Linux-Misc Digest #918, Volume #24               Sat, 24 Jun 00 20:13:03 EDT

Contents:
  Re: Internet backup services for Linux (Mark Wang)
  Re: linux as a client :-( (Grant Edwards)
  Re: Do I need linux-2.2xx source in /usr/src? ("Tom Hoffmann")
  talkd logging (Toni)
  Re: Can't mount cdrom ("Peter T. Breuer")
  Re: Installing Linux from a ZIP drive? (The Procrastinator)
  Re: Porting from X86 to RISC Processor.. (The Procrastinator)
  Re: Need help: PS/2 mouse not working in doom (David Efflandt)
  Re: zoom modem (David Efflandt)
  Re: Do I need linux-2.2xx source in /usr/src? ([EMAIL PROTECTED])
  Re: Why can't I unmount....????? ([EMAIL PROTECTED])
  Re: Netscape & Video (Craig McCluskey)
  Re: The X Server... (Mike Hoegeman)
  Re: Do I need linux-2.2xx source in /usr/src? (J Bland)
  Re: Porting from X86 to RISC Processor.. (J Bland)
  Re: Sun Sparc faster then intel pentium: is this true???? (Carl J. Boll)
  Re: find files with pattern (Craig McCluskey)
  Re: Changing hwclock settings to local and not UTC (Craig McCluskey)
  Re: Group membership - any limit ? (brian moore)

----------------------------------------------------------------------------

From: Mark Wang <[EMAIL PROTECTED]>
Subject: Re: Internet backup services for Linux
Date: 24 Jun 2000 21:05:14 GMT

Robie Basak <[EMAIL PROTECTED]> wrote:

: Hmm. Assuming you just bought yourself a new hard drive, it's got to
: be at least > 10 Gb or so. In which case, what kind of internet speed
: do you have? If you've only just started considering it, have you made
: sure you've calculated the time it'll take to do a full backup/money
: it will cost?

True, however, the size of the data that I really need backed up (ie,
stuff that can't be re-installed or re-created easily) is small relative
to this -- around 500 MB - 1 GB  or so.  Plus, the way most services work
is that they just incrementally back up changed files, so after the first
full backup, the amount of data is much smalaler.

I'm on a Ethernet connection, and a fast connection to the Internet
backbone, so bandwidth shouldn't be an issue for me.

If anyone knows _any_ suggestions, please let me know.  Seems to me this
is a potentially large market.

-- Mark

-- 
Mark Wang                           "Darkness cannot drive out darkness; only
Stanford University '99              light can do that.  Hate cannot drive out
[EMAIL PROTECTED]        hate; only love can do that."
http://www.cs.stanford.edu/~mwang/              -- Martin Luther King, Jr.

------------------------------

From: [EMAIL PROTECTED] (Grant Edwards)
Subject: Re: linux as a client :-(
Date: Sat, 24 Jun 2000 21:10:40 GMT

In article <8j1dhv$64t$[EMAIL PROTECTED]>, [EMAIL PROTECTED] wrote:

>   I long for the day when all software
>   and drivers will be as simple as double-clicking on setup.exe

And then sitting around watching the damn machine reboot itself
four or five times as part of the installation process, crash
during one of the reboots, and not work when it's done.  I am
not exaggerating.  That is exactly what I sat through one
afternoon last week trying to install something under NT.

I can hardly wait for Linux to "catch up".

-- 
Grant Edwards                   grante             Yow!  Have my two-tone,
                                  at               1958 Nash METRO brought
                               visi.com            around...

------------------------------

From: "Tom Hoffmann" <[EMAIL PROTECTED]>
Subject: Re: Do I need linux-2.2xx source in /usr/src?
Date: Sat, 24 Jun 2000 21:17:19 GMT

In article <8j2qcg$[EMAIL PROTECTED]>, [EMAIL PROTECTED] (Andrew
Purugganan) wrote:
> Do I need to keep the linux-2.2.13mdk source in /usr/src? I see a
> bunch  of subdirectories, and some *.h files in there that are taking
> up space.  Is it common to keep it in expanded form for say, compiling
> purposes or  something? Like the includes?

If you ever plan/need to recompile your kernel you will need to have
the source code.

------------------------------

From: Toni <[EMAIL PROTECTED]>
Subject: talkd logging
Date: Sat, 24 Jun 2000 23:23:48 +0200

Hy, I'm a Linux newbie, but before asking I searched all available
sources concerning this after asking in this forum...

I'd like to log the sessions on my Linux machine, because sometimes I
like to re-read them...how can I do this ? 
Is there any talk-daemon that supports logging, or do I have
to change the code manually ? (This would be very bad, because I can't
program C yet)

Any help welcome,
Greetz,
Toni.

------------------------------

From: "Peter T. Breuer" <[EMAIL PROTECTED]>
Subject: Re: Can't mount cdrom
Date: 24 Jun 2000 21:26:19 GMT

Robert Hampf <[EMAIL PROTECTED]> wrote:
: Peter T. Breuer <[EMAIL PROTECTED]> helt pessu fram:
: : Robert Hampf <[EMAIL PROTECTED]> wrote:
: : : Have you tried not to use modules?  To compile everything into the
: : : kernel.  It makes things a lot simplier. 
: : No it doesn't. How can making your choices at compile time be simpler
: : than making your choices at run time?

: Because you don't need to mess with them (remeber to compile them,

You mean that in those cases where there is a choice of driver per
device node (major/minor), you can rely on the built in drivers to
probe to auto-select themselves, while you have to edit conf.modules
to choose the right driver if it's in a module?

Presumably you know which driver you want or you wouldn't be thinking
about whether best to compile it in or leave it as a module!

And if you don't know which driver you want, loading modules to try
them out is easier than recompiling the kernel and rebooting.

: have the right versions in the right locations and so on).  For

You have to remember to compile them. That's the same whether you
choose to compile them in the kernel or out. You don't have to do
anything else after installing them (and I include depmod in that).
In particular, you don't have to reboot. You don't have to install
a new kernel, etc. There is no opportunity for a wrong location ..
there is only one location for them.

I do know what you are thinking of, btw! You mean that if you 
recompile the kernel you may change the exported symbols in such a
way as to leave a previously compiled module short of a symbol
it needs.

: somebody who doesn't know what he is doing it's a good thing to have
: fewer things to remember.

I see. Why don't they just forget to recompile the kernel? Then things
would work just fine.

I think recompiling the kernel should be off limits until someone has
proven they can at least recompile a module!

Peter


------------------------------

From: The Procrastinator <[EMAIL PROTECTED]>
Subject: Re: Installing Linux from a ZIP drive?
Date: Sat, 24 Jun 2000 23:02:09 +0100

andrey <[EMAIL PROTECTED]> wrote:
> Greetings!

> I have an old 486, it doesn't have a CD-ROM or a network card. I have 
> RedHat 5.1 on CD and a parallel port ZIP drive. Can I install linux
> from the ZIP drive? If so, how?

Strictly not the answer to your question but offers you another way ...

Another way to install is to use the parallel port or serial port to network
the 486 to another computer with a CD ROM drive.

For the parallel port networking you need a `laplink cable'
Set up the network using plip (Parallel line interface protocol) as the 
interface.
etc.

Adrian

------------------------------

From: The Procrastinator <[EMAIL PROTECTED]>
Subject: Re: Porting from X86 to RISC Processor..
Date: Sat, 24 Jun 2000 23:04:06 +0100

shiv agarwal <[EMAIL PROTECTED]> wrote:

> Hi,

> Could some body help me finding the issues regarding porting  drivers
> from the Intel based (x86) processors to RISC based processors.
> Please help me with the documents or URLs or ....

Search for ArmLinux

Adrian

------------------------------

From: [EMAIL PROTECTED] (David Efflandt)
Subject: Re: Need help: PS/2 mouse not working in doom
Date: 24 Jun 2000 22:09:46 GMT
Reply-To: [EMAIL PROTECTED]

On Sat, 24 Jun 2000, [EMAIL PROTECTED]
<[EMAIL PROTECTED]> wrote:
>
>Just installed Doom and I can't get it to work with the mouse.
>
>It aborts with "mouse init failed" or some such.
>
>I have attempted to edit the ".doomrc" file with no success, perhaps
>because I don't know what to enter for the "mousedev" setting,
>although I have tried variations of "psaux", "ps aux", "PSAUX", etc.
>"mousetype" is set as PS/2.

I forget if Doom does this (or if I even used the mouse in Doom), but
Quake uses svgalib to control the mouse, and in some cases it helps to
'killall gpm' before running it.

The problem is that the default for svgalib is a serial mouse, so you
likely need to edit /etc/vga/libvga.config and change 'mouse Microsoft' to
'mouse PS2'.  If libvga.config is set to the wrong type of mouse for
Quake, you might not know it until you touch the mouse, at which time you 
go wildly out of control.

-- 
David Efflandt  [EMAIL PROTECTED]  http://www.de-srv.com/
http://www.autox.chicago.il.us/  http://www.berniesfloral.net/
http://hammer.prohosting.com/~cgi-wiz/  http://cgi-help.virtualave.net/


------------------------------

From: [EMAIL PROTECTED] (David Efflandt)
Crossposted-To: comp.os.linux.setup
Subject: Re: zoom modem
Date: 24 Jun 2000 22:43:29 GMT
Reply-To: [EMAIL PROTECTED]

On 23 Jun 2000 13:54:47 GMT, Kevin Croxen <[EMAIL PROTECTED]> wrote:
>
>The question is what IRQ the modem has landed on. The default
>IRQ for com1 and com3 is 4. If your /dev/ttyS2 requires a 
>different IRQ (likely, if you have something on com1),
>this value will need to be set using setserial. 
>Many Zoom modems allow choosing whether to set
>jumpers or rely on PnP. If yours is one of these,
>you will of course find life easier if you manually
>jumper the modem.

The problem is that Windows still uses what ever it pleases for port and
irq even if a PNP capable modem is hard jumpered.  I usually found that
the same settings Windows used would work in Linux with my Zoom 2819.  But
I eventually just jumpered it for PNP and used isapnp.conf to set it.

>If, however, your modem is a 
>basic nothing-to-be-done-about-it PnP, you may 
>wish to set your hardware bios for "non-PnP OS", so
>that the hardware is responsible for allocating your
>PnP peripherals. 

PNP in CMOS setup should be disabled anyway, but I am not sure if it has
anything to do with this.

What he would need is make sure that isapnptools is installed, do a
'pnpdump > /etc/isapnp.conf' and then edit the resulting file to uncomment
the port and irq used.  And if the irq is not a standard one for that
port, then add a setserial line in /etc/rc.d/rc.local to inform Linux what
it is.

>In article <[EMAIL PROTECTED]>, Paul Eisenberg wrote:
>>Hey, I got Corel Linux a few weeks ago and have been trying to get my
>>modem to work with it.  First I had a US Robotics/3Com Modem which I
>>assumed to be a WinModem with all the trouble it gave me.  Then I put
>>in a old Zoom Modem which Windows Found as a 
>>Zoom 56K Internal Fax Modem Modem 2812 PNP (COM 3)
>>
>>Using a Dialup setup in Corel Linux I picked 
>>/dev/ttyS2/ for COM 3
>>
>>It says 
>>Modem Ready
>>Then
>>Sorry, Modem is not responding
>>
>>What should I do at this point?  Sorry if this a simple question, but
>>hopefully someone will take the time.  I really am grateful.  Take
>>Care.  Paul


-- 
David Efflandt  [EMAIL PROTECTED]  http://www.de-srv.com/
http://www.autox.chicago.il.us/  http://www.berniesfloral.net/
http://hammer.prohosting.com/~cgi-wiz/  http://cgi-help.virtualave.net/


------------------------------

From: [EMAIL PROTECTED]
Subject: Re: Do I need linux-2.2xx source in /usr/src?
Date: Sat, 24 Jun 2000 23:56:54 +0100

Andrew Purugganan <[EMAIL PROTECTED]> did eloquently scribble:
> Do I need to keep the linux-2.2.13mdk source in /usr/src? I see a bunch 
> of subdirectories, and some *.h files in there that are taking up space. 
> Is it common to keep it in expanded form for say, compiling purposes or 
> something? Like the includes?

The only things you need in there are the files in /usr/src/linux/include,
and only then if you want to compile your own rather than relying on
RPMs/.debs...

-- 
______________________________________________________________________________
|   [EMAIL PROTECTED]   | "Are you pondering what I'm pondering Pinky?"   |
|Andrew Halliwell BSc(hons)|                                                 |
|            in            | "I think so brain, but this time, you control   |
|     Computer Science     |  the Encounter suit, and I'll do the voice..."  |
==============================================================================

------------------------------

From: [EMAIL PROTECTED]
Subject: Re: Why can't I unmount....?????
Crossposted-To: alt.os.linux,nf.comp.linux
Date: Sat, 24 Jun 2000 23:59:52 +0100

In comp.os.linux.misc Hendrix <[EMAIL PROTECTED]> wrote:
> Hi guys,

> Why is it that unmounting doesn't always work...???  Sometimes I try to
> unmount a specific filesystem and the system will return a message
> saying that the device is busy, and cannot be unmounted...  

It normally means there's something open in the partition you want to
umnount. For example, attempting to unmount /cdrom when you're in the /cdrom
directory tree will always fail. You have to cd out of there before you can
unmount it.

-- 
______________________________________________________________________________
|   [EMAIL PROTECTED]   |                                                 |
|Andrew Halliwell BSc(hons)| "The day Microsoft makes something that doesn't |
|            in            |  suck is probably the day they start making     |
|     Computer science     |  vacuum cleaners" - Ernst Jan Plugge            |
==============================================================================

------------------------------

From: Craig McCluskey <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux.suse
Subject: Re: Netscape & Video
Date: Sat, 24 Jun 2000 18:19:49 -0500

Thierry Cornelis wrote:

> What do you use to view .mpg? I have no problem with 'mtv' and 'plugger
> 3.0' as a plugin for netscape.

Where does one get these plug-ins?

Craig

------------------------------

From: Mike Hoegeman <[EMAIL PROTECTED]>
Crossposted-To: nf.comp.linux,alt.os.linux,comp.os.linux.x
Subject: Re: The X Server...
Date: Sat, 24 Jun 2000 23:26:02 GMT

The Darkener wrote:
> 
> <Snip>
> 
> >   h> So really, the computer running the X server could be a stand-alone
> >   h> computer without an x client at all....
> >
> > Yes.  In fact, early on there were a lot of hardware called X terminals,
> > which were very small systems with some flash or similar storage, a
> > dollop of memory, a big monitor, a keyboard, and a mouse, and that's it.
> >
> > There is a protocol these X terminals use to connect to other "big
> > systems" where your home directory, etc. lived, and the clients ran on
> > the "big system" and the X terminal had essentially nothing but the
> > X server process on it.
> >
> > Then everyone decided that was too slow and they wanted their own
> > hardware.  Now that networks and "big systems" are getting fast enough,
> > it looks like things might be moving back that way again.
> 
> That is one of my projects on my home LAN... I've got a cheapo computer (P-120) in
> my living room that I'm going to install a PC-2-TV card in, hook it up to my
> television + my lan (which the server resides in my bedroom), and then I'll have
> WebTV! (Well, sorta, a remote X client running a wm and netscape works though!)
> I'm also going to run remote speakers (which are hooked up to an amp which in turn
> is hooked up to the server's sound card) and then run xmms off of the remote
> client, pulling mp3 files from the server in my bedroom, and hearing/controlling
> them in the living room (and/or in my bedroom depending on where I want to flip the
> main/remote speaker knob at that particular moment).
> 
> Ahh, the beauty of Linux networking. =)



yeah! .. the linux guys invented all this stuff. 
kinda like al gore inventing the internet.. 

sigh..

------------------------------

From: [EMAIL PROTECTED] (J Bland)
Subject: Re: Do I need linux-2.2xx source in /usr/src?
Date: 24 Jun 2000 23:32:36 GMT

>> Do I need to keep the linux-2.2.13mdk source in /usr/src? I see a bunch 
>> of subdirectories, and some *.h files in there that are taking up space. 
>> Is it common to keep it in expanded form for say, compiling purposes or 
>> something? Like the includes?
>
>The only things you need in there are the files in /usr/src/linux/include,
>and only then if you want to compile your own rather than relying on
>RPMs/.debs...

If you installed the kernel includes from a package, keep them, you need
them to compile most (all?) C programs.

If you've rolled your own kernel and still have the source there, then still
keep those includes (you really want the same includes as the kernel you're
using). But to save space you can use: make clean, to get rid of all the
extraneous .o files etc.

In summary, clean out the crap but leave it, it's a pretty useful (and
sometimes essential) part of a linux system.

Frinky

------------------------------

From: [EMAIL PROTECTED] (J Bland)
Subject: Re: Porting from X86 to RISC Processor..
Date: 24 Jun 2000 23:39:12 GMT

>> Hi,
>
>> Could some body help me finding the issues regarding porting  drivers
>> from the Intel based (x86) processors to RISC based processors.
>> Please help me with the documents or URLs or ....
>
>Search for ArmLinux

To give you a starting advantage:

http://www.arm.uk.linux.org/

Although I would seriously look at other RISC ports before ARM Linux, unless
you're going to be using an ARM that is.

You could do worse than start with an architecture that's actually in the
main kernel tree...

Frinky

------------------------------

From: [EMAIL PROTECTED] (Carl J. Boll)
Crossposted-To: comp.sys.sun.hardware,comp.sys.sun.misc
Subject: Re: Sun Sparc faster then intel pentium: is this true????
Date: 24 Jun 2000 14:30:05 -0500

On a foggy, gloomy day, Dave Schanen's fingers typed:
> 
> > When it comes to pure processing speed the current winner in
> > this debate is Intel, no doubt about it.
> There is a doubt, or it wouldn't be debated at all.  If all you define speed
> as is clock frequency without regard for architecture, then intel systems
> are faster, but it's obvious a sun will outperform an IA32 machine clocked
> higher, even on the many overhyped benchmarks.

The proof please?  I've seen the benchmarks.  In a processor
battle with >NO< other considerations the Intel processor >WILL<
win.  If you had read my entire post before replying you would
understand that I am not pro-Intel nor am I pro-Sun.  In a
small business Suns are just not affordable.

> >
> > However, when looking at a system to serve files, process
> > data and manage multiple tasks processing speed isn't the
> > only criteria.
> >
> > Sun builds expensive, reliable well designed (for the most
> > part) servers and clients that can be installed easily and
> > do what they are designed to do.  Solaris (and SunOS) has
> > gone through many revisions and even without any add on
> > suites it is remarkably reliable.
> >
> > This isn't the case with Intel machines.  First you have
> > to decide either on a motherboard (and chipset) or a PC
> > manufacturer (like Dell, Compaq or H. P.) and then go from
> > there.  If you make a poor choice in motherboards the CPU's
> > speed won't count for much.
> >
> > I personally have a Tyan Tomcat IV with dual P-233 MMX CPU's
> > for my own personal server.  I also have several older
> > Suns on my personal network as well.  They are very good
> > machines and extremely reliable.  My main server up until
> > 4 years ago was a Sun 3/160.  I still use it as a Point of
> > Sale server at antique computer shows with 16 serial terminals
> > and at least as many clients on the ethernet.
> >
> > If I am installing a network in a small business for P. O. S.
> > (point of sale) I look at exactly what is required.  One of
> > the biggest factors is going to be cost and reliability.  For
> > such an application I generally suggest that an Intel based
> > server be used with Windows 9X as an OS and Intel based
> > computers be used for terminals.  In a small application
> > like this where only employees have access, there are no
> > external data lines and the most intensive access is still
> > small (less than 5 terminals) there is no need for stringent
> > security, high bandwidth, etc.  Cost is the main focus in a
> > very small business application as well as maintainability.
> > They can't afford a system administrator so it becomes a very
> > important factor that there be only one OS and that it be
> > easy to use.  I haven't found NT to be as reliable in these
> > cases (calls back to installations for troubleshooting).
> >
> > In a medium sized install where money is tight I am now
> > looking at Intel machines (Dell, HP and Compaq) for servers
> > running LINUX and VMWare and running Windows on the server
> > (NT or 9X) especially if it is a WAN or is used to access
> > the internet.  LINUX can be used as a firewall in this case
> > and requires only minimal maintainance while providing very
> > good security.
> It's funny that anything better than NT provides 'good' security these days.
> If you want security openbsd or solaris has a considerably better track
> record than Linux, and both run on intel hardware.
> >

NT can provide decent security too but it isn't well documented.
This probably so that you have to get certified, a real pain in
the butt and just Microsoft's mentality.  Out of the box NetBSD,
Solaris, AIX, SCO, LINUX and even NeXtstep provide better
security than NT.  

> > At this point I should mention that P. O. S. software runs
> > under MS-DOS.
> The machine I operate at a factory this summer runs it's program on MS-DOS,
> we have to throw the main switch every time we set up for a new set of parts
> for no particular reason...

Sure there is a reason.  The software is poorly designed or the
hardware is poorly designed.  This isn't the fault of the OS.
MS-DOS provides plenty of hooks for monitoring IRQ's and there
is no reason you need to reset the power >IF< the software is
written properly.  It's obvious that the software is either
unable to properly reset the hardware without a powerdown or
the hardware wasn't designed to change configurations without
a powerdown.  One of three things is happening here, either the
hardware needs a powerdown, this being not computer related or
the interface doesn't provide a way to reset without the computer
being reset.  The third thing it may be is that the program is
not written properly.  It should be able to reset the computer
but that may have been overlooked.  No matter what, it isn't
a problem with the processor or the OS.  The same program ported
to Solaris on a Sun would have the same problems.

> >
> > In large installs I recommend that a Sun be used, especially
> > if it is going to be an internet based business.  Suns are
> > without a doubt reliable.  This is a case where money is
> > usually secondary to reliability and security is a must.
> >
> > So what is my point?  Suns have a place, they are a tool
> > just like many other computers are tools.  The same is
> > true of an operating system.  No one OS or computer is the
> > answer to all installations or problems.  A good consultant
> > understands that and makes decisions decided on the needs
> > of his customer not on his preferences.  He tries to give
> > his client the proper tools to do the job.  I love Suns
> > but they aren't affordable for a lot of small applications
> > and they just can't be recommended because I like them.
> The "suns are good for internet servers only" mentality is silly.  For any
> sort of scientific data analysis suns are the only way to go, hence their
> brisk workstation sales.  A friend of mine does uses an mri scanning
> technology for part of a academic cancer research project, and the analysis
> of the data is done by, guess what, an ultra 60.  The pci rev 2.1 slots
> which all the ultra's but the 5's and 10's use are 66 mhz, I'm not aware of
> any stock desktop pc systems which support this speed outside of the AGP
> slot. Sun workstations are cheap as far as workstations go as well, take a
> look at ibm's rs6000 line sometime.  All the of decent scientific software
> is only available for Unix systems as well, try and find a decent fortran
> implementation for intel and NT.  As far as sun systems being too expensive,
> these are engineering/scientifc machines and servers, they're not meant for
> use for playing games and doing spreadsheets.  If you wanted a cost
> effective solution for a WAN, the sunray systems are available, and it's
> many times more efficent to have a half dozen sunrays sharing a servers
> files (and more secure) than to have 12 dells which spend 99 percent of
> their 600 mhz pIII clock cycles idle anyway.   Suns are a better long term
> solution as well, since solaris still supports it's legacy systems with each
> new release, unlike NT 5 which requires, what? A 266 with 96 megs of ram to
> play solitaire and run word?  Linux is nice, and would certainly make intel
> machines shine but is in need of some serious standardization and a little
> less hype and more 3rd party support.
> 

But I never said they were good for internet servers only, now did I?
As to Suns being engineering/scientific platforms, what a load of B. S.
They are computers designed to handle a tremendous amount of network
traffic and throughput.  This makes them good engineering and good
scientific platforms but they are designed for long amounts of uptime,
reliability and excellent throughput.  

As to the mri scanning, I have a friend who is in the field.  Many
mri machines are Intel based running OS-9 (a real time OS by Microware).

I am familiar with the RS6000 too as well as AIX.  These machines are
overpriced and AIX is a pain in the butt, just like Solaris, just
non-UNIX-standard enough to make the learning curve high.  As far as
the hardware goes though it is an excellent machine, just not very
cost effective when compared to a Sun.

Now, how would you explain your Sunray theory?  6 machines on a server
as opposed to 12 Dells?  If I only needed 6 machines why would I buy
12 Dells?  Seems like we are talking about apples and oranges here.

Finally, Solaris supports their legacy systems?  Since when?  I have a
3/160 that is locked in at SunOS 4.1.1U, a 4/200 that is locked in at
SunOS 4.3 (I believe), an IPX, an IPC, a Sparc 1+ that are locked in at
Solaris 7 since Sun dropped support for these platforms in Solaris 8.
Get your facts straight.

It makes no sense to support outdated platforms beyond a reasonable
period and Sun does that just as NT does.  Each version of SunOS has
required more hard disk space and more memory and has run a little
slower.  This is due to more processes, more features and probably
a little to poor programming.  If you don't believe me just boot up
a Sparc 1+ with SunOS 2.4 and then boot the same machine with the
latest version of Solaris that supports it, Solaris 7.  Try doing
something now.  How about copying say 20 Megs from one drive to
another.  Want to guess which version of the OS will do it faster?

No, Sun is a good machine, Solaris is a good OS but as I stated they
aren't for every application.  In this world cost is always a real
factor.

> Dave
> 
> 
> 
> 

Carl
-- 
Carl Boll:  [EMAIL PROTECTED]

-- 
Carl Boll:  [EMAIL PROTECTED]

------------------------------

From: Craig McCluskey <[EMAIL PROTECTED]>
Subject: Re: find files with pattern
Date: Sat, 24 Jun 2000 18:45:19 -0500

Dances With Crows wrote:
> Almost exactly the same in Linux.
> 
> find . -name \*.h -exec grep -H "FNM_FL" {} \;
> 
> The \; at the end signifies the end of the exec statement.  The * is
> escaped so the shell doesn't try to expand it.  HTH,

Or you can do it without the exec (a way I find more intuitive):

grep -H FNM_FL `find . -name *.h -print`

(Bash doesn't expand the *.)

Craig

------------------------------

From: Craig McCluskey <[EMAIL PROTECTED]>
Subject: Re: Changing hwclock settings to local and not UTC
Date: Sat, 24 Jun 2000 18:47:22 -0500

Kerry Cox wrote:
> 
> For some reason a machine that I now administer keeps wanting to change
> the clock settings to UTC time, thus placing it 6 hours ahead of the
> local time and the hardware clock.

Have you re-booted your system and checked the bios setting?

Craig

------------------------------

From: [EMAIL PROTECTED] (brian moore)
Subject: Re: Group membership - any limit ?
Date: 24 Jun 2000 23:59:08 GMT

On Sat, 24 Jun 2000 01:00:48 +0100, 
 Colin Smith <[EMAIL PROTECTED]> wrote:
> On Fri, 23 Jun 2000 23:43:14 +0200, Bertrand Renuart <[EMAIL PROTECTED]> 
>wrote:
> >> "R> Hi there,
> >> "R>
> >> "R> I was wondering if there is a limit to the number of groups a user
> >account
> >> "R> can belong to ?
> >>
> >> Not under Linux.  SunOS whines if you are in more than 8 groups and
> >> Solaris bitches and moans if you are in more than 16.
> 
> Yes, there is.
> 
> >>
> >Great...
> >So, according to you, what could be the cause of the problem I described
> >below ?
> 
> Linux allows 32 groups. You'll have to edit a couple of header files and rebuild
> a whole lot of software to increase it.
> 
> You could try NIS netgroups instead.

Or more likely "rethink what you're trying to do".

-- 
Brian Moore                       | Of course vi is God's editor.
      Sysadmin, C/Perl Hacker     | If He used Emacs, He'd still be waiting
      Usenet Vandal               |  for it to load on the seventh day.
      Netscum, Bane of Elves.

------------------------------


** 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.misc) 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-Misc Digest
******************************

Reply via email to