Linux-Advocacy Digest #843, Volume #25 Mon, 27 Mar 00 22:13:07 EST
Contents:
Re: Packaging Tools ("John W. Stevens")
Re: DCOM vs CORBA Re: Weak points (R.E.Ballard ( Rex Ballard ))
Re: Windows 2000 has 63,000 bugs - Win2k.html [0/1] - Win2k.html [0/1] (Tim Hockin)
Re: Rumors ... (Andrew)
Re: Penquins Forever! Was (Re: A pox on the penguin?) ("ax")
Re: Enemies of Linux are MS Lovers (Roger)
----------------------------------------------------------------------------
From: "John W. Stevens" <[EMAIL PROTECTED]>
Crossposted-To: comp.unix.bsd
Subject: Re: Packaging Tools
Date: Mon, 27 Mar 2000 17:58:11 -0700
Christopher Browne wrote:
>
> No, it's *NOT* intuitive. Not reasonably *at all.*
>
> If you sit someone down at a computer, and suggest that they type in
> whatever comes to mind, it is not likely that they'll come up with the
> GNU build commands.
Oh, well, if that's your argument, then you aren't really saying
anything. "Intuitive" is a relatively meaningless noise . . . you need
to define the term.
Now, if you want to actually communicate, the phrase is "leveraging off
of existing knowledge".
> The GNU build commands may be *readily learnable* by those that are
> literate with UNIX. Unix is *not* particularly intuitive.
No OS is particulary intuitive. And for those that do do a good job of
"leveraging off of existing knowlege", you end up with a very short,
very shallow learning curve that levels off at a much lower point than
that of Unix.
> It has
> built up over time, having complexity that arises from its history.
Unix is not particulary complex.
> It has got a steep learning curve which tends to discourage many;
Actually, the learning curve is quite shallow. Those that get
discouraged do so 'cause they weren't given the proper expectations.
> the
> "upside" is that being steep, those that *do* climb the hill can
> quickly get at powerful functionality.
Again, the curve is relatively shallow. Consider the transition from:
1) Typing in individual commands.
To
2) Scripting.
on a Unix box.
It's trivial on Unix, and the curve is shallow. On systems that
exclusively GUI'ed, or that are primarily GUI'ed, you really *DO* have a
steep learning curve!
> That is quite a different thing from being "intuitive."
There is very, very little in the computer world that is actually
"intuitive".
--
If I spoke for HP --- there probably wouldn't BE an HP!
John Stevens
[EMAIL PROTECTED]
------------------------------
From: R.E.Ballard ( Rex Ballard ) <[EMAIL PROTECTED]>
Subject: Re: DCOM vs CORBA Re: Weak points
Date: Tue, 28 Mar 2000 01:15:36 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (CAguy) wrote:
> On Fri, 24 Mar 2000 14:36:50 GMT, R.E.Ballard ( Rex Ballard )
> <[EMAIL PROTECTED]> wrote:
>
> >> DCOM was never ment to be used between components 'in process'.
> >
> >Correct. DCOM provided the ability to combine smaller modules
> >that could be executed either on the same machine or distributed
> >across multimple machines. Unfortunately, the performance
> >consequences of using DCOM instead of COM was so expensive that
> >most programmers simply stuck with COM.
> >
>
> I think we're mixing apples and oranges here,
There is some clarification required.
> If you're talking about cross network/process
> communications between components..you're
> talking about DCOM.
Correct. DCOM can be used for interprocess communication, this
can be two processes on the same machine, or two processes on
separate machines.
> COM is a spec for in process components only.
Correct, COM is the application programmer interface for connecting
to COM objects stored in a DLL. Especially for NT, these must be
Intraprocess. The application must link the calls to the COM control
into the application. The result is an application which cannot be
reconfigured without recompiling the code.
> So to say people stick with COM instead of DCOM implies that they
> were using DCOM for inprocess stuff
I was unclear about this. The benchmarks were based on intraprocessor
communications using both DCOM (interprocess) and COM (intraprocess).
> ...yeah, THAT would be slow
> and nobody in there right mind would do that.
Unless you wanted to be able to dynamically reconfigure your system
or be less vulnerable to DLL upgrades.
> >> It's just an extention to COM for distributed components.
>
> Again, we're mixing applies and oranges. CORBA is cross network
> while XPCOM is an in process component model, just like COM.
CORBA can be used for interprocess communications between processes
within the same processor (in which case the ORB simply passes the
messages using RAM buffers) or between processes across a network
(in which case the ORB serializes the messages and sends them via
network - usually TCP/IP using Internet Inter-Orb Protocol (IIOP).
Applications simply call the ORB. The ORB decides the best connection
between the Client and the Server.
> But you're right, KDE and GNOME are using CORBA for their
> cross network communications. However, they are using a
> component model similar to COM/XPCOM for in process stuff.
With KDE, which is based on the QT library, the QT APIs hide the
COM/CORBA interface and can call either. On NT the QT library may
indeed call COM or DCOM. On UNIX, the QT library calls CORBA.
> I'm not sure if Mozilla has a cross-network model yet, they
> could use CORBA, or something similar to DCOM...but I think
> their going with XML.
XML is being used as an alternative to HTML. Since XML is extensible,
and HTML can be expressed as an XML DTD, it's pretty easy to implement
HTML on an XML system.
> Well, CORBA and DCOM are always going to be slower then COM,
> obviously.
Correct. COM overhead is about 1 microsecond. Interaprocessor CORBA
is about 4 microseconds). Interprocessor CORBA is about 20
microseconds (not including transmission time), and DCOM is about 2
milliseconds.
> Calling a COM method is equivalent to a virtual function
> call in C++
Which implies a hard link-time binding which cannot generally be
changed. A change in the DLL often (though not always) means
that the orginal application must be recompiled or relinked.
> ...which is slightly slower then a standard C function
> call.
Not significantly slower. Generally the flexibility of being
able to replace the underlying DLL and the ability to share reentrant
code makes the cost of the COM call worth the effort. In Linux,
this is the equivalent to calling a routine within an ELF .so library.
>
--
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: Tim Hockin <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.redhat
Subject: Re: Windows 2000 has 63,000 bugs - Win2k.html [0/1] - Win2k.html [0/1]
Date: 28 Mar 2000 01:35:44 GMT
In comp.os.linux.misc jdaspinw <[EMAIL PROTECTED]> wrote:
: The simple truth is that Microsoft makes a damned fine
: operating system, and if you can't run Windows without
bwahahahahah! you have no idea about OS, do you?
: And Linux is no more stable, secure, or easy to manipulate
: than Windows. All you're doing is regurgitating bull crap
: you picked up somewhere in a magazine, or from one of these
um, it really is. truly.
: discussion groups. Do any of you have what it takes to
: write an operating system? Have you contributed to the Open
doing my own right now, so I'd guess that is a yes.
: Source kernel? Are you a guru, or just a slob playing at
yes I have contributed, and I'd say I'm closer to guru than slob, though
I'd never call myself a guru.
: They all make mistakes. You can't write operating systems,
: and you never will. hell, half of you probably couldn't
umm, troll?
: program your way out of a round room with one door. I don't
troll
: It does you no good, and just makes you all look stupid.
and this post does what for your appearance...
--
Tim Hockin
[EMAIL PROTECTED]
[EMAIL PROTECTED]
This program has been brought to you by the language C and the number F.
------------------------------
From: Andrew <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Rumors ...
Date: Mon, 27 Mar 2000 20:57:25 -0500
Craig Kelley wrote:
>
> All we need from Microsoft and it's partners is a promise not to
> develop Windows-centric hardware; publish all calls for every
> piece of equipment. The source code to Windows would do nicely,
> unless Microsoft would care to cooperate in other, less barbaric
> ways....
I don't think this goal has ANYTHING to do with Windows itself. This goal would
be much better served by making the drivers, not the OS open source. Oddly
enough, as Linux gets more and more popular, hardware companies are releasing
drivers for Linux... which are being released closed-source. Ironically this
defeats the very thing you want.
Andrew
------------------------------
From: "ax" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Penquins Forever! Was (Re: A pox on the penguin?)
Date: Tue, 28 Mar 2000 02:10:18 GMT
"JEDIDIAH" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> On Sun, 26 Mar 2000 14:42:26 GMT, ax <[EMAIL PROTECTED]> wrote:
> >
> ><[EMAIL PROTECTED]> wrote in message
> >news:[EMAIL PROTECTED]...
> >> In article <ereD4.411$[EMAIL PROTECTED]>,
> >> "ax" <[EMAIL PROTECTED]> wrote:
> >> >
> >> > But in Linux, all penguins are lazy sitting with round belly.
> >> > They cannot walk or fly. They cannot even stand up
> >> > with fat belly. Linux penguins must have been eating
> >> > too much "free" stuff.
> >> >
> >> > Totally amazing to see sitting penguins everywhere!
> >> >
> >> > Can Linxu penguins fly someday?
> >> >
> >> > It must be scary if those sitting penguins all suddenly
> >> > stand up and walk. They'd better be sitting!
> >>
> >> Nothing could be further from the truth. Linux is moving so fast,
> >> in so many different areas (desktop, servers, embedded devices),
> >> that it is taking the s/w indusrty by storm. Those not myopic
> >> from the M$ affect on this industry are adopting Linux in their
> >> droves. In the sea of s/w the Linux penguin is supreme. :-)
> >
> >Linux penguins are cloned UFO with programmed mission
>
> Most of the software industry exists quite outside your
> narrow, consumer view of it actually...
>
Narrow or not, "consumer view" may not matter at all to Linux
Penguins. Linux Penguins are all blind deaf-mutes.
> >of destroying the "Software Industry" on our planet. For the
> >protection of our mankind, should we re-program them or
> >take them to the zoo?
> >
> >
> >
>
>
> --
>
> It is not the advocates of free love and software
> that theare the communists, but rather those that |||
> advocate or perpetuate the necessity of only using / | \
> one option among many, like in some regime where
> product choice is a thing only seen in museums.
>
> Need sane PPP docs? Try penguin.lvcm.com.
------------------------------
From: Roger <roger@.>
Crossposted-To: alt.microsoft.sucks,alt.destroy.microsoft
Subject: Re: Enemies of Linux are MS Lovers
Date: Tue, 28 Mar 2000 02:10:25 GMT
On Sat, 25 Mar 2000 14:08:13 -0500, someone claiming to be T. Max
Devlin wrote:
>Quoting doc rogers from alt.destroy.microsoft; Wed, 22 Mar 2000 08:01:15 -0500
>>T. Max Devlin <[EMAIL PROTECTED]> wrote in message
>>news:[EMAIL PROTECTED]...
>>> Quoting doc rogers from alt.destroy.microsoft; Sat, 18 Mar 2000
>>11:32:25 -0500
>Why would he start such a thread on these groups? I believe (and may be
>mistaken, but despite Roger's empty but resolute insistence, it only happens a
>few times a day, not every time I open my mouth)
Please do post a cite where I have said any such thing. All I have
ever said is that you often wrong, and when I do, I can point to
specific instances of your being so.
>Am I going over the edge myself in being able to conceive of a way that a
>rather horribly implemented hardware platform (the Gateway laptop under
>discussion) is *Microsoft's* fault?
>
>Maybe, but the only evidence you have to refute it is correlative, since there
>are no other mainstream pre-load OSes which we can examine to determine if a
>less restrictive and more technically capable software/licensing combination
>on the manufacturer would mitigate the situation.
Of course, since we are talking about an enduser doing the
installation, the fact that at one point the OS was preloaded makes no
difference...
>>You can certainly have a hard time installing Windows in certain situations,
>>but what OS isn't that true of?
>There are no other OSes available for the hardware under discussion, so I
>really can't say.
Because if you tried to install Linux or BeOS on this box, MS would
break down your door and beat you to within an inch of your life. At
least in MaxWorld, where one can purchase a model 2600 from Gateway,
despite the fact that it is not a model they have ever offered...
>Snce Windows is supposedly the one that makes it easy to
>install an OS, Microsoft will easily and often take credit for making PCs
>accessible to the masses, and Bill Gates has been said to have invented "ease
>of use", I would say that the situation with other OSes is irrelevant for the
>most part.
Of course, that it was Max that said BG invented ease of use to make a
point he could not have done using reality makes his opinion of the
relevancy of other OSes suspect.
------------------------------
** 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
******************************
