbers of
programmers at a huge task is seldom a successful development strategy.
How many of MS's 37,000 employees are actually programmers? And how many of
these spend a substantial amount of their time actually designing or coding,
as distinct from documentation, meetings, testing, planning, courses,
support, and all the other activities that conspire to reduce the time
available for actual programming.
>
>> >> By using Microsoft you will be guaranteed the HIND END
>> >> of technology. You will always be running on obsolete hardware
>
>Really? New hardware is almost always supported on Windows first.
Windows is normally the first to support new hardware like graphics cards,
printers, and so on, as the drivers tend to come with the hardware. But
support is often limited to only the newest versions of Windows. For
example, where is the USB support for NT 4.0 ? MS claims it can't be done,
but there are third-parties producing NT4 USB drivers for their own devices.
With linux, support for new hardware can be put in old software, and often
without even needing to reboot the machine.
For more fundemental hardware, such as the processor, it is a different
matter. Windows now has no support for non-x86 processors (except for
Wince, which supports a small handful of CPUs, with the hand getting smaller
with each version of Wince), whereas Linux runs on dozens of processors,
even widely different architectures. MS has got some way towards making w2k
run on the itanium, although most of the code will run in its 32-bit
compatibility mode (i.e., very slowly); linux is running fine on simulators
and prototypes and is all recompiled for 64-bit mode. MS has no plans to
support AMD's 64-bit chips, and will rely on its 32-bit mode (albeit much
faster than itanium's), whereas Linux is already well on the way towards
full support.
>
>> >> as they simply can't keep up with the needed coding changes.
>
>The careful design of Windows insures that there aren't many coding changes
>needed, if any. The driver is all you should need.
Tell that to the people who have tried to install HP CD writers in a w2k
machine, and found that the drivers conflict with other parts of the system,
sometimes resulting in total destruction of the filing system.
Remember, you are talking about an OS that lets application programs (in
particular, MS Office) totally re-arrange fundamental parts of the system.
>
>> >> Microsoft can't compete with Linux - example in the IA64
project
>> >> where HP donates code to GNU/LINUX IA64 to put it over a
>> >> year ahead of Microsoft in getting a ready OS. Linux is ready
>> >> for IA64 right now - see redhat ftp site! Microsoft will not
>> >> be ready until late next year!
>
>It's very possible that the Merced will be the biggest flop Intel has ever
had.
>Foster, on the other hand, well look out Sun :-) And you'd better believe
that
>Microsoft is following this closely with their Datacenter server and
Advanced
>Server lines. The Datacenter hardware partners have elected to pass over
Merced
>because the IA32 architecture will likely be faster than it by the time
Merced
>is released.
Remember how long it took MS to properly support 32-bit 80386 processors? I
think it was around ten years between the chip's introduction and the its
support on MS's mainstream OS.
And what about SledgeHammer? And when the next generation of IA-64, which
will be faster than Merced, is available, Linux will have been running on
IA-64 for years while MS is still messing around with prototypes. And what
about other architectures? The StrongArm, running at 1 GHz, gives 1200 MIPs
at something like 200 mW. It does not have the floating point or 3D engines
of an x86 chip, but you don't need that for a server. If someone produced
SMP chipssets for these chips, they would be ideal for servers, at about 5%
of the cost and power of equivilent x86 chips. Linux would run on it, but
the only MS product to run on Arms is Wince.
>
>> >Thats ok. Even HP has lately admitted IA-64 is a prototype and will
never
>> >actually be in production of any scale.
>> >
>> >Who wants a 750mhz box with 128k cache when they can buy 1500mhz SMP AMD
>> >boxes by Christmas and Hammer boxes at 2ghz next year?
>>
>> It is interesting that you mention the AMD SledgeHammer - a chip that
Linux
>> will fully support in all its 64-bit glory when it comes out (simulation
and
>> testing is going on now),
>
>The kernel will support it. That's like 1% of the code in a linux distro.
None
>of the apps will support it just because the kernel does. Neither will the
>drivers. All that code must be ported to 64-bit. And with the propensity
I've
>seen for a lot of linux programmers to assume an int is 32-bits wide...
Actually, almost all code that is common on linux servers works fine when
compiled for 64-bit architectures - after all, it has been running on Alphas
and 64-bit MIPS for years. The really great thing about having all the
source code available is that if someone wants to run the code on another
architecture, they just recompile. Sometimes there will be small problems -
that is why most people let a distributer like Red Hat or Turbo Linux do the
work. And as for making assumptions about int sizes, I believe you can tell
gcc what size you want to use, and thus generate 64-bit native Merced code
with 32-bit ints. Full speed code on a new architecture with very close to
zero effort.
>
>> whereas MS has no plans to support it at all. And
>> as for SMP - going the Windows route, you have to buy the expensive w2k
>> server to take advantage of two processors, whereas Linux supports SMP on
>> every system.
>
>You can tell where this guy's been getting his info! Straight from the
linux
>adovcacy sites :-) You just tell my ABIT BP6 box that Windows 2000
Professional
>(AKA the lowly desktop version) doesn't support both of it's blazing
overclocked
>366/550MHz Celerons! And the I'd better cancel my order for it's
successor --
>the VP6 and those two 750MHz P3's (which I will clock to 1GHz each) --
because
>Windows 2000 won't use them either. :-)
Judging from your reply, and those of others, I think my info regarding SMP
on w2k must be outdated. It came from before w2k was actually released, but
it did come from MS's own website. The original plan was that w2k
professional would only support 1 CPU. Then they realised that people might
be a little miffed when upgrading a dual-processor NT 4 workstation, so they
made the exception that it would allow dual-processors when you upgraded
from a dual-processor NT system. Eventually it clicked that this was a
pretty unreasonable policy, so they must have rationalised it to follow the
same scalings as NT 4.0.
>
>> For a long time, Linux has been following Windows in hardware support
>> because manufacturers did not see it as a major market, and in user
>> interfaces because there were few non-expert users. But things have
>> changed. Which OS was the first to support IDE66 and IDE100, for
example?
>
>Windows was the first to support it. All it needed was a driver.
>
>> Not Windoze, but Linux.
>
>OK. Prove it.
I read a lot of articles (almost all from general computing news sites, not
linux-specific ones), so I have no idea where I read this. But the
harddrive manufacturer (I don't even remember who it was) that started IDE66
and IDE100 got things working under Linux much faster than under Windows.
>
>> Which UIs support advanced theming? Serveral wm's,
>> and especially KDE and Gnome, have supported theming for a long time. MS
is
>> trying to catch up with themes for the windmill media player.
>
>What are you talking about? Oh themes -- try looking at Windows 98. It
has
Changing your background picture, icons and mice pointers is a far cry from
the theming abilities of KDE and Gnome. Some windows applications, like mp3
players, do theming.
>themes -- good one's too :-) KDE and Gnome were barely even around then.
I
>think it's a tad more important to support global usability features like
cut
>and paste and drag and drop before fooling with "window dressing" like
that.
I am not claiming that linux has all the features of windows, or even that
it is easier or faster to use (there are plenty of others who will claim
that). All I am saying is that linux has some features that windows does
not, and that it windows is now trying to catch up with linux *in some
areas*. I will not argue that linux has plenty of catching up to do itself
in other areas, just that even for user interfaces and ease-of-use the
balance is not as one-sided as it used to be.
>Actually I have this theory that a lot of Linux users don't really use it
for
>anything. It's just an OS for it's own sake. It makes them feel special,
>superior because they could figure out how to install it and they get a big
>charge out of the challenge of building their own software with make. All
those
>geeky build messages flow by and you end up with a program. How cool.
Start
>with text, end with program. That lost it's novelty for me a decade ago.
>
>> >>> #5. The cost. At Microsoft's current rate of inflation, by 2005 the
>> >> cost of the Microsoft operating system will be over $1,000 a
copy.
>> >
>> >Hmmm. Linux people have math problems.Besides, I see Red Hat plans to
sell
>> >people Red Hat subscriptions that will make it more expense than Win2k.
>>
>> With Linux, you can pay for what you get (support, installation help,
>> printed manuals, etc.), or you can get most of it free. Even if you buy
a
>> Linux distribution, you can still install it on as many machines as you
>> want. You have zero cost for client access to a server, unlike MS.
>
>It takes longer to develop business systems in linux, the expertise is more
>expensive and most solutions involve "roll-your-own" shell script
programming.
>The plenitude of software available for Windows means that if you want to
do it,
>someone has probably already written it. Off-the shelf. Supported. When
an
>80K UNIX sysadmin writes a script that takes him 10 hours to author, debug
and
>test. That software costs $385. The same thing's probably available in
Windows
>for $49.95.
When that script needs to be installed on 200 client machines, it still
costs $385. When the Windows program is installed on 200 client machines,
it costs $10,000 (maybe $4,000 for volume discounts). And when the Windows
program does not do exactly what is needed, and takes 2 minutes extra per
day to use, that adds up to maybe $200 extra *per day*. There is also a
steadily increasing amount of linux software available (as the OS gets more
popular, so there is more software availabe). In many cases, this is open
source, so the sysadmin can take something ready-made, modify it to suit,
and install it on as many systems as he/she wants.
>
>> >> And at that time, the US court system will break Microsoft into
>> >> two separate companies, one the OS company and one the
>> applications
>> >> company. This will cause you to have to BUY your Microsoft
>> >> Operating system rather than just have it handed to you on your
>> new
>> >PC.
>> >
>> >Well, I wouldn't bet real money a Microsoft breakup.
>>
>> It may fall apart by itself before that time comes.
>
>And you _may_ be eaten by a dragon. Slim chance in either case. BTW, what
the
>hell makes you think that the breakup would affect preloads? Dell would
just
>preload Windows .NET 2.0 (AKA Blackcomb -- it'll be available by the time
any
>break-up could happen :-) and you'll get Office components ala carte as you
need
>them from Microsoft directly.
MS is betting a lot on .NET, and not everyone thinks it will be successful.
There is also MS's business model, which relies on continually increasing
share prices and continual growth. They have reached the point where their
growth is limited in current areas (when you have a monopoly, there is not
much room for growth). They are reduced to strategies such as enforcing
licences for large corperations that makes them pay twice for their software
(I am not going to say whether this is right or wrong, just that it is a
attempt to generate more income to keep up the growth rates), and are having
trouble with falling share prices. MS is unlikely to fall apart entirely,
but there may be big changes in the next couple of years, and not all to
MS's liking.
>
>> >> #6. The upgrade problem. In not one instance, since the inception of
>> the
>> >> company has Microsoft recommended you stick with last OS's
>> >> applications
>> >> when you upgrade your OS.
>> >
>> >Everything ran pretty good on Win2K when I upgraded without changing any
>> >applications at all.
>>
>> Experiances vary widly on that.
>
>Well most folks that have made the leap are happier with Windows 2000 than
any
>other operating system that they have ever used. Analyst's agree that as
long
>as you make sure your hardware is supported you shouldn't have any
problems.
>You have posted pure FUD.
>
Well, I have read reports both ways. Most agree that w2k is the most
reliable windows yet, but there are still plenty of potential problems,
especially for upgraders. Making sure your hardware is supported is a
particular problem, and software compatibility can be problematic too (there
are software compatibility problems upgrading from NT4.0 SP3 to NT4.0 SP6,
so a leap to w2k is not going to be pain-free for everyone). If you have
huge amounts of RAM, a fast CPU, and only use MS Office then you can expect
things to go quite well.
>> >In fact, a lot of games that wouldn't work on NT now work just fine on
>> >Win2k.
>> >
>> >I know you are just envious of the fact that Microsft plans to unfork
>> >Windows with Whistler while people twiddle their thmbs waiting for the
2.4
>> >kernel to come out knowing full wwll that forking in Samba and other
>> >applications is inevitable.
>>
>> MS have planned to "unfork" Windows for 8 years, so don't hold your
breath.
>
>All in the name of hardware vendors getting up to speed and software
>compatibility. When little Johnny's CD-ROMs won't play on Windows people
get
>cranky. Give it a couple years like Microsoft have done and many of those
>problems are solved by time. There's the old adage about how to cook a
frog:
>you can't throw the frog into a pot of boiling water because he'll just
jump
>out. To cook the frog you put him in the pot of cool water and slowly turn
up
>the heat. Frog soup. In a way Millenium is turning up the heat on
vendors. It
>grafts on many of the potentially incompatible features from 2000 and
forces
>them to fix them so they will work in 2000.
>
I will believe the "unforked" windows when I see it. Given the huge
hardware requirements of w2k now, and the price tag, I think it will be a
while before they have some way to charge businesses the premium they want
while offering the same thing small enough and cheap enough for consumer
use.
>> The dozen versions of windows that are currently in common use have huge
>> differences and compatibility problems.
>
>There aren't a dozen versions of Windows in common use. The few programs
that I
>still use back from the Windows 95 days that haven't been updated run fine
on
>Windows 2000.
>
Win95, Win95 OSR2, Win95 OSR2.5, Win98, Win98SE, WinMill, WinNT4.0,
WinNT4.0SP3, WinNT4.0SP6a, Win2k, Win2kSP1 (11 varients - ok, "a dozen" was
an exageration) - there are enough differences between these that any
hardware or software manufacturer should be testing on all of these systems.
Why should people have to upgrade their application software to run on a new
OS? Remember the other old adage - if it ain't broke, don't fix it.
>> And as everyone says, 2.4 will be ready when it is ready, unlike MS
>> releases.
>
>But when Windows 2000 slipped and Microsoft said the same thing you are
about
>your precious kernel 2.4, they were a bunch of vaporware fuck-ups. FUD,
FUD,
>FUD. You folks are amazing.
MS promised release dates. Then the date slipped, and they promised a new
one. This happened again and again - it forces people (developers and
purchasers) to think that the new version is just round the corner. Had MS
been honest about the release date of NT 5.0 (as w2k was originally called),
more people would have shifted to different systems that were available at
the time rather than waiting. MS knows that people will put up with waiting
for 3-6 months at a time, but will not stand to wait 3 years. But as for
linux 2.4, no date has ever been promised. Linus has always been completly
honest - 2.4 will be ready when it works. Distributers are preparing
distributions based on 2.4, but no one will be releasing anything, or giving
release dates, until the system is ready.
>
>> >Say ... has Linus ever admitted how much stock he owns in Transmeta? ANn
>> how
>>
>> Transmeta is a hardware company which makes CPUs.
>
>They hired Linus for some reason. He doesn't design CPUs :-)
>
They hired Linus because he is an expert programmer. The Transmeta CPUs
make use of a software translation layer, and for that they need programmers
who know all about the x86 chips. They may also have hired him for prestige
and publicity reasons.
------------------------------
From: [EMAIL PROTECTED] (Bob Hauck)
Subject: Re: Claire Lynn
Reply-To: bobh{at}haucks{dot}org
Date: Sat, 21 Oct 2000 15:43:41 GMT
On Fri, 20 Oct 2000 23:48:51 GMT, [EMAIL PROTECTED]
<[EMAIL PROTECTED]> wrote:
>My field, unfortunately, is full of them. I can spot one at 1000
>yards, and yes my company keeps them away from the clients as well.
And I'm sure the "geeks" prefer it that way just as much as you do.
Why does the company keep these "geeks" around anyway? Do you suppose
that they have some skills that the compnay needs? Having nice
blow-dried hair and being able to grease customers is not the be-all of
useful skills you know. Somebody somewhere actually has to make things
work. And having that somebody out doing sales calls is not the best
use of their skills.
--
-| Bob Hauck
-| To Whom You Are Speaking
-| http://www.haucks.org/
------------------------------
From: [EMAIL PROTECTED] (Bob Hauck)
Crossposted-To:
comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy,comp.os.ms-windows.advocacy,comp.unix.advocacy
Subject: Re: Pros and Cons of MS Windows Dominated World?
Reply-To: bobh{at}haucks{dot}org
Date: Sat, 21 Oct 2000 15:43:44 GMT
On Fri, 20 Oct 2000 16:08:02 -0700, Bruce Schuck <[EMAIL PROTECTED]> wrote:
>http://www.freebsd.org/security/#adv
>http://www.redhat.com/support/errata/rh62-errata-security.html
http://www.microsoft.com/technet/security/current.asp>
Whats your point?
--
-| Bob Hauck
-| To Whom You Are Speaking
-| http://www.haucks.org/
------------------------------
From: [EMAIL PROTECTED] (Bob Hauck)
Crossposted-To: comp.os.linux.misc
Subject: Re: Is there a MS Word (or substitute) for Linux?
Reply-To: bobh{at}haucks{dot}org
Date: Sat, 21 Oct 2000 15:43:43 GMT
On Fri, 20 Oct 2000 15:17:43 GMT, Grant Edwards <[EMAIL PROTECTED]> wrote:
>Yes, it is possible to use Word the "right" way. Marking
>things by content and applying format with style sheets. But
>I've never seen anybody actually _do_ that.
I've done it. The company I work for has a set of styles for writing
specifications and test procedures. For the most part, it works
reasonably well. Making global changes is fairly clumsy, what with
having to do it through tabbed dialogs and not having a clear overview
of which styles are inherited from where, but it can be done.
Of course you're screwed if someone inserts hard formatting, since
there's no way to prevent that. Plus there's no way to change all of
your documents without loading each of them into Word. These problems
limit the usefulness of styles as compared with LaTeX or LyX or most
other packages built on top of TeX.
--
-| Bob Hauck
-| To Whom You Are Speaking
-| http://www.haucks.org/
------------------------------
** FOR YOUR REFERENCE **
The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:
Internet: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.advocacy) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Advocacy Digest
******************************
