Open source vs Microsoft vs public domain

[Malcolm McLean]

I do. [Use Open Office] I won't pay for Word for home use.

I do too. I also used it for my PhD thesis, but for a different
reason, which was that all the scientific software ran under Linux. I
think that was a mistake - I ended up making all the colour figures as
separate PDFs and interleaving them, which was a terrible job. Word
will handle complex structured documents a lot better.

However I wrote my novel (100,000 words, it's about a boarding school
set in a monastery in the Hebrides) using Open Office. A novel has
almost no structure in the typographical sense, other than a few
italics, bolds for chapter headings, and page breaks at chapter ends.
For that, Open Office was fine. It also saves as a PDF, which means I
can send non-editable versions to people, which is handy, because you
don't want more than one copy under edit at any one time.

 

[Chris H]

In message <[email protected]

I was with you, feeling the vibe, digging your sentiment, until this
last sentence. MSFT software is high quality? Since when?

Well there is another way of looking at this... due to MS giving 90% of
the world a single platform ie windows. It means that Sw is generally
developed for one (or two) targets. IE Windows or MAC or some times
Linux and Unix

Before this there were many OS and software was comparatively far more
expensive. So like it or not Windows created a de-facto desk top
standard and API for most of the worlds developers.

 

[Seebs]

Before this there were many OS and software was comparatively far more
expensive. So like it or not Windows created a de-facto desk top
standard and API for most of the worlds developers.

Several consecutive standards and APIs, *all of which sucked*.

Consider the amount of standardized working around of the 640k limit that
was done. That came to many millions of developer hours, *all completely
wasted due to stupidity*. Had Microsoft never existed, and the previous
state of affairs continued, we would have been much better off. Had someone
remotely competent at a technical level (rather than a marketing level) been
the one to provide the de-facto desk top standard, we would all be much,
much, much, better off.

If you were to compare our computing experience to basically any alternative
not involving Microsoft, people would call ours "catastrophic", and reasonably
so.

-s

 

[Malcolm McLean]

Consider the amount of standardized working around of the 640k limit that
was done.  That came to many millions of developer hours, *all completely
wasted due to stupidity*.  Had Microsoft never existed, and the previous
state of affairs continued, we would have been much better off.  Had someone
remotely competent at a technical level (rather than a marketing level) been
the one to provide the de-facto desk top standard, we would all be much,
much, much, better off.

I've used Windows and I've used X.

Unix provides a far far nicer environment for commandline tools.
However end-users
don't like the commandline.

X has its niggles and Windows has its niggles, but I wouldn't like to
say which is worse. I think the truth is that, just as the first cars
broke down all the time, weren't streamlined, killed pedestrians and
passengers in low-speed crashes because of the lack of rounding, and
so on, so the first computer windowing systems are full of
undesireable features. X was designed to run on separate servers and
clients, for instance, but in fact almost no-one uses it that way.
It's just too slow, and technology has developed in a way the
designers didn't expect. It's easy to criticise, harder to do it
yourself.

 

[Seebs]

Unix provides a far far nicer environment for commandline tools.
However end-users
don't like the commandline.
Irrelevant.

X has its niggles and Windows has its niggles, but I wouldn't like to
say which is worse.

I would. One of them was designed from the ground up to not only *allow*
people to write viruses and malware, but to *require* people to develop
that technology in order to accomplish anything. That's worse.

Imagine a language just like C, except in which file I/O could *only*
be done by writing code which overran buffers. That language would have
been a good fit for Windows.

-s

 

[BartC]

Several consecutive standards and APIs, *all of which sucked*.

Consider the amount of standardized working around of the 640k limit that
was done. That came to many millions of developer hours, *all completely
wasted due to stupidity*.

Not all Microsoft's fault.

Someone at Intel decided to shift a segment address left 4 bits before
adding the 16-bit offset. Simply by shifting 8 bits instead of 4, the memory
limit would have been 16MB instead of 1MB (well, plus adding 4 device pins
for the extra address lines).

And who was it decided to put memory-mapped video at the 640K location, IBM?

Had Microsoft never existed, and the previous
state of affairs continued, we would have been much better off. Had
someone
remotely competent at a technical level (rather than a marketing level)
been
the one to provide the de-facto desk top standard, we would all be much,
much, much, better off.

If you were to compare our computing experience to basically any
alternative
not involving Microsoft, people would call ours "catastrophic", and
reasonably
so.

After a few brief forays into Linux (where there was always some essential
component that didn't work at all, and a few more that worked badly), I was
glad to get back to Windows where it looked much more professional.

(And probably someone normally using Mac OS would say the same about
Windows.)

And while Win32/GDI was a nightmare to work with, I understand that X11
wasn't that much better...

 

[Seebs]

Someone at Intel decided to shift a segment address left 4 bits before
adding the 16-bit offset. Simply by shifting 8 bits instead of 4, the memory
limit would have been 16MB instead of 1MB (well, plus adding 4 device pins
for the extra address lines).

And who was it decided to put memory-mapped video at the 640K location, IBM?

I used machines based on the same hardware, which were compatible enough
to run some PC software, which didn't have that limitation.

After a few brief forays into Linux (where there was always some essential
component that didn't work at all, and a few more that worked badly), I was
glad to get back to Windows where it looked much more professional.

How do you configure a printer that uses TCP/IP?

Hint: It's not a "network" printer. It's a local printer directly connected
to this machine. Using one of COM1, LPT, or... "standard TCP/IP port".

Windows was fairly good at *looking* professional. But there were a couple
of years during which plugging a USB mouse into a NetBSD machine worked
nearly instantaneously and quite reliably, and plugging a USB mouse into a
Windows machine might or might not work at all, and if it did it took ten
seconds or longer to identify and install drivers.

(And probably someone normally using Mac OS would say the same about
Windows.)

To some extent, this is certainly true, but baby duck syndrome is not nearly
sufficient to explain some of these things. No amount of Mac users growing
up with it made the Classic MacOS "you must have N+1 megabytes of disk space
for backing store to add 1MB of virtual memory to a system, and most
applications require a lot more memory if you don't have virtual memory
enabled" design decision rational or justifiable. It was broken, whether by
design or otherwise.

And while Win32/GDI was a nightmare to work with, I understand that X11
wasn't that much better...

Perhaps it wasn't, but you could always replace it.

The thing is... Once you get past the initial baby duck syndrome and
not-used-to-that, and start looking at the documentation and writeups by
experienced professionals who really do like a given system... Windows
loses. By a gigantic margin. It's genuinely, objectively, awful. It
is unstable, insecure, and actively hostile to long-term code maintenance.

If you look at a Windows app, you can tell when it was written by the APIs
it is built for, because they get replaced every few years. There's Unix
apps I'm still using now that were last significantly updated in the early
90s. Anything new enough to have prototypes typically works.

There's totally some baby-duck syndrome involved, but... There's also some
real differences in design philosophies, and in the long run, differences
in design philosophies *matter*.

Think about the stuff sandeep keeps proposing doing to C. Windows is an
operating system where people like him were given carte blanche.

-s

 

[Marcin Grzegorczyk]

I used machines based on the same hardware, which were compatible enough
to run some PC software, which didn't have that limitation.

*What* same hardware? 8086-compatible? If so, they certainly had to
have at least a 1M limit, which I think you agree is not a big
improvement over 640K.

And yes, it was IBM who invented the 640K limit in their PC design.
Really, I thought you knew better than trying to blame Microsoft for that!

 

[Seebs]

*What* same hardware? 8086-compatible? If so, they certainly had to
have at least a 1M limit, which I think you agree is not a big
improvement over 640K.

It was a bit over a 50% improvement.

And yes, it was IBM who invented the 640K limit in their PC design.
Really, I thought you knew better than trying to blame Microsoft for that!

Fair enough, but with a saner design, it might have been less of a problem.
Why did so many people have to develop programs to reorganize memory, when
the OS could have done it right in the first place?

-s

 

[ImpalerCore]

I used machines based on the same hardware, which were compatible enough
to run some PC software, which didn't have that limitation.


How do you configure a printer that uses TCP/IP?

Hint:  It's not a "network" printer.  It's a local printer directly connected
to this machine.  Using one of COM1, LPT, or... "standard TCP/IP port".

Windows was fairly good at *looking* professional.  But there were a couple
of years during which plugging a USB mouse into a NetBSD machine worked
nearly instantaneously and quite reliably, and plugging a USB mouse into a
Windows machine might or might not work at all, and if it did it took ten
seconds or longer to identify and install drivers.

It doesn't matter how well Windows was designed or not from a
developer standpoint. Windows won because it had the environment with
the best word processors and spreadsheets, which in my opinion was the
main motivator for the general public to start buying a computer in
the first place. Compare the options on Windows at the time and those
on UNIX and it's no contest. In college at the university UNIX labs I
had access to what, pico, emacs, and vi, and there wasn't any
spreadsheet software; 'xv', matlab, and netscape navigator were the
cool apps. Windows had Word and Excel, which made up the bulk of my
work. Windows has been carried by that inertia from those apps for a
long time.

If some of the early UNIX community were greedy little Scrooges
instead of a bunch of nerds, maybe they would have seen that at the
beginning, all the world was a word processor or a spreadsheet.

Best regards,
John D.

 

[Malcolm McLean]

It doesn't matter how well Windows was designed or not from a
developer standpoint.  Windows won because it had the environment with
the best word processors and spreadsheets, which in my opinion was the
main motivator for the general public to start buying a computer in
the first place.

Microcomputers were originally toys, useful for nothing except hobby
programming and playing games.

Everyone loves games. When IBM launched a micro, suddenly business
owners had an excuse to buy a machine to play games on. And, to be
fair, you could also type letters on it. Then came Lotus and the
spreadsheet, and that was actually useful to the financial types.
IBM offered 2 operating systems, an expensive, fully-featured one, and
a cheap one. Most people opted for the cheap one. Because there's such
an advantage in having the same operating system as everyone else, the
expensive OS died.

Windows came a lot later, and had to beat rivals. However Microsoft
always had the advantage because they also made the operating system.
For quite a long time you needed to run some programs under Dos,
others under Windows.

 

[Chris H]

In message <[email protected]

It doesn't matter how well Windows was designed or not from a
developer standpoint. Windows won because it had the environment with
the best word processors and spreadsheets, which in my opinion was the
main motivator for the general public to start buying a computer in
the first place. Compare the options on Windows at the time and those
on UNIX and it's no contest. In college at the university UNIX labs I
had access to what, pico, emacs, and vi, and there wasn't any
spreadsheet software; 'xv', matlab, and netscape navigator were the
cool apps. Windows had Word and Excel, which made up the bulk of my
work. Windows has been carried by that inertia from those apps for a
long time.

If some of the early UNIX community were greedy little Scrooges
instead of a bunch of nerds, maybe they would have seen that at the
beginning, all the world was a word processor or a spreadsheet.

The PC+ Windows made computing mass market and affordable. It gave a
large standard platform for developers. Thus the cost of software
declined to an affordable level. This was the break though.

 

[ImpalerCore]

Microcomputers were originally toys, useful for nothing except hobby
programming and playing games.

Everyone loves games. When IBM launched a micro, suddenly business
owners had an excuse to buy a machine to play games on. And, to be
fair, you could also type letters on it. Then came Lotus and the
spreadsheet, and that was actually useful to the financial types.
IBM offered 2 operating systems, an expensive, fully-featured one, and
a cheap one. Most people opted for the cheap one. Because there's such
an advantage in having the same operating system as everyone else, the
expensive OS died.

I was a wee lad during that time, but growing up if you wanted games,
you bought an Atari, a Commodore 64, or a Nintendo. I do have many
fond memories of my dad's TI-99 and the many games I eventually played
with it, some from copying code, and it's likely why I got into
computers. At the consumer level, why spend money on a PC for gaming
when you could spend the money on a Nintendo or Genesis.

Windows came a lot later, and had to beat rivals. However Microsoft
always had the advantage because they also made the operating system.
For quite a long time you needed to run some programs under Dos,
others under Windows.

Let me make a substitution.

X Windows came on the scene, and had to beat rivals. However UNIX
always had the advantage because they also made the operating system.
For quite a long time you needed to run some programs under 'sh',
others under X Windows.

From what I understand, X11 and the OS/2 project were around at the
same time. They both had the opportunity to establish themselves as a
front-runner. In my opinion, I don't think the idea of using the PC
as a gaming platform was what led to Windows overtaking the
competition, not with Nintendo like gaming platforms running around.

X11 - 1987, academia oriented
OS/2 - 1987, business oriented
Windows - consumer + business + academia

Creating a platform by itself does not create a mass-market for it
(X11 is a testament to that). Something needs to be made to pull that
platform into the collective eyes of the mass market. My opinion is
that it was the Windows word processor apps (like Word, and eventually
Office as a whole) that attracted the consumer populace to actually
make the PC+Windows platform widespread enough to make it a cheaper
platform to develop software to reach a wider audience.

If Seebs had access to time travel for a minute and wanted to rewrite
history to eliminate Windows and make UNIX king, my best suggestion
would be to convince the X11 group back in 1985 to make the equivalent
of Word. I don't know if there was anyone specifically influential
enough to actually make it a reality (Seebs may be more convincing if
he's in a Delorian and wearing Doc's sunglasses).

Best regards,
John D.

 

[Rui Maciel]

It doesn't matter how well Windows was designed or not from a
developer standpoint. Windows won because it had the environment with
the best word processors and spreadsheets, which in my opinion was the
main motivator for the general public to start buying a computer in
the first place.

Microsoft sold Microsoft Word for the Apple Macintosh since 1984, along
with another half dozen platforms. Microsoft sold a Excel version for the
Apple Macintosh since 1985. Adding to that, and contrary to popular
notion, Microsoft Word was never good. A good proof of that was the
dominance of other word processors such as WordPerfect.

And anyone who uses/used Microsoft Word and has any experience with other
document preparation systems can quicky tell you how poor Microsoft Word
fairs when compared with alternatives. Even TeX and LaTeX, tools which
are pushing 30 years, are constantly better than what Microsoft Word has
to offer.

Compare the options on Windows at the time and those
on UNIX and it's no contest.

Microsoft sold Word for a couple of UNIX platforms.

In college at the university UNIX labs I
had access to what, pico, emacs, and vi, and there wasn't any
spreadsheet software; 'xv', matlab, and netscape navigator were the
cool apps. Windows had Word and Excel, which made up the bulk of my
work. Windows has been carried by that inertia from those apps for a
long time.

If some of the early UNIX community were greedy little Scrooges
instead of a bunch of nerds, maybe they would have seen that at the
beginning, all the world was a word processor or a spreadsheet.

If you try you can make your trolling less obvious.


Rui Maciel

 

[Rui Maciel]

Creating a platform by itself does not create a mass-market for it
(X11 is a testament to that). Something needs to be made to pull that
platform into the collective eyes of the mass market. My opinion is
that it was the Windows word processor apps (like Word, and eventually
Office as a whole) that attracted the consumer populace to actually
make the PC+Windows platform widespread enough to make it a cheaper
platform to develop software to reach a wider audience.

Your opinion isn't compatible with the facts. There were quite a lot of
word processors and spreadsheets back then, including for non-windows
platforms. What forced microsoft windows onto the people was the fact
that, due to Microsoft's partnership with IBM, MS-DOS and MS Windows
copies were bundled with IBM's computers and it was cheap and easy to
copy, distribute and install them in "IBM PC compatible" systems.


Rui Maciel

 

[ImpalerCore]

ImpalerCore wrote:


If you try you can make your trolling less obvious.

You're right. I'm definitely not as subtle as 'lccwin fails to
generate NOPs'.

 

[Chris H]

And anyone who uses/used Microsoft Word and has any experience with other
document preparation systems can quicky tell you how poor Microsoft Word
fairs when compared with alternatives.

That is not true.

Even TeX and LaTeX, tools which
are pushing 30 years, are constantly better than what Microsoft Word has
to offer.

That is not true either or TeX would have a far better take up than it
does.

 

[Seebs]

That is not true.

Right you are. Many people can't tell.

That is not true either or TeX would have a far better take up than it
does.

What makes you think that "take up" follows any particular objective measure
of quality?

FWIW, there's a fair bit of stuff that uses TeX under the hood. YMMV. In
all my years of using TeX, though, I've never had a recurring problem where
data got lost because if you had two windows open, it could wedge itself
into a state where neither window, nor any other part of the program, could
receive any form of user input, nor could it save or be told to quit
gracefully. That was, by contrast, a fairly *typical* failure mode for
the last version of Word I used.

-s

 

[Chris H]

Right you are. Many people can't tell.

Well I have used many WP's and Word is fine for me, In fact it's
nearest rival is Open Office not TeX so the system must be good.

What makes you think that "take up" follows any particular objective measure
of quality?

That is true.

FWIW, there's a fair bit of stuff that uses TeX under the hood. YMMV. In
all my years of using TeX, though, I've never had a recurring problem where
data got lost because if you had two windows open, it could wedge itself
into a state where neither window, nor any other part of the program, could
receive any form of user input, nor could it save or be told to quit
gracefully. That was, by contrast, a fairly *typical* failure mode for
the last version of Word I used.

I have not had any problems with word. At least no more than any other
office software and I used Windows, MAC and Unix. MS Word on the MAC is
very nice. I have tried TeX several times but did not find it very
usable. IT may be better but it is not easier to use.

 

[Malcolm McLean]

I have not had any problems with word. At least no more than any other
office software and I used Windows, MAC and Unix. MS Word on the MAC is
very nice.  I have tried TeX several times but did not find it very
usable. IT may be better but it is not easier to use.

There's always a tension in software between power and ease of use.

Scripts are very powerful, but if you go too far in the scripts
direction you end up producing a programming language rather than a
user application. If you add lots of buttons and menus, it's easy for
the casual user to get lost, or to put the software into a state he
doesn't understand. If you cut down the options, there's always
something conceptually trivial which the software won't do.

Microsoft have the advantage of huge resources to invest in usability
labs. So Microsoft software is typically quite easy for the casual
user to pick up and get core functionality out of the package. However
it also has lots of options hidden away for advanced users.
The problem is that the transition from novice to advanced user isn't
always easy. I had two days of training on "Word for long documents",
for instance. However by the time it came to write my thesis I'd
forgotten most of it. It's possible to set Word up to automatically
produce contents pages and link them to numbered paragraphs, keep
track of references, and so on, but really it's a job for a
professional secretary.