Linux-Advocacy Digest #251, Volume #34 Sun, 6 May 01 11:13:05 EDT
Contents:
Re: IE (Bob Hauck)
Re: Performance Measure, Linux versus windows (Charles Lyttle)
Re: Justice Department LOVES Microsoft! (Rick)
Re: Is StarOffice 5.2 "compatible" w/MS Office 97/2000? (Bill Vermillion)
Re: Richard Stallman what a tosser, and lies about free software (David Vaughan)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Bob Hauck)
Crossposted-To: alt.linux,alt.destroy.microsoft
Subject: Re: IE
Reply-To: bobh = haucks dot org
Date: Sun, 06 May 2001 14:43:50 GMT
On Sun, 6 May 2001 13:31:38 +0100, Michael Pye <[EMAIL PROTECTED]> wrote:
> "Bob Hauck" <[EMAIL PROTECTED]> wrote
> That's why I think there should be a complete overhaul of HTML and all it's
> components and add-ons.
The big question is how to get from here to there. It would have been
easy in '93, now it will be pretty hard. The path of least resistance
would probably be a plugin that implements your new protocol.
> Perhaps such web sites should be written in a different language or subset
> of XML, and purely functional pages written in HTML, be we need a new
> language because of the many available which would do the job (pdf, doc,
> flash etc) none are human readable or as compact as HTML.
TeX maybe? Or maybe a specialized DTD for XML.
> That depends. Some people do want to see a good looking web, other don't.
It isn't that people don't want to see a good looking web, it is that
most of them are more concerned with other issues. They don't have
broadband or the latest browser/plugin/player combination. They do not
approach the web in the same way as a magazine or TV.
> The main reason flash is unpopular at the moment is because the files are
> too large and the interpreter too slow. I hate it as well, but eventually,
> when everybody has a broadband connection and at minimum a 1GHz processor
That may be a while, given the slowing rate of replacement of PC's.
The broadband part might be even further off, what with the providers
dropping like flies and all.
> Perhaps the inherent simplicity of HTML is one of it's major downfalls.
It is a two-edged sword. Low cost of entry, but not very powerful in
terms of presentation. Deliberately so, because lots of people were
still using text terminals at the time. Gopher was the main competition.
> That's what annoys me in many ways. After producing something crap, netscape
> just sat back for two and a half years and let their previous reputation
> spread their piece of shit browser all round the web.
The _did_ work on Mozilla. From what I understand, the problem with NS4
was that it had reached a state where further development was going to
be exponentially more difficult. Yeah, they shouldn't have let that
happen, but then they were in the process of being put out of business
by a much larger competitor. I'm sure that had something to do with it.
Unfortunately, Mozilla took longer than anyone thought it would. I
think some of that was because of trying to keep all the features of NS4
rather than just making a browser. But that is only my opinion.
> Mozilla is great, but it is huge and slow. There really is no need for the
> kind of power gobbling stuff that is in both NS6 and Mozilla.
I agree. There are a number of projects to use the Mozilla rendering
engine in a lighter-weight browser. Don't know if any work on Windows,
but there are at least two for Linux.
> Opera is the only reasonable one for working within a system spec and
> that costs money, something a browser can't really afford to do in a
> market where EVERY alternative is free...
Opera is free if you don't mind the ads. OTOH, I liked Opera enough to
pay for it. Price is less an issue than bundling though. Not being
bundled with anything will prevent it from gaining a majority share.
> What is this Konqueror? I would like to take a look as I still haven't found
> an up to date browser I can run on my P200!
It is the browser for KDE. It (KDE 2.1.1) is a bit slow to start on a
P-120 but runs ok after starting. It starts and runs well on a P-300.
Haven't tried it on a 200. You have to be on a Unix-y OS though.
> > Perhaps, but if your site has useful content then people will come back
> > even if it is not flashy.
> But if we are talking about e-commerce sites here, they have to look and
> function better than their competition too. Just like in the real world,
> they have an image to project of themselves in order to impress clients.
I think the key words there are "function better". Looking good is less
important than function. Glitches in the software will cause people to
abandon the transaction *after they have made the decision to buy*. You
are just throwing away good money if that happens.
> > Sure. I am far more interested in the content than in the presentation
> > of that content though.
> Perhaps is was, but things have moved on... I enjoy both parts of the web.
Some sites are all about presentation and design. Those are perhaps the
ones that have a legitimate gripe with HTML. People interested in such
things would probably be more likely to be willing to download a plugin.
> > Even e-commerce sites are not about flash or presentation. If I go to
> > Amazon, I am already sold on buying something, usually something in
> > particular.
> Perhaps you are sold on something. But many people aren't as decisive. They
> visit to browse the bargains, just as you might do in the high street.
Right, and that is why Amazon puts all of those implulse items on the
screen when you are shopping. Maybe one will catch your eye and they
will make an extra sale. However, what makes you go to Amazon in the
first place is that you want to buy a book (or a car or a house or a
bathtub or...). You don't just accidentally wander in like you might a
real store. I really doubt that people browse online stores the way
they do brick-and-mortar shops.
I could be wrong though. People watch the shopping channels on TV after
all. I've always been amazed by that.
> > It is not fundamentally important that the site look to the user
> > exactly as envisioned by the designer.
> But many people want to provide precise layout and that is why HTML has
> become so messed up...
Oh, I understand that. I just question whether that is the right
approach to the medium.
--
-| Bob Hauck
-| To Whom You Are Speaking
-| http://www.haucks.org/
------------------------------
From: Charles Lyttle <[EMAIL PROTECTED]>
Subject: Re: Performance Measure, Linux versus windows
Date: Sun, 06 May 2001 14:44:56 GMT
I agree with pretty much everything you wrote. My statement was intended
to be just an oversimplified remark about the logic of benchmarking. Any
published benchmark has to be treated with a great deal of suspicion.
There is probably a hidden agenda in it somewhere and, either
deliberately or not, false assumptions built in. When testing something
as high level as an OS, it is almost impossible for anyone outside the
testing lab to find them.
Weevil wrote:
>
> Charles Lyttle <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > If a benchmark can be rigged to prove anything, then benchmarks are
> > crap.
> > I can rig benchmarks to prove anything.
> > Therefore, benchmarks are crap.
> >
> > A fundamental element of logic is that a false assumption can be used to
> > prove anything.
> > By making a false assumption when writing a bechmark, a benchmark can be
> > used to prove anything.
>
> No, not really. You can rig a benchmark to make it appear that it supports
> your case, but you cannot rig a benchmark to prove your case. If you're
> "making a false assumption when writing a benchmark," and it is this false
> assumption that makes it appear to support your case, then you haven't
> proven anything at all. You've merely made a mistake (or lied, as the case
> may be).
>
> A similar (atcually identical) misconception is that statistics can be
> massaged so that they prove anything. Not true, of course. Statistics do
> not lie: they say what they say, no more and no less. What muddies the
> waters is that statistics and benchmarks can be interpreted incorrectly, and
> it is often extremely difficult to tell when this is being done. So they
> get abused all the time.
>
> What they do not do is prove anything that is not true. That is impossible.
>
> > The only benchmarks that count are those you write to test your
> > application. And they are meaningless for anyone else.
>
> Yup. I agree , oh, 85% or so. Statistics and benchmarks are best when you
> can compile (no pun intended) them yourself, but other people's numbers can
> be very enlightening and useful. You just have to be really careful that
> you're getting the whole, accurate story.
>
> > --
> > Russ Lyttle
> > "World Domination through Penguin Power"
> > The Universal Automotive Testset Project at
> > <http://home.earthlink.net/~lyttlec>
>
> --
> Weevil
>
> ----------------------------------------------------------------
>
> "The obvious mathematical breakthrough [for breaking encryption schemes]
> would be development of an easy way to factor large prime numbers."
> -- Bill Gates
--
Russ Lyttle
"World Domination through Penguin Power"
The Universal Automotive Testset Project at
<http://home.earthlink.net/~lyttlec>
------------------------------
From: Rick <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.ms-windows.nt.advocacy,comp.sys.mac.advocacy,comp.os.ms-windows.advocacy
Subject: Re: Justice Department LOVES Microsoft!
Date: Sun, 06 May 2001 10:54:27 -0400
Ayende Rahien wrote:
>
> "Rick" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > Ayende Rahien wrote:
> > >
> > > "Roy Culley" <[EMAIL PROTECTED]> wrote in message
> > > news:[EMAIL PROTECTED]...
> > > > In article <dJZI6.6119$[EMAIL PROTECTED]>,
> > > > "Daniel Johnson" <[EMAIL PROTECTED]> writes:
> > >
> > > > > I *strongly* suggest that printing *must* be
> > > > > addressed as soon as possible. I honestly see
> > > > > very little movement on that front.
> > > >
> > > > I don't understand this printing problem that Unix supposedly
> > > > has. Unix apps normally produce postscript. Unix lpr uses filters to
> > > > see what the data type to be printed is. If you don't have a
> > > > postscript printer the filter will use ghostscript to convert it to
> > > > pcl or whatever. I now have a Lexmark z52 which has its own Linux
> > > > driver. To me it is totally transparent. Before I had an
> > > > HP850C. Again printing was totally transparent. Where is the problem?
> > >
> > > First, there is a news server that is dedicated to printing in Linux,
> that
> > > is a problem all in itself.
> > > Second, we aren't talking about end users' problems here, we are talking
> > > about support for developers.
> > > Let's say that we take two comparable products that does the exact same
> > > thing, one for windows, the other for linux.
> > > The windows one could have printing support in a matter of no time,
> because
> > > GDI abstract the output device from the developer, so you don't have to
> > > change the code at all to print to a printer or display on the screen.
> > > The only worry you have is page breaks.
> > >
> > > On Linux, however, you need to have seperate code that does it, which is
> > > much harder than adding few "if"s for page-breaks.
> >
> > Gee, I thought all I had to do was add a printer drive, and all a vendor
> > had to do was write a driver. Much like a printer driver is written for
> > any other OS.
>
> On user side, maybe, I really didn't deal with printing on linux on the user
> side.
> I'm talking on the *developer*'s side here.
> There is no abstraction on linux of the printer. In Windows, I don't *care*
> what printer it is, or if I print to printer or screen.
> The only thing that I need to change in my code to change displaying on the
> screen to printing is page breaks. On Linux, I would've to do all the
> formatting myself, and create a PostScript output, which is good for
> printers that support it, but what about those that don't? The other option
> is to support *myself*, every brand of printer in existance.
>
I have an Epson 740. It doesn not support Postscript. Thats why I have
Ghostscript.
> Again, it's *not* a user-side issue, it's a developer side issue. If mean
> that it's far easier to print your application output on Windows than on
> Linux.
--
Rick
------------------------------
Crossposted-To: comp.unix.advocacy,alt.solaris.x86,comp.unix.solaris
From: [EMAIL PROTECTED] (Bill Vermillion)
Subject: Re: Is StarOffice 5.2 "compatible" w/MS Office 97/2000?
Reply-To: [EMAIL PROTECTED]
Date: Sun, 6 May 2001 14:20:41 GMT
In article <[EMAIL PROTECTED]>,
T. Max Devlin <[EMAIL PROTECTED]> wrote:
>Said Bill Vermillion in comp.os.linux.advocacy on Sat, 5 May 2001
>>In article <[EMAIL PROTECTED]>,
[older attribs between Max and I removed. wjv]
>>Xenix was not just on a PC platform. Seeing something like an
>>Altos with an 8086 processor, 386K of RAM, a 25MB HD, and 5
>>terminals and a couple of printers, running a business
>>successfully, shows just how efficient it was.
>That far back, what was or was not "a PC" was imprecise enough that
>this box might well count. Though I would have to admit that having
>5 terminals would disqualify it. Still, running Unix on anything
>approaching an 8080 series processor was not going to happen unless
>Bill got a cut, in his mind.
I'd have to go back and search but I'm fairly certains Gates'
involvement with Xenix pre-dated the IBM/PC-DOS affair.
> [...]
>>Of course the DEC platform had more user seats - if you based that
>>on totals - but Tandy has more physical machines running.
>See, these are "platforms". The PC was not such a platform, which
>is why the definition dissolves away if you're not careful. Would
>MS have loved to crush DEC with Xenix-based systems? No, they
>preferred to just slowly wedge them out with DOS-based systems.
>Bill was a big BASIC advocate, remember. He probably still doesn't
>know squat about the Bourne shell.
The PC - as we know it - did not exist. In that era the ONLY way
you could call a system Unix was if you compiled the source tape
from AT&T and made no changes. I friend of mine showed me some
docs a very long time ago. The basic instructions were to mount
the tape on your Vax, go through some startup procedures, start the
compile, and come back in three days when it was done.
Because of the extreme AT&T restrictions on the use of the name
UNIX - all of the licensed and modified variants had different
names. There was usually a thread [about 15 years ago] on the 'net
about the 'names of Unix'. The largest list I recall had about 80
names. Cromix - from Cromemco. Xenix - from MS. Serix ?? from
Siemans, Venix ??, and on and on, almost all with an *IX in their
name.
>>The Tandy Xenix [they did their own port on the 1.x] was what got
>>me hooked in the Unix world 18 years ago. No one ever considered
>>a Unix system a viable entity on a PC system in those days. It was
>>not until the '386 - where you could forget the 64K paragraph
>>limitations - that made *ix really workable on that architecture.
>Was it really and truly just this "64K paragraph limitation"? What is
>that?
Part of the limits of the original 8088/8086. I should have said
pages, sorry about that. You could have a 16 byte offset in each
page - so their were 16 byte paragraphs in 64 byte pages.
That gave you the 'models' for the compile. Small model where
both the program and data resided in a 64K block. Medium with one
64K for program and one 64K for data. Large with one 64K for
program and many 64K blocks for data, and huge with multiple blocks
for program and data. I have NOT been near any of that stuff for
about 15 years - nor do I want to revisit it. If this is wrong
anyone please feel free to correct it.
>>The Altos used the 8086 - not the 8080 - so at least is was 16
>>bits across the board, not the hybrid 8/16 in the 8088.
>I'll tell you I'm certainly not interested in such low-level
>details, in general. Don't let me stop you; I know there's
>an audience, and we're cross-posted. I can appreciate them,
>when they're worth appreciating, but most of the time they're
>meaningless trivia.
Well I only read the comp.unix side. Been hooked on the 'net since
I got access in '85.
>>Xenix was also the orignal OS in the Apple Lisa - which you had to
>>use to write and OS for the Mac, as the Mac was so limited it
>>couldn't do anything as complex as compiling an OS.
>To be honest, I find this less then credible.
Well you may find it less than credile, but Apple was selling the
Lisa to the developers. The first Mac had a 128K memory limit.
I remember when Apple demo'ed the Lisa at our computer club.
$10,000. Had a 5 or 10MB hd. I forgot the memory size. My first
impression was that it was a lot like the Xerox Star - which Xerox
had shown us a year or so earlier, along with a system they had
invented to be able to wire all the computers in a single building.
They called that "Ethernet". :-) [that stuff surely seemed
complicated then]
I have some price sheets from 1984 showing the SCO price list for
Xenix on a Lisa. [SCO was a porting house then and made a lot of
cross-compilers].
>>>B) To suck enough as a Unix on a PC that they didn't have to work
>>>that hard and could get away with easy, crapware, DOS.
>>We can't totally eliminate point A - Gates wanting to monopolize
>>the OS - but at that point in time MS was primarly a language house
>>and Xenix was the ONLY OS they produced. They did have the highly
>>successful Z80 add-in card for the Apple so the Microsoft licensed
>>version of CP/M was a big hit in the Apple world.
>This time, it is the definition of "OS" which turns to sand,
>historically. I know, the term itself is not variable, but the
>ROM BASIC was 'an OS' to Gates; DOS itself only happened on IBM's
>insistence. (I say happened because "created" would be overly
>generous; it was plagiarized more than it was created, as was BASIC
>and Xenix, of course.)
And the OS part of DOS was stretching it a bit IMO. It was really
more of a 'file-handler' than an operating system. The very first
PC's [I got to try out DOS 1.0 on a PC about 6 weeks after they
were introduced and I was quite un-impressed. 160K drives when
everyone else had 360K. 64K memory limit. A warmed over CP/M and
the screen display method made the sucker really creep compated to
the memory mapped screen displays in most of the other CP/M
systems].
>>We can totally elminate point B - 'suck enough on a PC ..' because
>>what we call the PC [the iNTEL based platform] had not been
>>introduced at that time.
>Are you sure we're talking the same Microsoft? BASIC was there
>first product Microsoft? I didn't realize Xenix was that old. Just
>where did it come from?
BASIC from MS came out in 1975 with the Altair. The ROM'ed BASIC -
based on MS's 4.5 BASIC as I recall - was what ran Radio Shack's
Model One computer. {I bought the 4th one sold in Orlando}. The
Apple variant was poorer in many respects. I have the original
Apple Basic and OS ?? manuals. I bought those because I thought
the color was really neat on the Apple. I paid $50 for the
manuals. One of the better investments in computers, as I read
them and saved myself from making [for me] a very expensive
mistake.
>>I remember Gates ranting and raving and calling all hobbyiest
>>thives for 'stealing' his $150 cassette basci [highly over-priced
>>in the 1977 world] while the competition was in the $35 to $50
>>range.
>And Xenix already existed? Wow. [He wasn't even selling the same
>thing as 'the competition', I'll bet, licensing where they actually
>*sold*.]
Xenix came along after his cassette BASIC. This was back when
Micro Soft [the orignal name was two words] was still located
in Alberqureque [sp?] New Mexico. There are classic Gates quotes
from that era such as "There is no way a company can make
$10,000,000 selling software, and I have the figures to prove it".
A year or so later he bumped the $10,000,000 to $100,000,000.
That was about the time he re-wrote the contracts with he an Balmer
and cut Balmer's share to a much smaller point. He's always been
ruthless that way. Stories seem to indicate he made all his money
in his short college career by playing poker.
I don't know the exact year MS introduced Xenix. The worst Xenix
[buggiest] that I worked on was IBM Xenix 2.0. It was their first
'supported' version and IBM Xenix 1.0 was released totally as is.
2.0 was so buggy I'm surprised it was released as supported. I was
writing something for some radio modems for pipeline monitoring
software. I had code that worked on my Televideo running
MicroPorts V.2 Unix on a '286, and the code would NOT work on the
IBM. In places where you were to 'and' the values in the IOCTL -
IBM's version totally replaced the values.
In the 'creat' function, the program did not take the values you
used, but 'and'ed them with the existing values, so you
specifically had to set umask to 000 before entering the program.
I'm glad I don't work at that level anymore.
>>>>Then he had an 'office suite'. Multimate, Multiplan, and Multi???
>>>>Then IBM knocked on the door and things changed.
>>>Office didn't happen until after Win3.1,
>>"Office" by that name didn't happen until Windows but the 'office
>>suite' concept existed in Xenix. MultiMate [Word Processor],
>>MultiPlan [spread sheet] and Mulit ???? [I can NOT remember what
>>that app was] came long before even PC-DOS 2.0
>This was definitely not Microsoft's creation. MultiMate was
>Ashton-Tate, IIRC. So I presume Xenix came from somewhere other than
>MS, as well. Where?
It may have been licensed. There were three "Multi" packages. I
see an RS MultiPlan manual high on shelf over there - and too hard
to get too - so I can't look further.
Xenix was the MS license of Unix from AT&T. [See above comments].
If it says Xenix on it, MS gets royalties.
>>> Previously, the app space was only monopolized by directly
>>>killing off competition with Windows (Lotus, Wordperfect, dBase).
>>>Multimate was Ashton-Tate's (dBase) word processor, IIRC.
I remeber being at Comdex when Windows was first announced. Took
it two years to actually get out. Gawd what a pile of crap.
Windows 2.0 [the only one I remember using at all on an IBM]
painted the word "Windows" in crude ASCII blocks.
>>I had frined running Vulcan - the predecesor to dBaseII - on their
>>CP/M machines. That dates to about 1978 - when MS was only
>>producting Cobol, Fortran, Basic and a couple of other languages -
>>primarily in the CP/M market, and licensing ROM verison of MS Basic
>>to Commodore, Apple, Tandy, et al.
>I was going to say that, from what I've read, "producing" would be
>over-stating the case (Microsoft is/was notorious for announcing
>products they didn't actually have yet, or ever) but then I realized
>that your "producting" might not be [sic]. Was that intentional?
Oh yes. That's the typical MicroSoft tactic. The book "Start Up"
- about he life and death of Go computers - reveals that all too
well. Go had an OS which recognized handwritting. MS announced
Pen???. So all the potential Go customers [and State Farm was
going to equip all agents with them] waited.
As to producing, MS had written their versions of the above
languages. When the PC first came out they cross-compiled their
8080 code to the 8088/8086. Made for some very sluggish
performance as I recall. I don't know how long it took them to
write native code. In those days everything was typically written
in assembler - so cross-compilers were the norm - and not every
efficient.
MS bragged it only took them $7 million to kill a $40 million
company. [I'm not sure on the exact figures but you get the idea.]
This never came up in the government anti-trust investigation as it
was long gone by that time.
Looking back no one would have guess where we'd be today. It's
sort of interesting to look back at this [my secondary major was
history in college so that may account for my fascination with the
past]. I'd surely NOT like to go back there. I try to keep
history from being re-written to badly [as far as my memory
permits] but I'm not one who builds old computers, or wants to go
back. I'm having enough trouble just keeping up with a small
subset of today's world.
--
Bill Vermillion - bv @ wjv . com
------------------------------
Crossposted-To: gnu.misc.discuss,comp.os.ms-windows.advocacy,misc.int-property
From: David Vaughan <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Subject: Re: Richard Stallman what a tosser, and lies about free software
Date: Sun, 06 May 2001 14:58:45 GMT
I dont see how you can say the GPL keeps us from using the code in useful
ways, they simply ask us to share our coding efforts. This prevents
reinventing the wheel every time something needs updated and the original
author cant/wont work on it.
just my $0.02 worth
On 14 Mar 2001, Sam Holden wrote:
> On Wed, 14 Mar 2001 05:38:10 GMT, Les Mikesell <[EMAIL PROTECTED]> wrote:
> >
> >"Sam Holden" <[EMAIL PROTECTED]> wrote in message
> >news:[EMAIL PROTECTED]...
> >wrote:
> >
> >Of course the original authors have the right to prevent others from
> >being able to use the code in many ways that would be useful
> >to others, as the GPL restrictions do. I don't think anyone questions
> >that right. The question is, why does anyone else consider this to be
> >a good thing and how can they mention it in the same sentence with
> >freedom or sharing?
>
> Because they look at it from the users point of view, not the developers
> point of view. I'm not going to argue the case again - there have been enough
> posts and www.gnu.org has enough explanation of the rationale behind the GPL.
>
> If you think that non-free software is OK, then you will not agree with
> the GPL and it's rationale. Since that is the basis of it.
>
> No amount of arguing over definitions of free is going to change that.
>
> --
> Sam
>
> PC's are backwards ... throw them out! Linux is ok though.
> --Rob Pike (on the subject of CR/LF etc)
>
>
------------------------------
** 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 by posting to comp.os.linux.advocacy.
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
******************************
