Linux-Advocacy Digest #177, Volume #30 Sat, 11 Nov 00 10:13:03 EST
Contents:
Re: A Microsoft exodus! (Curtis)
Re: Another Silent Computer :( (mlw)
Re: A Microsoft exodus! (Curtis)
Re: A Microsoft exodus! (Curtis)
Re: A Microsoft exodus! (Curtis)
Re: wtb: Old Digital DEC PDP-8 computer or software (mlw)
Re: Aaron R. Kulkis - Who is this guy? (A transfinite number of monkeys)
Re: wtb: Old Digital DEC PDP-8 computer or software ("MH")
Re: so REALLY, what's the matter with Microsoft? ("Chad Myers")
Re: Linux 2.4 mired in delays as Compaq warns of lack of momentum ("Chad Myers")
Re: Linux 2.4 mired in delays as Compaq warns of lack of momentum ("Chad Myers")
Re: Linux 2.4 mired in delays as Compaq warns of lack of momentum ("Chad Myers")
Re: Linux 2.4 mired in delays as Compaq warns of lack of momentum ("Chad Myers")
----------------------------------------------------------------------------
From: Curtis <alliem@kas*spam*net.com>
Crossposted-To:
comp.os.ms-windows.nt.advocacy,comp.os.ms-windows.advocacy,comp.sys.mac.advocacy,comp.os.os2.advocacy,comp.unix.advocacy
Subject: Re: A Microsoft exodus!
Date: Sat, 11 Nov 2000 08:01:14 -0500
Les Mikesell wrote...
> In outlook email if you do the same, you are executing it with the
> application
> the sender wanted to be associated, not any choice that you have
> made.
I don't know about you, but I control my associations and hence I'm the
one determining which application is used and not the sender.
Take for example, I associated .reg files with my text editor. Are you
saying that if a sender sends me a file with a .reg extension with the
intention of messing up my registry, that they will determine what
application is used to run the reg file (regedit in this case)?
Other files I'm uncertain of, I open with my editor, TextPad. Where is
this sender control that you speak of?
> It does pop up a warning about how this .txt and that .png can
> harm your system. And if you ignore those, you will probable
> ignore the real dangerous ones for consistency.
For text files as well?!!! That's very bad. This is why I wonder how you
guys use that damn application, among other things. I use The Bat! and I
can easily manipulate which extensions produce a warning and which don't.
> Then you think it is correct that the sender should chose what happens
> when you open instead of the recipient? And you continue to think
> that knowing all the trouble it has caused?
I've already dealt with this ridiculous way of looking at things.
> > > > .... if the user chooses to executes it or(white line) not execute
> it
> > >
> > > No - the lines should distinguish *what* you are going to execute.
> > > Dragging the attachment to a program is going over the white line,
> > > but that is your business if you crash.
> >
> > I see no difference. I set my file associations so that I don't need to
> > be dragging and dropping.
>
> Then you always assume the sender's choice is good for you?
The sender never chooses for me. *I* choose what application is used
automatically. I can choose not to use the associated application and
instead use another. All this is readily available at your finger tips.
> Why?
> If you didn't name the attachment, this isn't something you chose.
I definitely cannot choose that. :-) My lack of control however ends
there. Once it reaches my Inbox, even when using the annoying OE, I'm
still in control. It's the ignorant user who isn't.
> > :-) At least you concede that you're the one erecting the wall and not
> > me. Bad!! You'll make Windows unfriendly to those who are competent and
> > the ignorant will remain ignorant forever.
>
> Keeping you from knowing what is about to happen is not friendly.
> Doing what the sender chose instead of the recipient is not friendly.
When a typical, ignorant user double clicks a file, he/she hopes that
Windows will open or run the file with the appropriate or a compatible
application. They get a thrill out of things happening for them in that
fashion. In fact they often get annoyed when the dialog asking what to
open the file with, comes up. MS knows this and Apple knows this. This is
why they implemented file associations. I've had complaints from users
when they groped with an unassociated files context menu and could only
get their attachment viewed, which is meaningless to them, because they
don't understand what they're seeing and they know that the sender
wouldn't have intended to send the gibberish that's in front of them.
They saw this experience as a bad and tedious one.
The sender can take advantage of such users plain and simple. They
correctly realize that this is the type of user that like decisions to be
made for them. Ignorance is the problem.
> The wall does not prevent any choice you might make, it is to
> keep someone else's choice from being the wrong one for you.
I'm quite alright thank you. I make my choices and not the sender. Other
users choose to allow the sender to make choices for them by being happy
with things happening automatically for them. But hey, having them learn
that such an approach is dangerous and teaching them a more informed
approach is too much. One can't burden them with that.
I agree that they should be provided simple means of controlling their
associations but they simply have to learn the basic principles.
> Yes, that is what I meant. While it is unlikely that other systems
> would have (or want) any .vbs files, other types are usable
> across platforms and on other systems may not have the
> names you think control their actions.
As I said, names don't control my systems actions. I control the systems
actions. I may choose to automate how certain file types are dealt with
through associations but that's my choice.
> > I fail to see how my argument is system specific.
>
> That is pretty scary... Nothing but windows does that, and
> in email the content type is specified by MIME headers which
> are a cross-platform standard.
Nothing but Windows has had to cater to the joe user in such a way and at
such a large scale. It's through Windows that computers have reached the
homes of the joe user. You can't expect Windows to be and behave the same
as a system designed for professionals who are very much in the know?
Look at Linux development as they try to cater to the average userbase.
Look at where the Linux GUI is going.
> > > It is like advocating that
> > > people learn to drive only one peculiar kind of car.
> >
> > You're misunderstanding me.
>
> No, I don't think so. Everything you have mentioned
> so far is incredibly platform-specific.
So?
> You can't see that this 'harsh reality' is caused by a bad design in
> the first place? It shouldn't be masked, it shouldn't exist in the
> first place. Email attachments are not the same as desktop files.
> It is the pretense that they are that causes the problem in the first
> place. Drop the pretense and the problem goes away.
Drop the ignorance level and you don't have to create the pretense. Since
dropping the ignorance level is a daunting and a slow task in general,
creating the pretense to increase system usability for such users is a
necessary evil with its obvious drawbacks. File associations are created
without the user really knowing what is happening. They don't wish to
know anyway, since they just want the thing to work. I think you also
said that they shouldn't be expected to learn.
> > People will always be exchanging files. Having to name the file will
> > always be necessary. Having to know what type of file it is will also
> > always be necessary whether it be done through icons or extensions. I
> > don't see how this can or should be hidden from a user that exchanges e-
> > mail and attachments.
>
> People rarely need to exchange programs that run when you open them.
Oh? You really need to get out. I get a lot of jokes in the form of
executables. People exchange applications which are packaged as
executables.
> Maybe never... Getting something that does execute should be noted
> as something different than a text file or image. This isn't impossible
> for the computer to do for you, unless you start with the wrong model.
As I said, most joe users, out of ignorance, just want the thing to be
run. They don't care how MS does it. Just do it, they say.
> > If you don't know what to look for when viewing these files 'unsafe'
> > files, how are you going to bless them as being safe. Don't you have to
> > !!learn!! what to look for when viewing them??
>
> No, the MIME types tell what data type is enclosed and the mailer
> should use a safe handler for that type or warn you that there is
> no safe handler.
An what does the joe user say? Especially when their animated jokes keep
being labelled as dangerous. And they all have the little .exe at the end
of their names?
> This is something the mailer should do for you, and it shouldn't
> allow its handlers to be changed accidentally as a side effect of
> installing a new program.
You need to work with some home users and not only those in the office
setting where help is always near when using your superior setup.
--
ACM.
________________________________________________________
"A thing is not necessarily true because a man dies for it."
------------------------------
From: mlw <[EMAIL PROTECTED]>
Subject: Re: Another Silent Computer :(
Date: Sat, 11 Nov 2000 08:08:13 -0500
[EMAIL PROTECTED] wrote:
[snippage]
Finding a single issue and saying it is a system wide issue is a bit
self serving for your purpose, isn't it?
There are many installation issues with Windows which fall under the
term difficult. Just upgrade to an ATI RAGE IIC in a functioning Windows
98 box and many things will stop working. Why? because they can't
allocate a suitable frame buffer. Make sense? of course not, it isn't
supposed to. It means that ATI only supports a newer version of
direct-draw, different from what is currently installed, and you must
download a new version and install that from Microsoft.
For every installation issue in Linux, you can easily find worse ones
from Windows.
This is why I always say installation is not usability. People with
knowledge of the systems must do an install. Chances are, with either
system, your install will be fine, but it is not guaranteed.
Once up and running, Linux is more problem free, more predictable, more
stable, and way more easy and fun to use.
--
http://www.mohawksoft.com
------------------------------
From: Curtis <alliem@kas*spam*net.com>
Crossposted-To:
comp.os.ms-windows.nt.advocacy,comp.os.ms-windows.advocacy,comp.sys.mac.advocacy,comp.os.os2.advocacy,comp.unix.advocacy
Subject: Re: A Microsoft exodus!
Date: Sat, 11 Nov 2000 08:08:49 -0500
Aaron R. Kulkis wrote...
> > You simply cannot use a computer effectively and safely without first
> > learning how to use it.
>
> But Microsoft claimst that THEIR PRODUCTS are "intuitive", and thus,
> all of that messy "learning stuff" is no longer necessary.
>
>
> Since your argument in defense of Microsoft products DIRECTLY CONTRADICTS
> one of the foundation-level premises of Microsoft product design,
> this defense is null and void by reason of internal contradiction
>
> GAME
> SET
> MATCH
I'm playing no games. My advocacy for Win2k doesn't encompass,
advertising and media hype statements coming from Redmond. Only an idiot
would take these sales pitch claims seriously or to the extreme that you
seem to.
Those statements can be taken on relative terms. There are intuitive
aspects to Windows. There's no doubt about that. It does prevent a lot of
the messy learning stuff that would have been necessary to achieve the
same goals with a UNIX system.
But of course wise guys like yourself will try to take these statements
to the extreme and in so doing make yourselves look stupid rather than
those you intended to.
> You LOOOOOOOOSE!
Me? Nah! :-) Read again.
--
ACM.
________________________________________________________
"A thing is not necessarily true because a man dies for it."
------------------------------
From: Curtis <alliem@kas*spam*net.com>
Crossposted-To:
comp.os.ms-windows.nt.advocacy,comp.os.ms-windows.advocacy,comp.sys.mac.advocacy,comp.os.os2.advocacy,comp.unix.advocacy
Subject: Re: A Microsoft exodus!
Date: Sat, 11 Nov 2000 08:14:23 -0500
Aaron R. Kulkis wrote...
> > > it is precisely BECAUSE MS FAILS TO DISTINGUISH BETWEEN "OPENING" AND "RUNNING"
> > > that e-mail viruses are so rampant on Microshaft Operlosing Systems.
> >
> > Actually we do know the difference.
> >
> > How is joe user going to be able to tell what's what in a perl script
> > file when he opens it to view the contents rather than just run it?
>
> Maybe he can understant it,...maybe he can't.
MOST will not understand. Most use Windows. If you think otherwise, you
need to get out more.
> At least he knows it's executable code.
How?
> And he has the OPTION of finding someone who can read the code
> before executing it.
Oh really now? Hahahahahaha. You really don't know what you are talking
about do you? This is why you sound so ridiculously unrealistic in your
arguments.
> > Hmmmm?
>
> So....what you're saying is....because some user's are not able
> to properly evaluate scripts....ALL users should be penalized by
> having scripts run for them without an opportunity for evaluation.
All users are not. All users have the option to evaluate the script
before running it in Windows. Didn't you know this?
> You write like an 'everybody must suffer equally" communist.
>
> Why is that?
Would you please learn some Windows before wielding these Strawman
arguments in such an arrogant manner? You guys who claim to know so much
can be so darned disappointing to read. <sigh>
--
ACM.
________________________________________________________
"A thing is not necessarily true because a man dies for it."
------------------------------
From: Curtis <alliem@kas*spam*net.com>
Crossposted-To:
comp.os.ms-windows.nt.advocacy,comp.os.ms-windows.advocacy,comp.sys.mac.advocacy,comp.os.os2.advocacy,comp.unix.advocacy
Subject: Re: A Microsoft exodus!
Date: Sat, 11 Nov 2000 08:30:39 -0500
Aaron R. Kulkis wrote...
> Sam Morris wrote:
> >
> > Facts which apparantly have not prevented you from acting like one of the
> > most stupid and prejudiced people I have ever seen posting to these NGs. I
> > guess the .sig really does say it all.
>
>
> It serves the purpose for which *I* want it to serve.
<shaking head> And you claim that you know so much, flaunting your
supposed credentials. Do you know how much bandwidth you're wasting with
that unnecessarily long signature that is *repeatedly* included with
every single message that you send? It's such an unnecessary waste. Those
who really know would be annoyed to the max with your practice.
--
ACM.
________________________________________________________
"A thing is not necessarily true because a man dies for it."
------------------------------
From: mlw <[EMAIL PROTECTED]>
Subject: Re: wtb: Old Digital DEC PDP-8 computer or software
Date: Sat, 11 Nov 2000 08:33:51 -0500
MH wrote:
[snip]
You call me a liar, in a public forum, with no proof. That is a bad
thing, whether you admit it or not. Don't you think that is a serious
charge? One which you should be certain before you accuse? You do not
call someone a liar without proof.
It is quite all right to say someone is mistaken and explain why, but
that is different, do you understand why?
I am not "thin skinned" at all, I can get into an insult fight just as
well as any person, but it is tiresome in a discussion when people
substitute insults and character assassination for reason. If I'd wanted
that I be in politics. You do it almost all the time, and it just a
waste of bandwidth.
When someone makes a valid point, you insult them or call them liars in
an attempt to diminish the point in the discussion. This is the behavior
of children, geez, my 9 year old can discuss things in a more rational
manner.
--
http://www.mohawksoft.com
------------------------------
From: [EMAIL PROTECTED] (A transfinite number of monkeys)
Subject: Re: Aaron R. Kulkis - Who is this guy?
Date: Sat, 11 Nov 2000 13:41:27 GMT
On Fri, 10 Nov 2000 13:23:20 -0500, Aaron R. Kulkis <[EMAIL PROTECTED]> wrote:
: Also 20 credit hours of calculus
20 hours of calculus, eh? Wow, I've only got a BA in Mathematics, so take
my results with a grain of salt. If you want to talk just calculus, I had
12 hours. Calculus I through IV, 4 semesters of class-time. Now, if you
count 2 semesters of Real Analysis and a semester of Complex Analysis, as
being "calculus", then I've got 21. I wouldn't count the Analysis stuff
as calculus, however. Most of the material is the same as what you cover
in calculus, except rather than accepting all of those things the instructor
told you are "just true, and given facts", you prove them. You prove things
like the Fundamental Theorem of Algebra, and other fun facts.
: Aaron R. Kulkis
: Unix Systems Engineer
:
: [alt.fan.warlord candidate clipped]
--
Jason Costomiris <>< | Technologist, geek, human.
jcostom {at} jasons {dot} org | http://www.jasons.org/
Quidquid latine dictum sit, altum viditur.
------------------------------
From: "MH" <[EMAIL PROTECTED]>
Subject: Re: wtb: Old Digital DEC PDP-8 computer or software
Date: Sat, 11 Nov 2000 09:40:59 -0500
"mlw" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> MH wrote:
> [snip]
>
> You call me a liar, in a public forum, with no proof. That is a bad
> thing, whether you admit it or not. Don't you think that is a serious
> charge? One which you should be certain before you accuse? You do not
> call someone a liar without proof.
This is truely a tiresome tirade. Again you do not respond to the issue, you
rarely do unless it's to your advantage to do so. This advantage I cannot
acsertain, unless it is to see your "expertise" in print. Please, once and
for all, quit saying I called you a liar, will you?
I never once used that term and you and everyone else knows it. If you think
I did so then supply the post where I did, and repost it for all of us to
see. You'll no sooner do that than you will address an issue you aren't
equipped to deal with. Which seems to be any that call into question your
"fling snot and duck" approach to linux advocacy.
> It is quite all right to say someone is mistaken and explain why, but
> that is different, do you understand why?
I haven't a clue of what you are referring to, where this fits into the
context of what is being discussed, or what this has to do with anything in
this thread. I made a joke by stripping away a piece of your post and making
a comment totally in fun -- with a smiley face. That's the thread. Can you
stick to it? Talk to your 9 year old some more. I imagine you're having a
more equitable conversation than you are here, because you sound like G.W.
Bush.
> I am not "thin skinned" at all, I can get into an insult fight just as
> well as any person, but it is tiresome in a discussion when people
> substitute insults and character assassination for reason. If I'd wanted
> that I be in politics. You do it almost all the time, and it just a
> waste of bandwidth.
I don't agree. I think you are thin skinned, and I think everytime you
respond as you do when questioned during one of your free-range MS bashing
sessions, it shows.
I don't think we've broken the surface yet.
>someone makes a valid point, you insult them or call them liars in
> an attempt to diminish the point in the discussion. This is the behavior
> of children, geez, my 9 year old can discuss things in a more rational
> manner.
This simply isn't true at all. You wonder why I ever question you on typos
and misspellings?
I'll tell you why. You once stated in this very group that you were once a
proof reader.
Next, you're a software engineer, among countless other professional
endeavors. Now, once and for all, I'm not saying you're lying about these
claims. But I don't find it obtuse to make note when someone who makes these
very claims misspells the word architect, then uses "phonetic" as the
excuse. Proofreader + Software engineer & architect != more than the rare
typo or misspelling. Your postings don't follow this logic. I don't feel
that is an outrageous expectation or conclusion.
Unless you truly ARE using the fling snot and duck approach.
Your rhetorical stance is, and has always been one of expertise backed up by
these very claims. But you should realize that to hold yourself as a paragon
among the peanut gallery is to invite just what you seem to be in such
distaste of. In other words, get a clue.
No, don't. I take that back. Just go on trying to "take the high ground"
when your wits seem to fail you in your ability to argue a point. It's
amusing to me. You see, I graduated from the 10th. grade, went on to get a
GED, a technical diploma from a one year school, and 50 credits from a
community college. I'm not an "engineer", or a "proof reader", or an
"architect". I guess I'm just one of your lowly MS peons so keep it
flinging!
------------------------------
From: "Chad Myers" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.advocacy,comp.os.ms-windows.nt.advocacy
Subject: Re: so REALLY, what's the matter with Microsoft?
Date: Sat, 11 Nov 2000 14:29:38 GMT
"Mike Byrns" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> chrisv wrote:
>
> > "Chad Myers" <[EMAIL PROTECTED]> wrote:
> >
> > >Therefore... *PL0NK*
> >
> > Oops! You lose! And you were doing so well up until then....
>
> Heh. Losing is when you are forced to concede not when the other yahoo
> loses all by themselves :-)
Exactly, the debate was pretty one-sided and I grew tired of
continually explaining to him what the topic was about and trying
to keep him on focus.
He wasn't particularly interested in the topic, but was perfectly
content to continue bashing Windows with not even the slightest attempt
at providing logic behind his claims.
The word "because" is not in his vocabulary (along with several thousand
other words).
- Chad
------------------------------
From: "Chad Myers" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Linux 2.4 mired in delays as Compaq warns of lack of momentum
Date: Sat, 11 Nov 2000 14:32:26 GMT
"Ayende Rahien" <[EMAIL PROTECTED]> wrote in message
news:8uipuq$o9i$[EMAIL PROTECTED]...
>
> "Les Mikesell" <[EMAIL PROTECTED]> wrote in message
> news:pM2P5.18368$[EMAIL PROTECTED]...
> >
> > "Chad Myers" <[EMAIL PROTECTED]> wrote in message
> > news:FYJO5.2465$[EMAIL PROTECTED]...
> > >
> > >
> > > You can't have an enterprise-ready application with a faulty and
> > > half-baked file system (ext2). Even the MS you consider to be so
> > > inferior knows this.
> >
> > Ext2 has no more problems that NTFS - less if you ever let an
> > NTFS fill with tiny files.
>
> Fixed.
>
> > NT has been under development for years. Have you warned
> > people away from using it?
>
> One problem I've with OSS model is that it's always under development.
> I don't know about you, but I think that at some point (check topic) someone
> has to say "This is as much as we are going to put into this version. From
> now on, we will only pull out the bugs."
That's one of the many fundamental problems with OSS. No one takes charge
and therefore no one accepts blaim. If they put a product out and claim it's
great and wonderful, a bunch of people use it and it fails miserably (as they
usually do) they just back up and say, "Well, hey, it's under development!"
There's no responsibility and therefore no trustworthiness.
-Chad
------------------------------
From: "Chad Myers" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Linux 2.4 mired in delays as Compaq warns of lack of momentum
Date: Sat, 11 Nov 2000 14:33:38 GMT
"Craig Kelley" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> "Ayende Rahien" <[EMAIL PROTECTED]> writes:
>
> > "Goldhammer" <[EMAIL PROTECTED]> wrote in message
> > news:vD%O5.75454$[EMAIL PROTECTED]...
> >
> >
> > > Yes, that's entirely correct. The use of Raw partitions have nothing
> > > to do with database sizes. It's an issue of performance. However,
> > > it must be said that you may not gain anything from raw partitons
> > > at all. IIRC, Oracle has also been recommending the converse route, if,
> > > for example, you are intending to set up Oracle on Sparc machines
> > > running Solaris.
> > >
> > > The bottom line is: people can and do maintain >50 Gb databases
> > > (Oracle or otherwise) under linux without dedicated partitions,
> > > and thus, of course, Myers's comment is complete nonsense.
> >
> > Very few db, if any, keeps databases in one huge files.
> > It's inefficent to do so.
>
> Just ask the computing department in my organization that runs
> Microsoft Exchange. It stores all mail, calendar and related items
> for everybody in one huge file.
>
> They came to us for help when it became corrupt. We had to use linux
> and some open-source tools to dissect a bad backup tape.
And you're blaiming your incompetent computer dept. operators on Exchange?
They didn't do good backups, and now it's Exchange's fault?
-Chad
------------------------------
From: "Chad Myers" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Linux 2.4 mired in delays as Compaq warns of lack of momentum
Date: Sat, 11 Nov 2000 14:34:26 GMT
"Craig Kelley" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> "Les Mikesell" <[EMAIL PROTECTED]> writes:
>
> > "The Ghost In The Machine" <[EMAIL PROTECTED]> wrote in
> > message news:[EMAIL PROTECTED]...
> >
> > > >So you can't use Oracle on Linux for >2GB databases without fancy
> > > >techniques or special filesystems.
> > > >
> > > >Thank you for finally ending this thread of this topic.
> > >
> > > I'd put it even more strongly than that. It is IMPOSSIBLE to use
> > > Oracle on Linux for >2GB databases on *any* file system, on
> > > *any* 32-bit machine, without a recompile using the above options
> > > and/or explicit use of the xxx64() API and/or multiseeking techniques.
> >
> > Note that Postgres does it whether or not the OS supports 64 bit
> > operations, so it is certainly not impossible. Oracle may not
> > do without raw partition access but it would not be impossible.
>
> You don't need to use raw partitions either. We have an 80GB Oracle
> database which runs on top of ext2 on an unmodified Linux 2.2 kernel.
Please enlighten us how you manage to do this? Multiple tablespaces?
Do you span the files out into multiple files? This has a large impact
on performance...
-Chad
------------------------------
From: "Chad Myers" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.ms-windows.nt.advocacy
Subject: Re: Linux 2.4 mired in delays as Compaq warns of lack of momentum
Date: Sat, 11 Nov 2000 14:40:36 GMT
"Paul Colquhoun" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> On Sat, 11 Nov 2000 05:07:41 GMT, Chad Myers <[EMAIL PROTECTED]>
wrote:
> |
> |"Goldhammer" <[EMAIL PROTECTED]> wrote in message
> |news:8uij0g$a2s$[EMAIL PROTECTED]...
> |> In article <[EMAIL PROTECTED]>,
> |> [EMAIL PROTECTED] (The Ghost In The Machine) wrote:
> |>
> |>
> |> > I'd put it even more strongly than that. It is IMPOSSIBLE to use
> |> > Oracle on Linux for >2GB databases on *any* file system, on
> |> > *any* 32-bit machine, without a recompile using the above options
> |>
> |>
> |> Pardon me - and I mean this in the gentlest manner - but aren't
> |> you being just a bit dense, in this instance? Properly concieved
> |> dbm systems support constructs analogous to Oracle's dbspaces
> |> and tablespaces, effectively rendering database & table sizes
> |> independent of machine architecture or OS constraints.
> |>
> |>
> |> > Note that two relative seeks should accomplish what one absolute seek
> |> > cannot, but extending this for files that could be as huge as
> |> > 20 gigabytes -- or 200 gigabytes -- or 200 terabytes! -- becomes
> |> > a royal pain.
> |>
> |>
> |> Why are so many of you hung up about filesizes when the
> |> comment being discussed had to do with database sizes? A database
> |> is not a file.
> |
> |When you create a tablespace in Oracle, is not the syntax:
> |
> |CREATE TABLESPACE <tablespace name>
> |DATAFILE '<path to data file>'....
> |
> |What, exactly is the "datafile" if not a file, then?
>
>
> And when that file is almost full, you add more tablespace by
> telling Oracle about *another* *differnet* *file* that it can use.
>
> At that point, the "datafile" consists of 2 OS files.
>
> Repeat as necessary.
Hmmm... not the way I do it. I use Autoextend and make sure
that:
a.) the db file is large enough to make autoextending rare
b.) make the autoextend size large enough that it will prevent
frequent extending
Here's a sample SQL script I use to create a 10M example database.
Note: I'm not a dba, so this script probably sucks, but it
illustrates my point well enough.
CREATE TABLESPACE TEST_DATA
DATAFILE 'c:\Oracle\oradata\sidname\testdata01.dbf'
SIZE 10M
AUTOEXTEND ON
NEXT 10M
MAXSIZE UNLIMITED
DEFAULT STORAGE (
INITIAL 10K
NEXT 10K
MINEXTENTS 1
MAXEXTENTS UNLIMITED
PCTINCREASE 1
)
ONLINE
PERMANENT
/
(and then I'd create the index, but you get the point)
Wouldn't having two files slow down performance?
But then, I guess, if you're using Oracle on Linux,
performance obviously isn't a concern.
Especially if you were planning on having an 80GB db which
would result in 40 db files at 2GB a piece!
> Sorry I can't remember the exact commands involved. I did peer
> over our DBA's shoulder while he did this once, but my notes
> are at the office.
I know what you're talking about, and I believe you, but that's
not the optimal way to do things, it's more of a hack to get
around the completely worthless Ext2 filesystem.
-Chad
------------------------------
** 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
******************************
