Linux-Advocacy Digest #304, Volume #26 Sat, 29 Apr 00 01:13:05 EDT
Contents:
Re: linux as Netscape platform (R.E.Ballard ( Rex Ballard ))
Re: The truth is often painful... (Bob Hauck)
Re: linux as Netscape platform (Christopher Browne)
Re: "Technical" vs. "Non-technical"... (was Re: Grasping perspective...) (Loren
Petrich)
Re: Beer Tosser a Sabre Fan? Nope. (Charles Philip Chan)
Re: Awfulness of WordPerfect (Tim Adams)
Re: which OS is best? (Jim Richardson)
Re: which OS is best? (Jim Richardson)
Re: Government to break up Microsoft ("Otto")
Re: which OS is best? ([EMAIL PROTECTED])
Re: which OS is best? ([EMAIL PROTECTED])
Re: So what is wrong with X? (mlw)
----------------------------------------------------------------------------
From: R.E.Ballard ( Rex Ballard ) <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking
Subject: Re: linux as Netscape platform
Date: Sat, 29 Apr 2000 03:13:08 GMT
In article <[EMAIL PROTECTED]>,
Yeoh Yiu <[EMAIL PROTECTED]> wrote:
> I'm thinking of setting up a web kiosk.
> (user accounts/security is not an issue)
> Is Netscape on linux as stable as Netscape on NT ?
It depends on which version of Netscape. Navagator 4.01
and 4.7 aren't too bad, Navigator 6.0 was redesigned from
the ground up to use the features of KDE rather than trying
to suck in the entire Mosaic library as a static link (bringing
the entire memory image into RAM/Swap). A minor bug made
text entry a memory pig - I cloud grow Netscape from 16 meg
to 80 meg by interactively editing a dejanews post.
> My past experience with Linux was that the
> networking (tcp/ip, http) was great, but
> the user experience was poor because when switching between
> an ssh window, and Xterm and Netscape, when switching to
> Netscape, the keyboard would not be recognized
> so eg I could not use the search box of the web page
> I just downloaded.
Yes - that was a big problem, especially in 3.5x and 4.1 through
about 4.3, I think that was fixed in 4.5. Again, 6.0 was built
to exploit Linux toolkits.
To get really good performance out of Netscape for Linux, you
had to purchase the Motif Library and order/download the
shared-library version of Netscape. After a few minutes, the
initialization routines would swap out leaving a very lean run-time
module in memory. 6.0 provides this as well.
I have had troubles with Mozilla over firewalls. I would guess
that this has been corrected by now.
> squid.
>
--
Rex Ballard - Open Source Advocate, Internet
I/T Architect, MIS Director
http://www.open4success.com
Linux - 60 million satisfied users worldwide
and growing at over 1%/week!
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED] (Bob Hauck)
Subject: Re: The truth is often painful...
Reply-To: bobh{at}haucks{dot}org
Date: Sat, 29 Apr 2000 03:22:00 GMT
On Fri, 28 Apr 2000 07:15:09 -0400, Stuart Krivis <[EMAIL PROTECTED]> wrote:
>Just because a businessperson can't operate a computer well doesn't mean
>they are stupid.
Tell us, then, what we should call someone who uses tool X for several
hours each day, who depends X for results in their job, and yet refuses to
learn how to use X? Shall we call this person a "visionary genius"?
Fact is, most people that use a computer at work use it as a data-entry
device. Order takers of various kinds, hospital admissions, credit card
companies, etc, etc, hire thousands and thousands of people to basically
process forms. These people could give a shit what they are using as so
long as they are told what to click on and what to type where so that they
may take customer orders or fill out insurance forms or whatever they are
supposed to do to justify their paycheck at the end of the week.
The point being that for these sorts of applications you could swap out
Windows for BeOS or OS/2 or TOPS-20 and nobody would much care after a few
days as long as you had a half-hour meeting in the cafeteria to explain
how to use the new thing.
--
-| Bob Hauck
-| To Whom You Are Speaking
-| http://www.bobh.org/
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Crossposted-To: comp.os.linux.networking
Subject: Re: linux as Netscape platform
Reply-To: [EMAIL PROTECTED]
Date: Sat, 29 Apr 2000 03:28:33 GMT
Centuries ago, Nostradamus foresaw a time when R.E.Ballard ( Rex Ballard )
would say:
>Yes - that was a big problem, especially in 3.5x and 4.1 through
>about 4.3, I think that was fixed in 4.5. Again, 6.0 was built
>to exploit Linux toolkits.
>
>To get really good performance out of Netscape for Linux, you
>had to purchase the Motif Library and order/download the
>shared-library version of Netscape. After a few minutes, the
>initialization routines would swap out leaving a very lean run-time
>module in memory. 6.0 provides this as well.
Ah. You find that Netscape 6.0 works _substantially_ better when you
install Motif libraries?
_Right_, _sure_... We believe you...
Have you run ldd on any of the recent mozilla.org releases of Mozilla?
That might be an instructive exercise. It might show how you're making
this up as you go along, as usual.
--
Isn't it a bit unnerving that doctors call what they do "practice"?
[EMAIL PROTECTED] - - <http://www.hex.net/~cbbrowne/lsf.html>
------------------------------
From: [EMAIL PROTECTED] (Loren Petrich)
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: "Technical" vs. "Non-technical"... (was Re: Grasping perspective...)
Date: 29 Apr 2000 03:47:50 GMT
In article <8e0ch8$133e$[EMAIL PROTECTED]>,
Leslie Mikesell <[EMAIL PROTECTED]> wrote:
>No, that's not really the point of the difference. With a CLI, especially
>general purpose ones like unix shells, anything you do repeatedly
>can be automated by putting exactly the same commands you would
>use manually into a script, perhaps with variable substitution for
>the variations of each run. with a GUI, if automation is possible
>at all it has to be done in some macro language embedded into
>each program with its own peculiar syntax. There is almost
>never a way to do something by hand, then paste those commands
>into a script for re-use.
An interesting systemwide GUI scripting system is Apple's
AppleScript. It's one of those numerous snazzy things that Apple
introduced, but never followed up on. If one's app is properly written,
one can not only script it with AppleScript, one can even record an
AppleScript of one's actions and then edit one's transcript to one's
heart's content.
AppleScript will be making it into MacOS X, and it looks as if
Apple will be building AppleScript support into Cocoa, its OpenStep update.
I think it would be interesting for the KDE and Gnome guys to try
to create an AppleScript clone (KScript? GScript?). I'm not in any
position to take up this cause, but someone else may be interested.
>>>IMHO, Linux was born out of a sheer dislike of Microsoft, unlike its BSD
>>>cousins, who seemed to originate from a need for a BSD-like OS.
>Both Linux and the *bsd's are updated clones of unix systems that
>pre-date windows by at least a decade. Microsoft is pretty
>much irrelevant.
Linux was invented because Linus Torvalds wanted some alternative
to Minix, whose creator refused to try to improve it beyond being some
teaching OS. So when he issued his first kernel, a lot of frustrated
Minix users were happy to have something that they could work with.
--
Loren Petrich Happiness is a fast Macintosh
[EMAIL PROTECTED] And a fast train
My home page: http://www.petrich.com/home.html
------------------------------
Subject: Re: Beer Tosser a Sabre Fan? Nope.
From: Charles Philip Chan <[EMAIL PROTECTED]>
Date: 28 Apr 2000 23:32:43 +0500
>>>>> "Leslie" == Leslie Mikesell <[EMAIL PROTECTED]> writes:
> I might have agreed back in the days of hand-cranked computers.
> Now that a whole install takes 15 minutes or so, I'd rather make
> choices about config files before my system is broken instead of
> notices after the fact. As I understand it, the pre/post
> install scripts can't assume that stdin is connected to the
> installing terminal, so this is impossible.
Maybe Redhat should add an interactive mode to RPM that assume that
stdin is connected to the installing terminal. This will allow for
batch intallation --> Email notification and interactive installation
--> dialog boxes.
Charles
------------------------------
From: [EMAIL PROTECTED] (Tim Adams)
Crossposted-To: comp.sys.mac.advocacy,comp.os.os2.advocacy
Subject: Re: Awfulness of WordPerfect
Date: Sat, 29 Apr 2000 00:13:37 -0400
In article <8ec4uf$g8k$[EMAIL PROTECTED]>, "Christopher Smith"
<[EMAIL PROTECTED]> wrote:
> "Christopher Browne" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]...
> > Centuries ago, Nostradamus foresaw a time when Full Name would say:
> > >On Tue, 25 Apr 2000 04:59:37 -0400, Aaron Kulkis <[EMAIL PROTECTED]>
> > >wrote:
> > >>You mean... what if Word Perfect was let back into the market place?
> > >>
> > >>The mere mention of such an idea makes for nightmares in Redmond...
> > >
> > >The DOS version of WP was quite a good package. The Windows version
> > >was the worst mass produced software package I've ever seen.
> >
> > That demands the question:
> > Was it The Worst because:
> >
> > a) It was badly designed? or
> > b) It crashed incessantly?
> >
> > The design of version 5.2 certainly didn't seem _great_; the serious
> > problem I saw with it was that it could only run short periods before
> > crashing. Saving documents often made it crash, which was Rather
> > A Problem.
> >
>
> Plus it didn't follow any of the Windows UI standards. And it was
> horrifically slow.
>
> > Questionable design I could readily blame on WPC.
>
> But, it seems, you prefer to blame Microsoft based on things you heard
> around the campfire ?
>
> > But it was a well-known "factoid" at the time that Microsoft didn't
> > consider Windows done 'til it made WordPerfect for Windows crash.
>
> Ah, is this like the "till Lotus won't run" "factoid" as well ?
Chris; You should go and look up some of the old 'Dr. Dobbs Journals' from
the late 80. They had one (or more) articles showing code from Windows
that literally caused problems for several major players in the word
processing and spread sheet world. MS was also quite nice in tech support
telling people that to insure a program ran right on the OS, get Word or
Excel, because coming from the same company, you knew they would work.
> > That "fact" wasn't backed up by specific news out of Redmond, but was
> > certainly consistent with the other things that MSFT was doing at the
> > time, such as the "AARD" detection of DrDOS, and having Windows spit
> > out error messages if the DOS it was run on didn't come from Microsoft
> > (or IBM).
>
> Only in beta versions, then ?
--
Tim
------------------------------
From: [EMAIL PROTECTED] (Jim Richardson)
Crossposted-To:
comp.os.ms-windows.advocacy,comp.sys.mac.advocacy,comp.os.ms-windows.nt.advocacy,alt.flame.macintosh
Subject: Re: which OS is best?
Reply-To: [EMAIL PROTECTED]
Date: Sat, 29 Apr 2000 04:15:48 GMT
On Fri, 28 Apr 2000 21:31:08 -0500,
[EMAIL PROTECTED], in the persona of <[EMAIL PROTECTED]>,
brought forth the following words...:
>On Sat, 29 Apr 2000 01:11:21 GMT, [EMAIL PROTECTED] (Jim Richardson)
>wrote:
>
>>>>acrobat is installed, if I dbl click on the pdf, it is opened in acrobat,
>>>>but dragging to the printer is a no go, wants a file association.
>>>
>>>Reinstall the latest Acrobat from Adobe.
>>
>>Why? it prints okay as is, it just has to open it first. With a working
>>windows 9X setup, you don't tinker uneccesarily. THe registry is fragile
>>enough as it is.
>
>There is a problem with the installation if dragging a pdf to the
>printer doesn't start Acrobat and print the object. The easiest fix
>is usually to simply remove the old Acrobat and upgrade to the newest
>one - it's free, and on a reasonably fast computer should take all of
>2 minutes. Your registry concerns are unwarranted.
The question remains, why open the file in acrobat? just print it, I don't
need to preview it, I all ready did that.
As for my concerns re: the registry, too many people have bad stories of reg
corruption for me to be blase about dinking with a machine I don't own. (it is
at a company I am consulting for. )
>>>>Ghostscript is a postscript interpreter, very versatile, knows pdf and ps
>>>>and a lot more. GV is the X-frontend for GS and its not NT, 95
>>>
>>>You'd have to tell me why I'd need that, given that NT puts the PS
>>>interpreter in the driver.
>>
>>Read my lips, it's 95, not NT.
>
>Same difference WRT the PS printing.
so why is it that dragging a ps file to the printer produces a printout of the
ps file text? not as a postscript document?
>>
>>Nevertheless, Win95/98 are the vast majority of systems out there, in small
>>companies all over, because the price of NT is too high, (not that Win9X is
>>cheap mind, but it's usually included on the new systems, and the cost is
>>hidden.)
>
>You pay for what you want, and people have decided they want 98. It
>would be easy enough for businesses to simply buy and roll out
>machines with NT if they wanted; apparently many feel it isn't worth
>the extra money, as 98 fills their need appropriately.
When a small business buys 25 new desktop PC's the cost of the Licences for
NT are, well, hefty to say the least. At the place I am contracting, no-one
likes the windows boxes, sometimes for reasons that are not M$ related, but
there are very few good words to be said about M$'s products in use here.
Word is a four letter word :)
>>>>don't trust the boys in Redmond.
>>>
>>>So you've not run NT / Win2k?
>>
>>At a cost in excess of $300 for no net gain over linux? no, nor am I likely to.
>> Allthough the newest desktop machine actually has the oomph to do so (PIII at
>>550 Mhz and 128MB ram, nice machine, came with linux) I doubt I'll spend
>>money on W2K. (I probably could score a copy illegally, but frankly, I don't
>>support piracy, even of M$)
>
>How do you know there's no net gain if you haven't run it?
Simple, it doesn't do anything I can't allready do, that I need to do, it
costs over $300 and it will only run on one of the 5 machines on the boat,
even if I were allowed to install it on more. So why should I try it?
Does it have gvim/vim? does it have LyX? a compiler and various development
utilities? Does it firewall? or have a news server?
W2k may be the best that Redmond has ever put out, but it's not good enough for
my needs. YMMV, but that's mine.
--
Jim Richardson
Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
Linux, because life's too short for a buggy OS.
------------------------------
From: [EMAIL PROTECTED] (Jim Richardson)
Crossposted-To:
comp.os.ms-windows.advocacy,comp.sys.mac.advocacy,comp.os.ms-windows.nt.advocacy,alt.flame.macintosh
Subject: Re: which OS is best?
Reply-To: [EMAIL PROTECTED]
Date: Sat, 29 Apr 2000 04:17:30 GMT
On Fri, 28 Apr 2000 21:31:37 -0500,
[EMAIL PROTECTED], in the persona of <[EMAIL PROTECTED]>,
brought forth the following words...:
>On Sat, 29 Apr 2000 01:14:46 GMT, [EMAIL PROTECTED] (Jim Richardson)
>wrote:
>
>>>>You'd have to tell me why I'd need that, given that NT puts the PS
>>>>interpreter in the driver.
>>>
>>>I suspect that he wants to read/preview PS files on the screen and
>>>have the ability to print them on non-postscript printers. GS/GV
>>>will do that under windows as well, but it could have been a
>>>native capability.
>>>
>>> Les Mikesell
>>> [EMAIL PROTECTED]
>>
>>This is correct, I was going to install GS/GV for windows, but never got
>>around to it, it was simply easier to ps2pdf the file on the linux machine,
>>and print from the windows box via acrobat, annoying, far from "integrated"
>>and slow, but it worked.
>> It just boggles the mind that win9X doesn't know what to do with a postscript
>>file...
>
>Why would it? It's not as if I (or most users) routinely run across
>raw .ps files.
what do you mean by raw?
Why does linux have no problems with ps files, but M$ does?
--
Jim Richardson
Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
Linux, because life's too short for a buggy OS.
------------------------------
From: "Otto" <[EMAIL PROTECTED]>
Crossposted-To: comp.lang.java.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: Government to break up Microsoft
Date: Sat, 29 Apr 2000 04:19:45 GMT
"JEDIDIAH" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> >We could run in circles discussing this issue. MS was able to use those
> >tactics because everybody wanted to sell their products, under a
favorable
>
> IOW, they were the biggest player. Selling their stuff was most
> profitable. That's far different from being able to claim that
> NO ONE wanted to buy DRDOS.
And who's claiming that "NO ONE wanted to buy DRDOS", other than you that
is? There is a rather obvious difference between "sell" and "buy".
>
> Oddly enough, all the serious DOS users I knew at that time weren't
> using msdog but something a bit more powerful. Doubledos and
> Desqview seemed quite popular with the SysOp crowd.
Oh, the good ol' days...
>
> >term, which means incresed profit for the OEMs. They were/are competing
with
> >each other and did things on the voluntary basis. OEMs are as much at
fault
>
> ...this is a contradiction. They didn't do things on a voluntary
> basis. You've admitted this yourself. They were slaves to their
> own business goals. If a competitor could get an edge by acquiring
> the 'free OS' for cheap, then their contemporaries would likely
> have to follow suit in order to remain price competitive.
Welcome to the real business where the bottom line does count.
>
> This only establishes what was cheapest for an OEM to shovel at their
> customer, not what was better.
Not the cheapest, it was rather the most profitable for the OEMs. Better is
a relative terminology anyway.
>
> >in this as Microsoft. Even now some of the OEMs would not sell you a PC
with
> >OS other than Windows. Some of them started to sell Linux pre-installed
and
> >then there are Linux companies where you can't get a Windows PC.
>
> ...exclusive contracts, built on potentially punitive price
> discrimitation.
It isn't the first, or the last time for that matter, that it's done in the
US. That fact in itself doesn't make it right, however that's how most
businesses operate. Home Depot can sell products cheaper than your local
hardware store can buy them. How's that for "price discrimination"?
> >That was only in Beta code and was disabled in the released version.
Don't
> >you think that MS had right to do so? Why should they tested Win3.x
unders
> >DR-DOS?
>
> Nope. What they did was clear defamation.
>
> They could have left well enough alone if they really wanted to.
> However, they didn't do that. They specifically sought out
> competitor products.
Give me one good reason why Microsoft should've developed Windows3.x for the
DR_DOS. Their DOS platform was competing against it, was it not? Did they
have any obligation to Beta test Win3.x on the DR-DOS platform? I don't
think so. Why spend their R&D money on developing the Win3.x platform for
DR-DOS? Anyone deciding to install Win3.x final version for DR-DOS was free
to do so. If it didn't work on the competitor's product, then oh well... Let
the other company fix the problem. Just ask the Samba team how many times
they are forced to fix Samba, after a hot fix or service pack is out for the
Windows platforms.
Otto
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To:
comp.os.ms-windows.advocacy,comp.sys.mac.advocacy,comp.os.ms-windows.nt.advocacy,alt.flame.macintosh
Subject: Re: which OS is best?
Date: Fri, 28 Apr 2000 23:32:17 -0500
On 28 Apr 2000 21:58:05 -0500, [EMAIL PROTECTED] (Leslie Mikesell)
wrote:
>>>This is correct, I was going to install GS/GV for windows, but never got
>>>around to it, it was simply easier to ps2pdf the file on the linux machine,
>>>and print from the windows box via acrobat, annoying, far from "integrated"
>>>and slow, but it worked.
>>> It just boggles the mind that win9X doesn't know what to do with a postscript
>>>file...
>>
>>Why would it? It's not as if I (or most users) routinely run across
>>raw .ps files.
>
>Errr... Lots of people have raw .ps files. The reason you don't is
>that Microsoft would have had to share some of their wealth with
>another company and follow published standards to include the tools
>for you.
Sorry; I don't see the point of it. If I need to print something, I'm
perfectly willing to re-send it thru the driver, have it re-rasterize,
and then sent to the print device.
I'm not a print shop with big, long-time-to-render jobs lying around.
If I were, I could see the use. I'm not, and I don't. I print mostly
text files, and I have no need to keep .ps files lying around.
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To:
comp.os.ms-windows.advocacy,comp.sys.mac.advocacy,comp.os.ms-windows.nt.advocacy,alt.flame.macintosh
Subject: Re: which OS is best?
Date: Fri, 28 Apr 2000 23:39:45 -0500
On Sat, 29 Apr 2000 04:15:48 GMT, [EMAIL PROTECTED] (Jim Richardson)
wrote:
>>There is a problem with the installation if dragging a pdf to the
>>printer doesn't start Acrobat and print the object. The easiest fix
>>is usually to simply remove the old Acrobat and upgrade to the newest
>>one - it's free, and on a reasonably fast computer should take all of
>>2 minutes. Your registry concerns are unwarranted.
>
>
>The question remains, why open the file in acrobat? just print it, I don't
>need to preview it, I all ready did that.
Windows itself cannot process the PDF file; AAcrobat is required for
that.
> As for my concerns re: the registry, too many people have bad stories of reg
>corruption for me to be blase about dinking with a machine I don't own. (it is
>at a company I am consulting for. )
Obviously not Windows consulting, I take it.
>>>>>Ghostscript is a postscript interpreter, very versatile, knows pdf and ps
>>>>>and a lot more. GV is the X-frontend for GS and its not NT, 95
>>>>
>>>>You'd have to tell me why I'd need that, given that NT puts the PS
>>>>interpreter in the driver.
>>>
>>>Read my lips, it's 95, not NT.
>>
>>Same difference WRT the PS printing.
>
>so why is it that dragging a ps file to the printer produces a printout of the
>ps file text? not as a postscript document?
Because the file is still a text file in the eyes of Windows, and
hence is seen as a text file (read: sent as text) by the owning
application (which is probably notepad or similar).
Perhaps someone more familiar than I with printing can tell you how to
send the data to the printer raw, without going through the Windows
print processing device, which sounds like what you
want....essentially, cp filename.ps lpr -> copy filename.ps LPT1:.
>>>Nevertheless, Win95/98 are the vast majority of systems out there, in small
>>>companies all over, because the price of NT is too high, (not that Win9X is
>>>cheap mind, but it's usually included on the new systems, and the cost is
>>>hidden.)
>>
>>You pay for what you want, and people have decided they want 98. It
>>would be easy enough for businesses to simply buy and roll out
>>machines with NT if they wanted; apparently many feel it isn't worth
>>the extra money, as 98 fills their need appropriately.
>
>When a small business buys 25 new desktop PC's the cost of the Licences for
>NT are, well, hefty to say the least. At the place I am contracting, no-one
>likes the windows boxes, sometimes for reasons that are not M$ related, but
>there are very few good words to be said about M$'s products in use here.
>Word is a four letter word :)
So why do they use it if it's as bad as you claim? Obviously they've
found nothing better - or else they just like to complain. I suspect
both are somewhat true.
>>>>>don't trust the boys in Redmond.
>>>>
>>>>So you've not run NT / Win2k?
>>>
>>>At a cost in excess of $300 for no net gain over linux? no, nor am I likely to.
>>> Allthough the newest desktop machine actually has the oomph to do so (PIII at
>>>550 Mhz and 128MB ram, nice machine, came with linux) I doubt I'll spend
>>>money on W2K. (I probably could score a copy illegally, but frankly, I don't
>>>support piracy, even of M$)
>>
>>How do you know there's no net gain if you haven't run it?
>
>Simple, it doesn't do anything I can't allready do, that I need to do, it
>costs over $300 and it will only run on one of the 5 machines on the boat,
>even if I were allowed to install it on more. So why should I try it?
To learn something?
> Does it have gvim/vim?
Not in the box, but I'm sure you could download similar tools.
>does it have LyX? a compiler and various development
>utilities?
Not in the box, but I'm sure ... "
>Does it firewall? or have a news server?
Not in the ... "
>W2k may be the best that Redmond has ever put out, but it's not good enough for
>my needs. YMMV, but that's mine.
------------------------------
From: mlw <[EMAIL PROTECTED]>
Subject: Re: So what is wrong with X?
Date: Sat, 29 Apr 2000 00:54:44 -0400
Stephen Cornell wrote:
>
> mlw <[EMAIL PROTECTED]> writes:
>
> > About once or twice a year, we get a crop of people trashing X for
> > various reasons. The reasons, however, never tell the whole story.
> >
> > "X is slow!"
> > Perhaps XFree86 is not as speedy as it could be, but it is really hard
> > to characterize as slow. Some optimized X servers are VERY speedy.
> >
> > "X is old!"
> > Yup, it is. Very proven and useful technology. Like most of the best OS
> > ideas, computers are growing in to them. X is no different. 20 years ago
> > virtual memory was totally impossible on a PC, now it is a necessity. X
> > is no different.
> >
> > I find it very funny that people are starting to kick X. Citrix and
> > Windows Terminal server are trying to do what X has done all along. I
> > guess if you can't beat the competitor with quality and innovation, you
> > have to discredit them with FUD and lies.
> >
> > Innovation to most people is doing something new to the world. Innvation
> > at Microsoft is doing something new to Microsoft.
>
> I may be speaking somewhat out of turn here, as I have no experience
> of programing X-applications, and I have only anecdotal understanding
> how the the X protocol works. What I say below is based on my
> experience of running X-applications remotely, and from what I know
> about where the various bits of software live when I do so. Please
> feel free to correct me where I am wrong - I'd like to be more
> educated on this topic.
>
> It appears to me that X *does* have some major deficiencies as a
> protocol for running remote applications. Isn't the X-server's job
> merely to render raw graphics on the screen, and to pass mouse and
> keyboard events to the client? Almost all of the the work is done on
> the remote machine, and not enough on the local machine (the one
> running the X-server). Thus, the local machine is not in a position
> to filter out the X-events that remote application does not need to
> know about, and there is a huge waste of bandwidth. For instance, a
> remote emacs window is almost unusably slow over a phone line. I can
> see the modem lights flashing when I merely move the mouse.
>
> Given that most GUI applications have pretty simple needs (buttons,
> sliders, text boxes, etc.), it's surely possible to pass the
> information over the network in a more parsemonious way (`slider 3 has
> been moved from position 1.1 to 2.3' rather than `mouse is being
> clicked and dragged over points x1,y1; x2,y2; x3,y3...'). The way
> that widgets respond to X-events could be controlled localy, without
> having to wait for calculations and updates from the remote client
> between each X-event.
>
> What I am suggesting is that the `X-client/X-server' division of
> labour is too unequal, and that a `remote server/local not-too-thin
> client' approach leads to much less wasted bandwidth. I know this is
> nothing new, and I also know that this approach can be (and is) used
> with applications that run under X. However, as far as I know there
> is no standard protocol (API?) for creating GUIs in this way, so
> programmers are not encouraged to write most software according to
> this model.
>
> I don't know how Windows terminal services works, nor have I ever used
> it, but I would have thought that, if it runs as a thin client so that
> only a subset of the terminal activity is sent over the network, it is
> much more efficient than running X-clients remotely.
>
> In summary: X is very powerful, but its very flexibility means that it
> is too inefficient to be used as the communication protocol for most
> remotely-run applications. Of course, this has nothing to do with X's
> suitability as the display protocol for locally running applications.
>
> Anyone care to comment on the above, or correct my ignorance?
Actually, I found your message very well thought out, however, a remote
Windowing environment such as X (terminal server too!) either have to
support regular applications or "special" applications coded for
remoteness. Perhaps it is possible to make "remote aware" optimized
programs, but I digress.
Neither Windows programs nor X programs are coded in such a way that the
applications expect to see anything but WM_xxx messages delivered to the
application. That is just the nature of the beast. No UI programmer
would accept any less.
You talk about the client end picking up more of the processing. Perhaps
sending and receiving meta information about what the user wants, having
the server provide the information whilst the client handles the whole
presentation end. This sort of remote server you are talking about is
very well represented by a web browser.
--
Mohawk Software
Windows 9x, Windows NT, UNIX, Linux. Applications, drivers, support.
Visit http://www.mohawksoft.com
"We've got a blind date with destiny, and it looks like she ordered the
lobster"
------------------------------
** 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
******************************
