Linux-Advocacy Digest #932, Volume #25 Tue, 4 Apr 00 01:13:06 EDT
Contents:
Re: Windows 2000: nothing worse (Jim Richardson)
Re: So where are the MS supporters. ([EMAIL PROTECTED])
Re: Ding Dong! ([EMAIL PROTECTED])
Re: it is OUT, MS is GUILTY ! ([EMAIL PROTECTED])
Re: So where are the MS supporters. ([EMAIL PROTECTED])
Re: Linux mail/news application questions (Janet)
Re: Why Linux on the desktop? (Jim Richardson)
Re: it is OUT, MS is GUILTY ! ("Bobby D. Bryant")
Re: linux users group? ("BadMan")
Re: Why Linux on the desktop? (Christopher Browne)
Re: OT:RANT:Long: If anyone develops an IDE for Linux PLEASE NO PROJECT FILES (or
MDI for that matter) (The Ghost In The Machine)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (Jim Richardson)
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Windows 2000: nothing worse
Reply-To: [EMAIL PROTECTED]
Date: Tue, 04 Apr 2000 03:24:54 GMT
On Mon, 3 Apr 2000 02:13:28 -0500,
Erik Funkenbusch, in the persona of <[EMAIL PROTECTED]>,
brought forth the following words...:
>George Marengo <[EMAIL PROTECTED]> wrote in message
>news:[EMAIL PROTECTED]...
>> Eric, follow the steps, one by one, and you'll be able to find the
>> story that shows your claim about IBM being unable to keep up
>> with sales to be fraudulent:
>
>I assume you have evidence of my intent when posting what I wrote. If not,
>using the term fraudulent is not only improper, but constitutes libel and
>slander. Fortunately for you, nobody would take any such statemtn by you
>seriously or you could have a serious lawsuit on your hands.
>
>My statements were made in good faith. At the time I purchased my z50 in
>january, dealers could not keep them in stock and I had not heard of the
>discontinuation of them.
>
Who told you that IBM couldn't keep up with production? the salesman who
was flogging something to you that IBM was either about to dicontinue, or had
allready done so? If that's your source, all I can say is, take a friend with
you when you buy a car...
--
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]
Subject: Re: So where are the MS supporters.
Date: Tue, 04 Apr 2000 03:16:25 GMT
In article <[EMAIL PROTECTED]>,
mlw <[EMAIL PROTECTED]> wrote:
>
> I just check on at, 10:25 PM EST to see what the MS supporters had to
> say about the court ruling, and low and behold, nothing.
>
> Hmmm, I am left to wonder.
They're busy dumping their stock on e-trade.
Seriously though, MSFT will be a good buy in the next few weeks for long
term investors. Really doesn't matter what the conclusions of law are
in the case. MSFT will still be a sound investment.
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: Ding Dong!
Date: Tue, 04 Apr 2000 03:26:07 GMT
In article
<[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
> On Tue, 04 Apr 2000 01:14:31 GMT, Don Clayton <[EMAIL PROTECTED]> wrote:
> >The witch is dead!
> >
> >Burn in hell Microsoft!
>
> Split MS up!
>
> part 1) the OS
> part 2) applications
> part 3) all the other crap like msnbc.
>
> It's worked w/ the phone company for years, it'll work just fine w/
MS.
Actually, my job was easier before the Bell-Co divestiture. I used to
just have to resolve comm problems, now I'm a referee in a tag-team
finger pointing contest where the participants change their name every
few weeks. Used to be I could just blame it all on AT&T. $DEITY, those
were the good old days.
So they split MSFT up, and some rogue MSFT application sh*ts all over my
system. Jolly good. I can hardly wait for this one....
Sent via Deja.com http://www.deja.com/
Before you buy.
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: it is OUT, MS is GUILTY !
Date: 3 Apr 2000 19:48:52 -0700
Looking at all the latest news today and inteviews with legal experts,
they all agree that this case will be concluded much sooner than others
think. May be by summer of 2002 at the most, depending on the direction
this case takes.
One direction will be to go all the way to the suprem'court. Another is
for the gov+states to negotiate a settlement.
Either way, freedom for the consumer from the tyron MS rule is just
around the corner, the consumer have suffered from lack of choice
and from lack of innovation, that one to two more years to wait is not
that long.
It is amazing how many idiots can come on the TV and talk about this, most
think that there is nothing out there but MS software. Some even claimed
that no matter what happen, people will still have to buy MS
operating systems becuase there is nothing else to run the PC on.
This is from so called computer experts being interviewed on national TV.
Guess which company stocks they own.
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: So where are the MS supporters.
Date: 3 Apr 2000 19:54:41 -0700
In article <[EMAIL PROTECTED]>, mlw says...
>
>
>I just check on at, 10:25 PM EST to see what the MS supporters had to
>say about the court ruling, and low and behold, nothing.
>
>Hmmm, I am left to wonder.
>
They are home licking their wounds and counting how much money they
lost in MS stocks today.
Not only MS is now officially guilty, any company now that does business
with this guilty company will have some cloud on its head. No one will want
to do business with a criminal company. No one should own stocks of a
company as slizzy as MS is. No engineer with any ethics should now continue
to work for such a criminal company.
I hope they throw the book at them.
------------------------------
From: Janet <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux,comp.os.linux.development.apps
Subject: Re: Linux mail/news application questions
Date: 03 Apr 2000 21:12:47 -0700
[EMAIL PROTECTED] writes:
> Just installed Linux Redhat 6.2 after a few years away from the OS.
> I'm stunned at how much is changed, but i'm beginning to miss the
> things which caused me to return to windows in the first place.
> I'd love a mail program that can sort and search mail. One that can
> automatically place mail in folders based on simple rules. A contact
> list that integrates with the mail program so I only have to maintain
> one list of contacts/email addresses. Netscape mail really sucks. It
> has corrupted my archives several times.
Never tried Netscape mail. At any rate, procmail can do everything you
want for the filtering, sorting, etc. Pretty much any mail program can
search mail; I use mutt (color, too!). Granted, these are text only
tools. However, if you just look on http://www.freshmeat.net, there are a
ton of other mail clients that aren't. You can get KDE, GNOME, simple X
ones, etc. etc.
> The news reading programs are very weak. I simply want the ability to
> select what articles I want to download, tell it to download, and have
> it happen. I'd also like the ability to have it automaticly combine
> and decode messages. Several windows programs, (outlook, agent, etc)
> have these abilities. I'm surprised Linux still doesn't.
> Are there any modern applications in development that meet these
> needs? Everytime I tried to search for an answer to this question, I
> found a lot of advice saying to use mail, trn, etc. Yeah, I used those
> programs for a while; but I didn't upgrade to linux to use the same
> text based programs I used 10 years ago.
Same with this, look on freshmeat.
Janet
------------------------------
From: [EMAIL PROTECTED] (Jim Richardson)
Subject: Re: Why Linux on the desktop?
Reply-To: [EMAIL PROTECTED]
Date: Tue, 04 Apr 2000 04:31:34 GMT
On Tue, 04 Apr 2000 02:49:57 GMT,
Jim Richardson, in the persona of <[EMAIL PROTECTED]>,
brought forth the following words...:
>On 3 Apr 2000 22:16:09 GMT,
> abraxas, in the persona of <[EMAIL PROTECTED]>,
> brought forth the following words...:
>
>>Donn Miller <[EMAIL PROTECTED]> wrote:
>>> "John W. Stevens" wrote:
>>>> [Mathias Warkus wrote:]
>>
>>>> > Writing HTML is not programming.
>>>>
>>>> Yes it is.
>>
>>> I agree. I would consider HTML and {TeX,LaTeX} programming
>>> languages. In html, you are programming a web browser to display
>>> something. The browser acts as an interpreter for HTML. And, of
>>> course, HTML is "Hyper Text Markup Language". Writing a TeX app is
>>> also programming, in my view. You are basically programming a
>>> typesetter when writing a [La]TeX document.
>>
>>You would be wrong. Theyre scripting languages, as are TCL, Perl,
>>Python, etc. Programming languages would include:
>>
>>C
>>C++
>>Fortran
>>Pascal
>>
>>If you do not understand the difference between scripting and
>>programming, you are a scripter, not a programmer. :)
>>
>>
>>
>>
>>-----yttrx
>>
>>
>
>Actually, they are interpreted vs compiled languages. Or do you believe that
>since there are at least 2 C interpreters, (EIC and another one the name
>of which escapes me) and compilers for python (the byte code compiler built
>into the interpreter for one.) and for tcl.
>
dang, I am tired, didn't finish the post, yawn anyway, since there are
compilers for some interpreted languages, and interpreters for some compiled
languages, do you believe (eg) Python is not a programming language? if
so, could you describe what feature python lacks, thus striking it from
the roll-call of "programming languages"?
--
Jim Richardson
Anarchist, pagan and proud of it
WWW.eskimo.com/~warlock
Linux, because life's too short for a buggy OS.
------------------------------
From: "Bobby D. Bryant" <[EMAIL PROTECTED]>
Subject: Re: it is OUT, MS is GUILTY !
Date: Mon, 03 Apr 2000 22:33:25 -0500
[EMAIL PROTECTED] wrote:
> It is amazing how many idiots can come on the TV and talk about this,
Yeah, I saw a guy on some news show telling everyone that a structural remedy
was going to cost consumers $10,000,000,000.00, and none of the moderators or
other pundits had the brass to ask him to justify his numbers.
So now lots of listeners will think that claim is a "finding of fact", too.
Bobby Bryant
Austin, Texas
------------------------------
From: "BadMan" <[EMAIL PROTECTED]>
Subject: Re: linux users group?
Date: Tue, 4 Apr 2000 00:55:26 -0400
How about in NYC ?
Paul Gowder <[EMAIL PROTECTED]> wrote in message
news:8c3ojh$p7i$[EMAIL PROTECTED]...
> Hiya -- I'm wondering if anyone knows of a linux users group or some
similar
> organization in the boston area.
>
> thanks,
> -Paul
>
> The address "[EMAIL PROTECTED]" goes directly to the trash.
> (anti-spam mechanism) replace "null" with "paul" to reach me.
>
> "Oh, it's just a mob war: go back to sleep honey."
> -Marge Simpson (on my favorite episode)
------------------------------
From: [EMAIL PROTECTED] (Christopher Browne)
Subject: Re: Why Linux on the desktop?
Reply-To: [EMAIL PROTECTED]
Date: Tue, 04 Apr 2000 04:55:00 GMT
Centuries ago, Nostradamus foresaw a time when Jim Richardson would say:
>>Actually, they are interpreted vs compiled languages. Or do you believe that
>>since there are at least 2 C interpreters, (EIC and another one the name
>>of which escapes me) and compilers for python (the byte code compiler built
>>into the interpreter for one.) and for tcl.
>
>dang, I am tired, didn't finish the post, yawn anyway, since there are
>compilers for some interpreted languages, and interpreters for some compiled
>languages, do you believe (eg) Python is not a programming language? if
>so, could you describe what feature python lacks, thus striking it from
>the roll-call of "programming languages"?
Right, and Lisp is an interpreted language and therefore slow...
... And on Crusoe, 80386 code gets *INTERPRETED,* which means that *any*
software that gets deployed obviously isn't written in a "programming
language."
--
Rules of the Evil Overlord #48. "If my trusted lieutenant tells me my
Legions of Terror are losing a battle, I will believe him. After all,
he's my trusted lieutenant."
<http://www.eviloverlord.com/lists/overlord.html>
[EMAIL PROTECTED] - - <http://www.ntlug.org/~cbbrowne/lsf.html>
------------------------------
From: [EMAIL PROTECTED] (The Ghost In The Machine)
Subject: Re: OT:RANT:Long: If anyone develops an IDE for Linux PLEASE NO PROJECT FILES
(or MDI for that matter)
Date: Tue, 04 Apr 2000 05:00:51 GMT
In comp.os.linux.advocacy, Andy Newman <[EMAIL PROTECTED]>
wrote on Mon, 3 Apr 2000 16:17:11 +1000 <[EMAIL PROTECTED]>:
>The Machine wrote:
>>In comp.os.linux.advocacy, Andy Newman <[EMAIL PROTECTED]>
>> wrote on Sun, 2 Apr 2000 16:29:56 +1000 <[EMAIL PROTECTED]>:
>>>I wrote:
>>>> @for i in $(SUBDIRS); do\
>>>> cd $$i && $(MAKE);\
>>>> done
>>>
>>>The cd $$i && $(MAKE) should be in a sub-shell of course.
>>
>>I would have thought it already is...? :-)
>
>No, the for...done is running in a shell started by make but
>the actual "cd $$i && $(MAKE)" needs to be in its own sub-shell
>so the change of directory doesn't affect the rest of the loop
>and break the change to the other directories. Adding a "cd .."
>doesn't cut it as the sub-dirs are not necessarily one level down.
>The correct method is to "( cd $$i && $(MAKE) )" which uses another
>shell level to run the make and the cd doesn't affect the parent.
>Clear :) (You could go through hoops to rememeber the cwd and go back
>there after the $(MAKE) but the sub-shell is simpler).
Ah....missed that the first time, yes. Nothing like a symbolic
link to screw up a whole slew of assumptions regarding 'cd ..' :-).
>
>>Mind you, I did run into something weird in a development
>>environment in NT, at one point. It turns out that
>>GNU Make on NT uses CMD.EXE to fork off its commands.
>
>(watch out, long rant follows :)
>
>Yes. And on Win95 COMMAND.COM is often used. It depends on how make
>is coded but if they use the C library system() function then
>the system's command interpreter is used (as it should according to
>the C standard). So that's command.com or cmd.exe. On Win9x
>command.com *always* returns 0 as an exit status (!!!!) which
>makes it useless as far as make is concerned. The MS C-library
>spawn() (or one the variants) is required if you want exit codes
>but then you don't get the command.com builtins (del, copy, et al)
>and have to look for them and treat them specially. Sigh. And of
>course Win9x and NT have different builtins and/or arguments. A few
>years ago I wrote a make front-end for a project that needed to
>cope with Win9x/Win-NT/Unix and discovered all of this crap. Coping
>with the Windows mess took the vast majority of the time. But I
>should've expected it. Years ago I suffered using MS-DOS 3.3 as
>a development environment and swore never to go back. Every time
>you want to do something there's this bloody wall you have to climb
>over only to discover the shards of broken glass waiting for you
>on the other side. It just hurts too much. If you're doing app.
>development a lot of this doesn't affect you but if you actually
>want to compete and do things that push the boundaries a bit you
>find it requires far more effort than it should because of the
>broken design. At least with Unix systems you can generally find
>out what is causing the problem and generally work around it easily.
>With Windows a lot of the cause is hidden and the solution far more
>time consuming. But the pay off has been quite good so people
>persist. I've worked in R&D for a major computer manufacturer for
>the last 10 or so years and watch it happen all the time.
>
>The make/command interpreter stuff is yet another example of Windows,
>especially 9x, brain dead system design that causes programs to be
>larger, slower, more error prone, etc... At least a lot of the NT
>people came from non-PC backgrounds and knew a better, it's a hell
>of a lot better than Win9x in nearly every aspect but is still hampered
>by the Win32 API and all its warts.
Another, slightly longer rant, this one based on the vagaries of
"how to wait for objects". I'm working on a multiple event-waiting
loop for various reasons for my current employer,
so I had a chance to fiddle a bit with, among other things,
Sleep(), WaitForMultipleObjectsEx(), and PeekMessage().
Very interesting quirks, especially when compared with their
Unix equivalents.
* * *
[1] sleep() is in seconds, Sleep() is in milliseconds.
This by itself isn't much, but it is another brain cell that could
probably have been better utilized elsewhere. :-) (To be
fair, there's also a usleep() in Linux and probably some other
Unices, which takes things in terms of *microseconds* -- but warns
that resolution might not be quite that precise. [Not that waiting
is all that precise anyway, on a multiprocess system.] And select()
uses a struct timeval *, which points to tv_sec and tv_usec fields.)
[2] It turns out WaitForMultipleObjectsEx() returns immediately, with
0 handles. Aaargh! (select() will happily sit there until the
death of the universe, or until 2038, whichever comes first... :-) )
[3] It also turns out pseudohandles aren't all that useful to this
call. (I forget what pseudohandle I passed it though, but it
had to do with named pipes.) In fact, I'm not sure what a pseudohandle
*is*...???
[4] Named pipes are....interesting. I still haven't figured them
out yet. I switched to Events, which do actually work more or less
as advertised. (A rough Unix equivalent of an Event might be waiting
on a semaphore, which would actually work pretty well, too.)
I'm still trying to figure out the difference between a named pipe,
and a named pipe instance...
[5] There's a useful concept called a WaitableTimer. Too bad it
doesn't work in NT 4. (Maybe in Win2K, or with a service patch.)
I was able to work around it by using the delay argument in
WaitForMultipleObjectsEx(), though; that seems to work pretty well,
although one has to be careful, lest one run into the 49.7 day bug.
(Yes, NT apps have it too, if one's sloppy about programming delays;
it turns out 49.710 days = 2^32 milliseconds.)
[6] MsgWaitForMultipleObjects() is an interesting philosophy.
In Unix, if one wants to wait for both X events and for, say,
a socket or pipe or two, one can set up a select() call giving the
fid of the connection (which, suprise surprise, is a socket itself)
as one of the items. (It's retrievable using ConnectionNumber(dpyptr).)
Once one detects activity on that socket, one can drain the display's
event queue using any of a number of X calls, the most obvious ones
being XPending(dpyptr), which checks to see if there are events
pending on the queue, and and XNextEvent(dpyptr, eventptr), which lifts
the desired event from the given display's queue. There are also a
number of specialized calls such as XCheckIfEvent().
But does Windows do it the obvious way? Noooooooooo....one is instead
required to use this particular call. The rest of the loop is fairly
straightforward, as it turns out, though, and an extra returned value --
corresponding to the magical handle added -- is not difficult to
process, but internally, I suspect it copies the handle array (plus
its magic slot!), waits on that, then frees the array. Either
that, or it builds a bitmap, just like select()...hmm...
I will say this, though -- the handle arrays are a bit easier to
build than the bitsets that select() uses. But three guesses which
one is faster, and the first two don't count! :-)
[7] One has to be careful with MsgWaitForMultipleObjects() (although
the same can be said for select(), too). Picking off one event
(be it an X event or a Windows event) and then returning to the wait
is not enough; one has to drain the entire queue, and only then
can one return to the wait.
[8] Who's bright idea was it to implement GetSystemTime() and
SystemTimeToFileTime()? Egads! time(time_t *) made much more sense.
(And yes, GetFileTime() requires a file handle. So much for
*that* idea.)
[9] What the heck is a "fiber"? At least, Java calls them
"thread groups"...
[10] CreateProcess() is interesting, too. Unlike Unix's fork/exec
paradigm, processes spawned off by CreateProcess() do not become
zombies if the parent forgets to wait for them, or exits prior
to their completion. (I think this issue is no longer prevalent,
however, on Linux. I'll have to check, though.) But it sure has
a lot of arguments... :-)
* * *
So, what could be a 1-call enterprise (a single select() call, which
can either sleep indefinitely, wait for a few microseconds, or wait
for something to happen, or wait for something to happen within a
few microseconds) turns into 5 function calls, depending on
the situation.
no handles, no delay, no windows:
Sleep(really big value) -- not all that useful, usually :-)
no handles, delay, no windows:
Sleep()
no handles, no delay, windows:
GetMessage()
no handles, delay, windows:
WaitMessage()
handles, delay or no delay, no windows:
WaitForMultipleObjectsEx()
handles, delay or no delay, windows:
MsgWaitForMultipleObjects()
Is it any wonder that, with this sort of philosophy, that Windows
NT requires 32 megs of virtual memory just sitting there?
Ow, my brain hurts.... :-)
(There is one silver lining. It turns out WaitForMultipleObjectsEx()
can also wait for process completions, which makes it a little
cleaner; I'd have to do things with SIGCHLD instead. However,
one could implement *that* using a flag (set by the SIGCHLD handler)
and some checking around the select() call [which would be interrupted
by the signal, if memory serves] -- so it's not that big of a problem.)
* * *
Socket programming is somewhat peculiar, too, although it's
not as bad; the main thing missing is a socket() call; one has
to use WSASocket() instead, which returns, bizarrely enough,
an unsigned value; one has to check for INVALID_SOCKET rather
than merely < 0 (I guess Windows NT expects to support 4
billion file handles per process someday :-) ).
No wonder MFC is so useful. It hides all of this glop. :-)
>
>--
>Chuck Berry lied about the promised land
--
[EMAIL PROTECTED] -- Bill Gates lied about Windows 95 :-)
------------------------------
** 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
******************************
