Linux-Development-Sys Digest #437, Volume #6 Sat, 27 Feb 99 19:14:22 EST
Contents:
Re: How do I fetch the cursor key block with vga_getch? (Dirko van Schalkwyk)
Re: The REAL question: Can Linux... (Steve A)
Re: Edlin in Linux? (Jakob Eriksson)
Re: The REAL question: Can Linux... ("Robert H. Thompson")
Re: The REAL question: Can Linux... ("David Sisk")
Re: Overclocking (was: Re: K6-2 and Linux, Are there any Bug?) (GBP)
Re: IP to process network interface? (Kip Rugger)
Re: monolithic kernel and source that can only be compiled as a module ("Bjorn
Wesen")
Re: Don't want to loose win '98 docs etc.., & want Linux NOW! (Adam P. Jenkins)
Re: Linux Threads use of SIGUSR1, SIGUSR2 (Rakesh Tiwari)
Re: /dev/zero (Konrad Mierendorff)
where got sis 6326 driver? ("Yeh Hung Tan")
Re: ATX Power Off problem (Dave Platt)
----------------------------------------------------------------------------
From: Dirko van Schalkwyk <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps
Subject: Re: How do I fetch the cursor key block with vga_getch?
Date: Sat, 27 Feb 1999 20:35:41 +0200
Reply-To: [EMAIL PROTECTED]
Norman Beran wrote:
> I want to react on the cursorkeys but I get for up and down the same
> numbers from vga_getch.
>
> Thanks a lot!
>
> Norman
Well I have'nt tried to get cursor-keys in linux yet, but I remember in
the old days of me dos-and-pascal livf you had to read twice to get a
cursor-key.
You would always get a 0 and then a number. Something like 0 and 78 is
up or something. Excuse my english.
Hope it helped.
Dirko
------------------------------
From: [EMAIL PROTECTED] (Steve A)
Subject: Re: The REAL question: Can Linux...
Date: Sat, 27 Feb 1999 18:14:55 GMT
[posted/mailed]
On Fri, 26 Feb 1999 17:05:44 GMT, [EMAIL PROTECTED] wrote:
> From a practical perspective, can Linux truly act as a full server to Windows
> 95/NT boxes? I want the server to have a SQL database that is accessed and
> updated by the Windows machines. For example, users would have to be able to
> use MS-Access to manipulate the database. How would the ODBC drivers on the
> Windows machines work in this case?
A treat. I do exactly that using mysql and adabas. I'm downloading a
trial version of Oracle 8i at the moment, so I can do it with that,
too.
Samba will give you Windows file'n'print services, and Apache will
serve web pages for an intranet.
For the price of that 128Mb upgrade for your NT server, you could run
a DBMS/web/mail/file/print server on the same machine and probably get
similar or better performance.
> If possible, give me a working example (with a SQL database whether commercial
> or open that would be comparible to what MS would offer).
Available on Linux at the moment: Informix, Sybase 11, Oracle 8, IBM
DB/2 5.2, MySQL and a large number of other less mainstream commercial
SQL DBMS's.
You won't get quite the same non-standard stuff as you would running
SQL Server on NT, but you'll actually be more in line with the
mainstream SQL people.
> This sql database in our example would also be published on the web, with the
> linux server running apache.
All fine.
> BTW Could you house access database files on a linux server and have the
> local Win boxes connect to it? are there odbc drivers in linux that could
> manipulate the access database files?
You can house the files on the server, but that isn't the same as
running a client/server DBMS. Access has some funny attitudes about
locking and suchlike, and a shared database file cannot (AIR - I'd be
pleased to be told I'm wrong) be updated by multiple users.
--
Warning: end of message imminent. Stop reading now.
We operate a zero-tolerance policy on UCE and spam.
------------------------------
Crossposted-To: comp.sys.amiga.hardware
Date: Sat, 27 Feb 1999 19:41:40 +0100
From: Jakob Eriksson <[EMAIL PROTECTED]>
Subject: Re: Edlin in Linux?
On 25 Feb 1999, Per Olsson wrote:
> BTW, is there a Linux equivalent to DOS edlin? I'm asking just out of
> curiousity.
>
> Edlin is installed with Windows NT 4, why I dont know. Maybe to satisfy
> those used to edlin that find it hard to learn notepad.
>
Should be "ed".
And if so, then "edlin" is rather the equivalent to UNIX "ed".
Jakob
------------------------------
From: "Robert H. Thompson" <[EMAIL PROTECTED]>
Subject: Re: The REAL question: Can Linux...
Date: Fri, 26 Feb 1999 12:46:32 -0500
Reply-To: [EMAIL PROTECTED]
[EMAIL PROTECTED] wrote:
>
> From a practical perspective, can Linux truly act as a full server to Windows
> 95/NT boxes?
>
Yes. Samba facilitates the networking of Linux and Win95/WinNT boxes. I
have this setup at home.
> I want the server to have a SQL database that is accessed and
> updated by the Windows machines. For example, users would have to be able to
> use MS-Access to manipulate the database. How would the ODBC drivers on the
> Windows machines work in this case?
>
Avoid this headache and do it on the web. Oracle, Sybase, Informix and
Ingres are all available for Linux web enabled. The database processing
could be done via a web browser, Perl/CGI and Apache.
> If possible, give me a working example (with a SQL database whether commercial
> or open that would be comparible to what MS would offer).
>
See the above. There are several others that come packages with most
Linux applications. The big question is the number of users, what type
of database/application server hardware are you talking about. The are a
lot of specifics that have to be considered. But as a proof of concept
this can be done.
> This sql database in our example would also be published on the web, with the
> linux server running apache.
>
> BTW Could you house access database files on a linux server and have the
> local Win boxes connect to it? are there odbc drivers in linux that could
> manipulate the access database files?
If its on the web and and the database is web enabled you don't need to
worry about it. If you use Samba any files stored on the Linux box will
be accessible to the Win95/WinNT machines including MS Access database
files.
<snip>
Hope this helps
Rob Thompson
DBA
Ford Motor Company
home: [EMAIL PROTECTED]
work: [EMAIL PROTECTED]
------------------------------
From: "David Sisk" <[EMAIL PROTECTED]>
Subject: Re: The REAL question: Can Linux...
Date: Sun, 28 Feb 1999 12:05:35 -0500
OK, let me make sure I understand. You want to run a database server on
Linux, and be able to access it from Windows95/NT clients, and via Apache on
the same box.
The commercial choices for a database server that runs on Linux are Oracle8
standard edition, Informix, Sybase, DB2 (I think), and some of the
open-source databases (like Postgre SQL). I'll use Oracle as an example,
since that's the one I know the most about.
You run the Oracle8 instance on the Linux. In addition to the Oracle8
instance, there will be a Listener process (the NET8 server) that listens
for client connection requests over NET8. The client machines (whether
Win95/NT or other Linux boxes) will need to have the Oracle NET8 client
installed. These workstations will then be able to seamlessly talk to the
Oracle8 database on the Linux box. To use through MS Access, you will have
to setup a DSN that references the Oracle8 ODBC driver. Access talks to the
ODBC manager which talks to the Oracle8 ODBC driver which talks to the
Oracle8 Net8 client which talks (over TCP or other protocols) to the Oracle8
Net8 server/Oracle8 database instance. As long as you have the NET8 or
SQL*Net servers and clients, any platform can talk to an Oracle instance on
any other platform.
Making Apache on the same box talk to the Oracle8 instance should be
not-too-difficult, but since I'm no Apache expert, I'll leave that one for
someone else to answer fully. You could write CGI programs in C that talk
to Oracle either by the Oracle8 pre-compiler or through the Oracle8 OCI. I
*think* I remember seeing some sort of "Apache modules" that allow Perl
scripts (???) to talk to Oracle from Apache.
An MSAccess database placed on the Linux box could be accessible to the
Win95/NT clients if you use SAMBA to make the storage location visible to
the Win clients (like a file share on NT). You would not be able to have
Apache talk to the MSAccess database though. The access database engine
runs locally on each client (that's why it has to be installed on each
client). It cannot run under Linux, so Apache can't talk to the MDB file.
(I'm not sure you'd even want to do this if you could. Depends on the
expected throughput requirements, though.)
By the way, running Oracle (and probably Informix as well) on Linux will
absolutely kick the snot out of MS SQLServer on WindowsNT, given the exact
same hardware!
Hope this helps.
--
David C. Sisk
The Unofficial ORACLE on NT site
http://www.ipass.net/~davesisk/oont.htm
[EMAIL PROTECTED] wrote in message
<7b6k90$fhj$[EMAIL PROTECTED]>...
>From a practical perspective, can Linux truly act as a full server to
Windows
>95/NT boxes? I want the server to have a SQL database that is accessed and
>updated by the Windows machines. For example, users would have to be able
to
>use MS-Access to manipulate the database. How would the ODBC drivers on
the
>Windows machines work in this case?
>
>If possible, give me a working example (with a SQL database whether
commercial
>or open that would be comparible to what MS would offer).
>
>This sql database in our example would also be published on the web, with
the
>linux server running apache.
>
>BTW Could you house access database files on a linux server and have the
>local Win boxes connect to it? are there odbc drivers in linux that could
>manipulate the access database files?
>
>Thanks in advanced. I have been using linux and have become familiar with
the
>environment...but the answers to these questions will allow me to devote
time
>into diving right in.
>
>AP
>
>-----------== Posted via Deja News, The Discussion Network ==----------
>http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: GBP <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Crossposted-To:
alt.os.linux.slackware,comp.os.linux.hardware,comp.os.linux.misc,comp.os.linux.setup
Subject: Re: Overclocking (was: Re: K6-2 and Linux, Are there any Bug?)
Date: Fri, 26 Feb 1999 13:36:11 -0500
Yeah point well taken. But is it true that overclocking can result in
system instability? When netscape crashes how am i going to know it was
a bug and not my CPU doing and instruction wrong or something? When
people say instability what exactly do they mean? do these machines
freeze?
gbp
------------------------------
From: [EMAIL PROTECTED] (Kip Rugger)
Crossposted-To: comp.os.linux.networking
Subject: Re: IP to process network interface?
Date: 25 Feb 1999 10:58:15 -0600
>Is there a way to get an IP network interface into the kernel so
>that the kernel can operate on it as any network interface, but
>have the interface really connect via a process on that machine?
>
>Any IP packets sent to that interface would be given to the process,
>and any packets that process produces through the appropriate API to
>this facility, would "arrive" into the IP stack in the kernel.
Sounds like the ethertap device is what you want:
CONFIG_ETHERTAP
If you say Y here (and have said Y to "Kernel/User network link
driver", above) and create a character special file /dev/tap0 with
major number 36 and minor number 16 using mknod ("man mknod"), you
will be able to have a user space program read and write raw
Ethernet frames from/to that special file. tap0 can be configured
with ifconfig and route like any other Ethernet device but it is not
connected to any physical LAN; everything written by the user to
/dev/tap0 is treated by the kernel as if it had come in from a LAN
to the device tap0; everything the kernel wants to send out over the
device tap0 can instead be read by the user from /dev/tap0: the user
mode program replaces the LAN that would be attached to an ordinary
Ethernet device. Please read the file
Documentation/networking/ethertap.txt for more information.
------------------------------
From: "Bjorn Wesen" <[EMAIL PROTECTED]>
Subject: Re: monolithic kernel and source that can only be compiled as a module
Date: Fri, 26 Feb 1999 20:49:05 +0100
Martin Recktenwald <[EMAIL PROTECTED]> writes:
> > Security. If someone hacks your box, what better back door than a
> > kernel module. IIRC, there's one out ther called "mal" that does
> > exactly this.
>
> If someone hacks you box he could
> - install a new kernel (you might notice the reboot)
> - install a new login program
> - a new inetd
> - hacked tcp-wrappers
> - sendmail with backdoor
> - <replace you favorite binary here>
>
> And in fact he would do all of this.
Nah.. most break-ins are by script kiddies and they'll just dump a canned bd
onto some daemon or login or whatever. Thorough auditing of all important
files regularly (by different systems) is a good way of keeping track of
breakins. What you forgot is, that with a kernel module hack, you have no
way of noticing that the system HAS been hacked, but as long as you can
trust the kernel itself, you have a LOT of options to detect the hack.
The cleaning up and restoring the system after a break-in is a totally
different matter.
I fully support the poster who want non-module kernels, I do that myself
also (very little point having modules on a server anyway).
/Bjorn
------------------------------
Subject: Re: Don't want to loose win '98 docs etc.., & want Linux NOW!
From: [EMAIL PROTECTED] (Adam P. Jenkins)
Date: 27 Feb 1999 16:52:54 -0500
[EMAIL PROTECTED] (Matthew Bloch) writes:
> On 27 Feb 1999 04:32:08 GMT, Kishore <[EMAIL PROTECTED]> wrote:
> >Hi Folks,
> >I have a Cyrix MII ,128 mb, 6.4 gb HD PC.
> >Win'98 is loaded throughout this disk.
> >I have lot of stuff(docs, prgs etc..,) on my PC. I don't want to loose
> >them.
> > Is there any way that I can load Linux to share this and make my PC dual
> >bootable. I want to allot 3gb to Linux and 3.4 gb should remain for MS.
> >Please give a detailed input if pos'
>
> Actual repartitioning will always involve scrubbing the data on your disc;
> there's no two ways about it. There's basically answers that don't
This is completely untrue. You can use the FIPS program to
non-destructively repartition your disk from DOS, or you can buy
PartitionMagic and do it directly from Win95. Both of these allow you
to shrink existing partitions without destroying their contents, and
create new partitions. FIPS is available for free at most Linux
sites.
I've done it several times with FIPS with no problem, without
destroying or backing up an existing Windows installation.
> involve repartitioning; buy another hard drive and entirely devote it to
> Linux (which is what I did; even finding an old 500Mb drive would do if
> you're just testing the waters), or there's a particular Linux
> distribution (the name escapes me) which uses DOS as its primary
> filesystem. It's a bit ugly, but saves you from repartitioning.
>
> --
> Matthew ( http://www.soup-kitchen.demon.co.uk/ )
>
>
--
Adam P. Jenkins
[EMAIL PROTECTED]
------------------------------
From: Rakesh Tiwari <[EMAIL PROTECTED]>
Subject: Re: Linux Threads use of SIGUSR1, SIGUSR2
Date: Sat, 27 Feb 1999 06:12:54 -0600
This relates to threads, may not relate to
exactly what you folks are talking about (sorry
for butting in).
I wrote a simple program in Solaris which
allowed me to suspend/resume threads from
keyboard within a process (kill SIGUSR1/SIGUSR2
pid).
Same program will crash on Linux since pthread (I
think) creates a process per thread.
My question is, what version of pthreads is likely
to have decent implementation of threads which will
allow me to suspend/resume threads arbitrarily ?(more
like NT, basically I want to implement ResumeThread/
SuspendThread calls) and not created stupid processes ?
Thanks,
Rakesh.
Peter Samuelson wrote:
>
> [Jim Cromie <[EMAIL PROTECTED]>]
> > > Speed? Signals are historically un-reliable, I kinda thought this
> > > meant that they were slow too.
>
> No reason the two should have any connection. Solaris is slow, Linux
> 2.1.44 was unreliable, NT is both. (:
>
> [David Wragg <[EMAIL PROTECTED]>]
> > Historically they have been seen as hard to program with, because
> > there are strict conditions on what it is safe to do in a signal
> > handler, and because there are pitfalls when using any kind of
> > asynchronous facility. As a result, people would write na=EFve code
> > that worked some of the time, but failed occasionally. I suppose
> > this gave the false impression that signals are unreliable.
>
> You have a good point ... but probably the main reason signals have
> been seen as unreliable is because on some older Unices (4.3BSD?) they
> really *were* unreliable. Signal handlers were single-use and had to
> reinstall themselves, and this meant a race window after a handler got
> called and before it called signal() to reinstall itself, and any more
> of the same signal would either get dropped or would kill the process,
> depending on the particular signal. Come to think of it, even in POSIX
> you will drop duplicate signals, since they don't get queued past a
> single bitmask, but at least you don't have to die from a SIGINT you
> wanted to handle.
>
> --
> Peter Samuelson
> <sampo.creighton.edu!psamuels>
------------------------------
From: Konrad Mierendorff <[EMAIL PROTECTED]>
Subject: Re: /dev/zero
Date: Fri, 26 Feb 1999 21:04:47 +0100
][ndigo - Stormy blue sky wrote:
>
> Hi.
> Can anyone explain me the purpose of this file (except for DCC send to
> friends)?
I use /dev/zero to create swapfiles ( something like 'dd if=/dev/zero
of=/swapfile ...')
Bye ...
------------------------------
From: "Yeh Hung Tan" <[EMAIL PROTECTED]>
Subject: where got sis 6326 driver?
Date: Sun, 28 Feb 1999 03:22:08 +0800
anyone who know where can download the sis 6326 driver for linux redhat 5.0?
my displaycard is built-in the motherboard. and i couldnt load into
xwindows...
------------------------------
From: [EMAIL PROTECTED] (Dave Platt)
Subject: Re: ATX Power Off problem
Date: Fri, 26 Feb 1999 20:44:31 GMT
>Thank you Phil, I indeed am considering buying a UPS, but my problem
>might still remain: after Linux give a shutdown command, will
>the machine goes up automatically when the power is back?
As far as I know, without explicit BIOS support or some external
hardware, you can't get the machine to re-power itself automatically.
I've seen some ATX systems with a BIOS which gives you a three-way
choice for what to do when the power is applied: don't power on,
always power on, or "restore previous state". You'd probably want to
do the always-power-on version.
--
Dave Platt [EMAIL PROTECTED]
Visit the Jade Warrior home page: http://www.radagast.org/jade-warrior/
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!
------------------------------
** 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.development.system) 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-Development-System Digest
******************************
