Printing from Solaris
perttu
I have a printer connected to parallel port (at least I think it is
parallel port). Ultimately I am trying to send print jobs to printer but
I am not able to get it work.
I have become to conclusion that the problem is with parallel port since
nothing happens when I do for example
ls /etc > /dev/lp0
or
ls /etc > /dev/lp1
The machine is Ultra 10
with SunOS 5.8
What am I missing.
Hopefully somebody can help.
/Perttu
Michael Tosch
I hope you will rather use admintool to install a local printer that
prints on /dev/lp0.
Most applications will print via lp or lpr.
Nevertheless, directly using /dev/lp0 should work. Check that you
really have a parallel port (// sign), and not a serial port
(named A or B).
--
Michael Tosch
IT Specialist
HP Managed Services Germany
Phone +49 2407 575 313
Mail: michae...@hp.com
Michael Laajanen
replace Solaris print system with CUPS or ESP, you will understand why
if you try both and compare againts Solaris printer system for easy of
use if you are no expert :))
http://www.easysw.com/printpro/
/michael
ML Starkey
wrote:
You are missing the proper name for the parallel port device on an
Ultra 10.
Try:
/dev/ecpp0
Richard L. Hamilton
You can do almost anything with the Solaris printing system that you can
with a replacement, although I won't pretend that it will be easy.
If you're not an expert, or at least willing to become one, you shouldn't
be a Solaris admin!
(that goes for other systems too; the MCSE solution of reboot, and if
that doesn't fix it, reload, is friggin' ignorant - you never learn
what the real problems are)
perttu
for my printer it started working.
Tony Walton
> Thanks a lot. This was the problem. As soon as I changed the device name
> for my printer it started working.
Now ls -l /dev/lp* and remove the ones that are files instead of devices...
--
Tony
Michael Laajanen
Richard L. Hamilton wrote:
> In article <z78Zb.83360$dP1.2...@newsc.telia.net>,
> Michael Laajanen <michael.laajanen.no-spam.@telia.com> writes:
>
>>HI,
>>
>
>
> You can do almost anything with the Solaris printing system that you can
> with a replacement, although I won't pretend that it will be easy.\
I know that.
>
> If you're not an expert, or at least willing to become one, you shouldn't
> be a Solaris admin!
That could be very much true, but in the world there are people using
Solaris that have no "professional admin", and they use a handy guy for
doing things and he has other things to do than learnig how to setup the
printersystem.
I for instance do HW design, infact I just did a HW design CONTROLLING
the blody printer head :)) and I did not use lpadmin nor ESP, I just
VHDL :))
Hmm, thats me by the way :))
>
> (that goes for other systems too; the MCSE solution of reboot, and if
> that doesn't fix it, reload, is friggin' ignorant - you never learn
> what the real problems are)
>
Yes, but this is not a B/W world, it has grayscale I never reboot a
machine unless really pain and that I can't rememeber when it was :)
No but to be serious, Some taskt can be done easier with some penalty,
patchproblems could be one, speed flexibility and so but you have to
draw a line just to get it working thats why many startups use Exchange
for mail,it has it's wizards and UNIX don't sadly.
/michael
admin on spare time
UNIX admin
Newsbeitrag news:z78Zb.83360$dP1.2...@newsc.telia.net...
> replace Solaris print system with CUPS or ESP, you will understand why
> if you try both and compare againts Solaris printer system for easy of
> use if you are no expert :))
One SHOULD NOT replace the native printing and spooling subsystems with
CUPS. Actually CUPS should be avoided at all costs and used only when all
other solutions have been exhausted. CUPS requires major tampering with OS,
including removal of the printing and spooling subsystems. Therefore, it
should be avoided like the plague.
Michael Laajanen
You must be one of the fellows that stick to something(vendorish) until
the vendor changes it's mind!
I can't stop thinking about ipfilter vs Sunscreen :))
Don't be folish, it's about making life easy if it's Sun orginal or not
who cares if it make the admins life easy!
If have not run +1000 people server group much smaller 50 people, but
those who have would not ask here, right!
/michael
Richard L. Hamilton
Glad to hear that.
Maybe it's just me, but almost always I'd rather figure out how to
configure (and perhaps even supplement) one component than to rip it out
and replace it entirely. Perhaps it's because the former is merely a
technical problem, but the latter is both a technical _and_ a people
problem (dealing with the multiple suppliers and support channels), and
people problems have a seriously adverse impact on my silicon-based
metabolism. :-)
> No but to be serious, Some taskt can be done easier with some penalty,
> patchproblems could be one, speed flexibility and so but you have to
> draw a line just to get it working thats why many startups use Exchange
> for mail,it has it's wizards and UNIX don't sadly.
>
Well, you understand about tradeoffs. Some of these folks that say
"just replace the [whatever] subsystem with WhizBang 6.0 and all your
[whatever] problems will go away" either don't (and are ignorant and
leading others to follow in their ignorance) or are being purposely
misleading. I guess you're neither of those, and just forgot to throw
in a TANSTAAFL* disclaimer.
*There Ain't No Such Thing As A Free Lunch (from Heinlein, I think);
that is, everything has costs, even if you don't see them or don't
_think_ you're paying for them.
>
> /michael
> admin on spare time
Michael Laajanen
Richard L. Hamilton wrote:
> In article <7VsZb.83453$dP1.2...@newsc.telia.net>,
> Michael Laajanen <michael.laajanen.no-spam.@telia.com> writes:
>
>>HI,
> and replace it entirely. Perhaps it's because the former is merely a
> technical problem, but the latter is both a technical _and_ a people
> problem (dealing with the multiple suppliers and support channels), and
> people problems have a seriously adverse impact on my silicon-based
> metabolism. :-)
>
You are absolutely right, especially when it comes to support.
It is a bit different some of these issues because alot of the people
here that are "real admins", and they can take the time to understand
things deeper and make a better end solution, somebody like me sometimes
have to make a shortcut and hopefully get back later when there is time
and redo thing properly.
>>
>
>
> Well, you understand about tradeoffs. Some of these folks that say
> "just replace the [whatever] subsystem with WhizBang 6.0 and all your
> [whatever] problems will go away" either don't (and are ignorant and
> leading others to follow in their ignorance) or are being purposely
> misleading. I guess you're neither of those, and just forgot to throw
> in a TANSTAAFL* disclaimer.
Sorry I am a non native english speaking guy, what do you mean with
TANSTAAFL* is that a common expression just or?
>
> *There Ain't No Such Thing As A Free Lunch (from Heinlein, I think);
> that is, everything has costs, even if you don't see them or don't
> _think_ you're paying for them.
>
Yes you pay sooner or later, lucky we have Yahoo as a backup for
Exchange users :))
/michael
Richard L. Hamilton
[...]
>> Well, you understand about tradeoffs. Some of these folks that say
>> "just replace the [whatever] subsystem with WhizBang 6.0 and all your
>> [whatever] problems will go away" either don't (and are ignorant and
>> leading others to follow in their ignorance) or are being purposely
>> misleading. I guess you're neither of those, and just forgot to throw
>> in a TANSTAAFL* disclaimer.
> Sorry I am a non native english speaking guy, what do you mean with
> TANSTAAFL* is that a common expression just or?
>
>>
>> *There Ain't No Such Thing As A Free Lunch (from Heinlein, I think);
http://jargon.net/jargonfile/t/TANSTAAFL.html (first hit in Google :-)
>> that is, everything has costs, even if you don't see them or don't
>> _think_ you're paying for them.
>>
>
> Yes you pay sooner or later, lucky we have Yahoo as a backup for
> Exchange users :))
>
> /michael
Greg Andrews
>
>Sorry I am a non native english speaking guy, what do you mean with
>TANSTAAFL* is that a common expression just or?
>
It stands for "There Ain't No Such Thing As A Free Lunch", meaning
one does not receive anything for free. The TANSTAAFL abbreviation
was made popular by the author Robert Heinlein in some of his books.
-Greg
--
Do NOT reply via e-mail.
Reply in the newsgroup.
Another Tom
Both Solaris printing and CUPS are tools to accomplish a job. I cannot
blame someone for choosing an easier tool to use. I don't see too many
mechanical engineers with slide rules in their pockets.
Sorry, semi-rant. SunOS printing soured my taste for host based printing solutions.
>
> (that goes for other systems too; the MCSE solution of reboot, and if
> that doesn't fix it, reload, is friggin' ignorant - you never learn
> what the real problems are)
With Windows you really cant.
UNIX admin
> You must be one of the fellows that stick to something(vendorish) until
> the vendor changes it's mind!
Absolutely. Except that vendors rarely change their minds, and Solaris is
particularly good about this. I don't have to worry about the next release,
because I know that what worked in Solaris9 will work as expected in
Solaris10. In an unlikely event that something has changed, a manpage will
explain what, why, and how to deal with it. So, I prefer a corporate UNIX
any day to the non-corporate one. And as an added bonus, a command in
Solaris exists on IRIX and HP-UX with 90% probability and operates
identically in 90% of the cases. As a professional, I have a true and deep
appreciation for that.
> I can't stop thinking about ipfilter vs Sunscreen :))
I actually use ipfilter and have been for a few years now. At any rate,
ipfilter will be included in Solaris10 by default. I've never touched
SunScreen; there was simply no need.
> Don't be folish, it's about making life easy if it's Sun orginal or not
> who cares if it make the admins life easy!
Because I'm NOT going to compromise system consistency and coherence for
some nitwit's comfort. One of the qualities of finest system administrators
is CONSISTENCY! There are other users AND uses I have to consider as well.
Anyway, if *I* am the admin, then I *know* how to set up printing using
Solaris's printing and spooling subsystems, and therefore, CUPS is as
useless to me as a pile of garbage.
In reality, I actually have a print server, and use lp/lpr/Samba's spoolers
simultaneously without any problems. But I wasn't a twit, because when I
went to buy my color laser printer, I picked one that supported PostScript
in hardware. It pays to ponder your steps before acting upon them, and not
just spew buzzwords like "CUPS" for solving a problem that never really
existed in the first place.
> If have not run +1000 people server group much smaller 50 people, but
> those who have would not ask here, right!
What...? This is a completely incoherent sentence.
Judging by your reactions, I can't help but be left with the impression that
you don't truly understand CLEAN system administration. Especially if you
have to revert to CUPS to solve your printing "problems".
UNIX admin
news:pan.2004.02.22...@snet.net...
> Sorry, semi-rant. SunOS printing soured my taste for host based printing
solutions.
That's a shame, because Solaris is one of the better print server platforms.
Michael Laajanen
UNIX admin wrote:
> "Michael Laajanen" <michael.laajanen.no-spam.@telia.com> schrieb im
> Newsbeitrag news:i8wZb.49730$mU6.1...@newsb.telia.net...
>
>
>>You must be one of the fellows that stick to something(vendorish) until
>>the vendor changes it's mind!
>
>
> Absolutely. Except that vendors rarely change their minds,
> and Solaris is
> particularly good about this. I don't have to worry about the next release,
> because I know that what worked in Solaris9 will work as expected in
> Solaris10. In an unlikely event that something has changed, a manpage will
> explain what, why, and how to deal with it. So, I prefer a corporate UNIX
> any day to the non-corporate one. And as an added bonus, a command in
> Solaris exists on IRIX and HP-UX with 90% probability and operates
> identically in 90% of the cases. As a professional, I have a true and deep
> appreciation for that.
Me thats whats so nice with Unix, the same for Linux.
>
>
>>I can't stop thinking about ipfilter vs Sunscreen :))
>
>
> I actually use ipfilter and have been for a few years now. At any rate,
> ipfilter will be included in Solaris10 by default. I've never touched
> SunScreen; there was simply no need.
>
true, but I know whats been said here regarding not using Sunscreen and
using ipfilter :))
>
>>Don't be folish, it's about making life easy if it's Sun orginal or not
>>who cares if it make the admins life easy!
>
>
> Because I'm NOT going to compromise system consistency and coherence for
> some nitwit's comfort. One of the qualities of finest system administrators
> is CONSISTENCY! There are other users AND uses I have to consider as well.
> Anyway, if *I* am the admin, then I *know* how to set up printing using
> Solaris's printing and spooling subsystems, and therefore, CUPS is as
> useless to me as a pile of garbage.
BUT, what is CUPS is default on the clients like a bunch of Linux
systems, then where is the CONSISTENCY for the users?
>
> In reality, I actually have a print server, and use lp/lpr/Samba's spoolers
> simultaneously without any problems. But I wasn't a twit, because when I
> went to buy my color laser printer, I picked one that supported PostScript
> in hardware. It pays to ponder your steps before acting upon them, and not
> just spew buzzwords like "CUPS" for solving a problem that never really
> existed in the first place.
never seen a printer that does PS in hardware, what printer is that?
>
>
>>If have not run +1000 people server group much smaller 50 people, but
>>those who have would not ask here, right!
>
>
> What...? This is a completely incoherent sentence.
I guess that you mean that I missed a few words, I mean that those of
run systems with 1000's of machines would not ask printer question here,
right?
>
> Judging by your reactions, I can't help but be left with the impression that
> you don't truly understand CLEAN system administration. Especially if you
> have to revert to CUPS to solve your printing "problems".
>
>
Nope, that is maybe true, but I am a non-professional admin, when not
doing my job as a HW eng, and then I try to make clean designs just like
you do, But what clean designs would you do in HW, would you not like to
cut corners then and by a assembled printer instead of assebling it
yourself?
In this matter Solaris is like IKEA,it constains everything (almost) but
you have to assembly it yourself :)
CUPS contains drivers/descriptions for the printers, what would you do
without them in Solaris!
/michael
Richard L. Hamilton
[...]
>> In reality, I actually have a print server, and use lp/lpr/Samba's spoolers
>> simultaneously without any problems. But I wasn't a twit, because when I
>> went to buy my color laser printer, I picked one that supported PostScript
>> in hardware. It pays to ponder your steps before acting upon them, and not
>> just spew buzzwords like "CUPS" for solving a problem that never really
>> existed in the first place.
> never seen a printer that does PS in hardware, what printer is that?
s/in hardware/without a host-based postscript interpreter/
although even Sun used to have something that did the latter (the
SPARCprinters). That particular point has little to do with replacing
the entire printing system anyway, as recent Solaris 9 includes gs,
so as long as one of the compiled-in drivers will handle the printer,
it doesn't take much more than some simple scripts to tie it all together.
[...]
>> Judging by your reactions, I can't help but be left with the impression that
>> you don't truly understand CLEAN system administration. Especially if you
>> have to revert to CUPS to solve your printing "problems".
>>
>>
> Nope, that is maybe true, but I am a non-professional admin, when not
> doing my job as a HW eng, and then I try to make clean designs just like
> you do, But what clean designs would you do in HW, would you not like to
> cut corners then and by a assembled printer instead of assebling it
> yourself?
>
> In this matter Solaris is like IKEA,it constains everything (almost) but
> you have to assembly it yourself :)
And that's a problem...how?
> CUPS contains drivers/descriptions for the printers, what would you do
> without them in Solaris!
Well, I for one have used SysV (with gs, when needed to drive non-ps
printers) since long before there was a CUPS or LprNG or any of that
(about 1993 on a 3b1 running its own variant of SVR2 (and later, on a
2nd-hand IPC running Solaris 2.4), printing to an Epson FX-80, and later,
a Canon BJ10e; yes, it was slow, but it worked!)
But to be fair, I was a programmer long before I was an admin, so that
seemed obvious enough at the time.
UNIX admin
You mean 'stubborn', as in 'stubborn as a dragon'?
> Me thats whats so nice with Unix, the same for Linux.
Linux *is* UNIX. It may be a low performing, home-brewed PC-bucket poor
man's OS, but that doesn't make it any less UNIX.
> true, but I know whats been said here regarding not using Sunscreen and
> using ipfilter :))
I'm not aware of what's been said about SunScreen versus ipfilter, nor do I
have a vested interest in it, especially after Sun announced SunScreen is
being discontinued.
If you are alluding that I'm against OpenSource or freeware, all I have to
write is that I use freeware/OpenSource extensively, as I've never been a
fan of paid software... although, there are a select few products that cost
a lot of money (20K Euro or more), that I would gladly buy if I ran my own
company. But otherwise, I run freeware on my corporate UNIXes at home,
although I try to run as much commercial software as I can get a hold of,
for the tasks I need.
Even I can readily acknowledge the benefits of running OpenOffice versus
StarOffice.
> BUT, what is CUPS is default on the clients like a bunch of Linux
> systems, then where is the CONSISTENCY for the users?
As far as I know, lprng is the default on most Linux clients/servers, unless
that changed recently. But here, you fail to recognize one important point:
in the event that CUPS indeed is the default spooling system on Linux, then
you wouldn't go replacing it with something else, but rather learn how it
works and how to integrate it with your print servers, or clients, depending
in which capacity you use that Linux machine.
Now, having said that, you'd only tamper with CUPS if there was something
seriously wrong with it or if it was lacking some very much needed
functionality. Otherwise, you want to leave everything as-is, because if
you have other Linux clients on your network, you want everything to be
CONSISTENT, or else the whole endeavor would quickly turn into an
administrative nightmare for both users and you, the administrator.
You have to consider what happens when something breaks, your users expect
that they have the needed functionality at their disposal, and now you would
have to fiddle with a non-standard printing system. It means downtime for
your users, and late work hours for you. And while I don't mind working
late hours when there is a need for it, I do mind working late hours BECAUSE
I WAS SLOPPY and "just wanted to get something working quick". That is too
high of a price to pay... it's the price of shortsightedness, and certainly
not artful or thoughtful system administration.
There is art in every discipline, and so too is that the case in system
administration. It can be a fine craft if the system administrator is a
master.
> I guess that you mean that I missed a few words, I mean that those of
> run systems with 1000's of machines would not ask printer question here,
> right?
Well, who knows? comp.unix.solaris and alt.solaris.x86 are chalk full of
people who are in system administration positions, and they're way over
their heads. They're drowning, not knowing what to do or where to turn to;
it's simply too much for them, yet they fear losing their jobs.
On the other hand, there are some highly skilled experts here as well; so in
the end, there is balance.
My point is, it doesn't matter if you run 50 or 1000 servers; if you do it
poorly, it'll be just as same nightmare. Even running 50 servers is a
daunting task, even in the event that you are the only user!
> Nope, that is maybe true, but I am a non-professional admin, when not
> doing my job as a HW eng, and then I try to make clean designs just like
> you do, But what clean designs would you do in HW, would you not like to
> cut corners then and by a assembled printer instead of assebling it
> yourself?
Well, that depends. You asking the wrong person. I'm a strictly hands-on
guy and a control freak when it comes to issues like these. If I knew how
to make my own printer, trust that I would rather do that than go buy one.
Example: I run my own firewall on a UNIX box, where I know every detail of
what's been done to it, rather than just walking to a nearest computer store
and buying a premade firewall appliance.
> In this matter Solaris is like IKEA,it constains everything (almost) but
> you have to assembly it yourself :)
Exactly the opposite! Solaris at least has a lot of built in support for
various things, and in my experience, after IRIX, it's the least difficult
to install and configure. In fact, you can have a fully operational Solaris
server in a matter of hours, without any extra work.
In this particular case, it takes about 20 minutes, maybe less, for the
printer to start spewing pages from clients using Solaris's native spoolers.
I bet you don't feel so hyped up about CUPS now?
> CUPS contains drivers/descriptions for the printers, what would you do
> without them in Solaris!
Use GhostScript, or even better, BUY THE RIGHT PRINTER right off the bat.
In time you will learn, that the most productive way is to buy your hardware
for your software, not the other way around. That is why hardware
compatibility lists exist; use them. You'll save yourself a lot of work and
nerves.
Richard L. Hamilton
"UNIX admin" <tripi...@hotmail.com> writes:
>> Hmm that is also sometimes called a stubbard way of thinking!
>
> You mean 'stubborn', as in 'stubborn as a dragon'?
>
>> Me thats whats so nice with Unix, the same for Linux.
>
>
> Linux *is* UNIX. It may be a low performing, home-brewed PC-bucket poor
> man's OS, but that doesn't make it any less UNIX.
It may be close and getting closer, but ask the folks at The Open Group
whether they think it's Unix.
Unix nowadays is no longer a code base, it's a trademark, and in addition
to license fees, one of the conditions of use is compliance with the
Single Unix Specification. Unless something big happened and I missed it,
Linux doesn't comply. One could perhaps create a distribution that
tried to get a good deal closer, although I think that some of the issues
are with the Linux kernel itself, not just with all the ancillary (mostly
GNU) commands and utilities.
[...]
UNIX admin
Newsbeitrag news:103l36d...@corp.supernews.com...
> > Linux *is* UNIX. It may be a low performing, home-brewed PC-bucket poor
> > man's OS, but that doesn't make it any less UNIX.
>
> It may be close and getting closer, but ask the folks at The Open Group
> whether they think it's Unix.
>
> Unix nowadays is no longer a code base, it's a trademark, and in addition
> to license fees, one of the conditions of use is compliance with the
> Single Unix Specification. Unless something big happened and I missed it,
> Linux doesn't comply. One could perhaps create a distribution that
> tried to get a good deal closer, although I think that some of the issues
> are with the Linux kernel itself, not just with all the ancillary (mostly
> GNU) commands and utilities.
Eh, I never wrote that it's POSIX compliant or that it follows standards (in
actuallity, it doesn't). It can't decide whether it wants to be BSD or
SVR4, although it's becoming more and more like the latter.
But if it walks like a duck, swims like a duck, quacks like a duck... then
it must be a duck.
Barry Margolin
Richard.L...@mindwarp.smart.net (Richard L. Hamilton) wrote:
> Unix nowadays is no longer a code base, it's a trademark, and in addition
> to license fees, one of the conditions of use is compliance with the
> Single Unix Specification.
Since it hasn't passed the conformance test, it's true that the Linux
vendors can't legally call their products Unix. But trademark law has
no bearing on what the rest of us say in informal communications. We
can say it's Unix, we can say it's Coca Cola, we can say whatever we
want. For practical purposes, it's as good as any "certified" Unix.
And if laypeople use Unix in a generic sense enough, perhaps it will
become the generic term that it should have been (as happened with
trademarks like Aspirin).
--
Barry Margolin, bar...@alum.mit.edu
Arlington, MA
*** PLEASE post questions in newsgroups, not directly to me ***
Thomas Dehn
Only if you compare the chicken (Linux) to the cockroaches.
Thomas
Michael Laajanen
Barry Margolin wrote:
> In article <103l36d...@corp.supernews.com>,
> Richard.L...@mindwarp.smart.net (Richard L. Hamilton) wrote:
>
>
>>Unix nowadays is no longer a code base, it's a trademark, and in addition
>>to license fees, one of the conditions of use is compliance with the
>>Single Unix Specification.
>
>
> Since it hasn't passed the conformance test, it's true that the Linux
> vendors can't legally call their products Unix. But trademark law has
> no bearing on what the rest of us say in informal communications. We
> can say it's Unix, we can say it's Coca Cola, we can say whatever we
> want. For practical purposes, it's as good as any "certified" Unix.
difference unless you go deaper but if you do you also apreciate and
understand things from both sides.
>
> And if laypeople use Unix in a generic sense enough, perhaps it will
> become the generic term that it should have been (as happened with
> trademarks like Aspirin).
Exactly.
/michael
Michael Laajanen
Richard L. Hamilton wrote:
> Michael Laajanen <michael.laajanen.no-spam.@telia.com> writes:
>
>>HI,
>>UNIX admin wrote:
>
> [...]
>
>>>In reality, I actually have a print server, and use lp/lpr/Samba's spoolers
>>>simultaneously without any problems. But I wasn't a twit, because when I
>>>went to buy my color laser printer, I picked one that supported PostScript
>>>in hardware. It pays to ponder your steps before acting upon them, and not
>>>just spew buzzwords like "CUPS" for solving a problem that never really
>>>existed in the first place.
>>
>>never seen a printer that does PS in hardware, what printer is that?
>
>
> s/in hardware/without a host-based postscript interpreter/
/michael
UNIX admin
news:403b1...@news.arcor-ip.de...
> > But if it walks like a duck, swims like a duck, quacks like a duck...
then
> > it must be a duck.
>
> Only if you compare the chicken (Linux) to the cockroaches.
Do pardon me, but I'm a professional admin, not a zealot. If there is
/dev/, if there is /proc, /var, /usr, if there is `ls`, `ps`, `fsck`,
`xterm`, and so on, then to me that means I'm on a UNIX system. I don't
care if it passed any conformance tests, but if it behaves like UNIX, looks
like UNIX and feels like UNIX, then it *IS* UNIX.
Again, I'm not an OpenSource and/or Linux/*BSD crusader, but a professional.
Petty nitpicking and OSS ZEALOTRY is of no interest to me.
* * * * *
I once read somewhere, that "Linux" is a contraction of "Linus" and "Minix",
which is also a flavor of UNIX; although, I tend to read that as "Lin-UX"
[Linuks], and to me it means "Linus's UX", or "Linus's UNIX". Which would
make sense. Quite frankly, I'M NOT INTERESTED where the name comes from, as
long as I pronounce it properly.
Richard L. Hamilton
Barry Margolin <bar...@alum.mit.edu> writes:
> In article <103l36d...@corp.supernews.com>,
> Richard.L...@mindwarp.smart.net (Richard L. Hamilton) wrote:
>
>> Unix nowadays is no longer a code base, it's a trademark, and in addition
>> to license fees, one of the conditions of use is compliance with the
>> Single Unix Specification.
>
> Since it hasn't passed the conformance test, it's true that the Linux
> vendors can't legally call their products Unix. But trademark law has
> no bearing on what the rest of us say in informal communications. We
> can say it's Unix, we can say it's Coca Cola, we can say whatever we
> want. For practical purposes, it's as good as any "certified" Unix.
For _some_ practical purposes, maybe. But I don't think I'm the only
one that would have problems if I wrote something in terms of a standard
and then tried to run it on something that didn't quite comply with
that standard.
> And if laypeople use Unix in a generic sense enough, perhaps it will
> become the generic term that it should have been (as happened with
> trademarks like Aspirin).
Oh, don't give me that advocacy crap. ISTR that TOG was considering
waiving the license fee for Linux if it could meet the compliance test,
but it still hasn't happened. Maybe the Linux groupies have too big egos?
(FWIW, I saw something that tried to enumerate the discrepancies between
LSB and POSIX, since LSB has some sort of standards sponsorship too.
One more fewkin' newbie coming along after all the big ($$) guys finally got
together on POSIX and SUS. Now we gotta reconcile it all over again, and
the newbies prolly think we ought to change to meet _them_. "crock" is
about the least rude way I'd describe that!)
Richard L. Hamilton
"UNIX admin" <tripi...@hotmail.com> writes:
> "Thomas Dehn" <thomas...@arcor.de> schrieb im Newsbeitrag
> news:403b1...@news.arcor-ip.de...
>
>> > But if it walks like a duck, swims like a duck, quacks like a duck...
> then
>> > it must be a duck.
>>
>> Only if you compare the chicken (Linux) to the cockroaches.
>
> Do pardon me, but I'm a professional admin, not a zealot. If there is
> /dev/, if there is /proc, /var, /usr, if there is `ls`, `ps`, `fsck`,
> `xterm`, and so on, then to me that means I'm on a UNIX system. I don't
> care if it passed any conformance tests, but if it behaves like UNIX, looks
> like UNIX and feels like UNIX, then it *IS* UNIX.
>
> Again, I'm not an OpenSource and/or Linux/*BSD crusader, but a professional.
> Petty nitpicking and OSS ZEALOTRY is of no interest to me.
If you were not only a an admin but a programmer too, you damn well would
care about compliance.
That all started back with SVR2 and the SVID really - a specification
and validation independent of implementation. From that came POSIX, and
eventually SUS.
A programmer would prefer to write a program as much as possible to a
portable specification, going outside it only in very isolated sections
of the code if absolutely necessary, so that any porting problems
should be isolated and 95%+ of the code should be trivial to port,
without powerful but ideally largely unnecessary tools like autoconf.
I don't think anyone would nitpick about their faves if they all
were pretty much interchangable and it was just the name on the box
(or download) that was in dispute.
Thomas Dehn
"UNIX admin" <tripi...@hotmail.com> wrote:
> "Thomas Dehn" <thomas...@arcor.de> schrieb im Newsbeitrag
> news:403b1...@news.arcor-ip.de...
>
> > > But if it walks like a duck, swims like a duck,
> > > quacks like a duck... then it must be a duck.
> >
> > Only if you compare the chicken (Linux) to the cockroaches.
>
> Do pardon me, but I'm a professional admin, not a zealot. If there is
> /dev/, if there is /proc, /var, /usr, if there is `ls`, `ps`, `fsck`,
> `xterm`, and so on, then to me that means I'm on a UNIX system. I don't
> care if it passed any conformance tests, but if it behaves like UNIX, looks
> like UNIX and feels like UNIX, then it *IS* UNIX.
I'm looking at this from a software developer's PoV.
For example, it bothers me a lot if multithreading is implemented
by splitting an application into several processes, and there
are various other threading oddities as well. Definitely
a POSIX standard about which I care. (In this case, Linux
is getting there with the NPTL)
Thomas
Barry Margolin
Richard.L...@mindwarp.smart.net (Richard L. Hamilton) wrote:
> For _some_ practical purposes, maybe. But I don't think I'm the only
> one that would have problems if I wrote something in terms of a standard
> and then tried to run it on something that didn't quite comply with
> that standard.
I think it's likely that many of those programs won't run on old
versions of AT&T or BSD Unix. Have those retroactively stopped being
Unix as well? I suspect that Linux probably conforms more closely than
the real Unixes that existed 10-20 years ago.
Richard L. Hamilton
Barry Margolin <bar...@alum.mit.edu> writes:
> In article <103moed...@corp.supernews.com>,
> Richard.L...@mindwarp.smart.net (Richard L. Hamilton) wrote:
>
>> For _some_ practical purposes, maybe. But I don't think I'm the only
>> one that would have problems if I wrote something in terms of a standard
>> and then tried to run it on something that didn't quite comply with
>> that standard.
>
> I think it's likely that many of those programs won't run on old
> versions of AT&T or BSD Unix. Have those retroactively stopped being
> Unix as well? I suspect that Linux probably conforms more closely than
> the real Unixes that existed 10-20 years ago.
No, because the definition of Unix has changed and evolved,
from a code base (with some forks) plus trademark (which was always
there - thus some of the early commercial variants weren't allowed
to advertise themselves as Unix), to code base + trademark + specification,
to ultimately just trademark + specification). And of course both the
code base (back in those days) and the specification (through SVID, X/Open,
and SUSv1 through v3) have evolved.
Is Linux closer to SUSv3 than to SysIII (for example?) Sure. And no
doubt that for some purposes it's "good enough". But if you're trying to
write code to a standard rather than just to run on one particular
implementation (with a bunch of #ifdefs to perhaps be added later if
porting is needed), close may not count for much; this isn't like
horseshoes or hand grenades. The difficulty of porting _from_ Linux
is illustrative, I think, although to be fair, it does illustrate that
Linux _has_ gotten closer over time, but that some of the folks writing
apps for Linux still don't get it about writing to a standard rather than
for an implementation.
UNIX admin
> If you were not only a an admin but a programmer too, you damn well would
> care about compliance.
I do care about compliance, and from what I can tell, Linux is at least
trying to be POSIX compliant. What I don't care about is whether some
artifical standards body gives it the official "approved" sticker or not.
It'd be nice, but if it doesn't happen, oh well. Life goes on, much as it
has this past age.
UNIX admin
> I'm looking at this from a software developer's PoV.
> For example, it bothers me a lot if multithreading is implemented
> by splitting an application into several processes, and there
> are various other threading oddities as well. Definitely
> a POSIX standard about which I care. (In this case, Linux
> is getting there with the NPTL)
But here we're NOT discussing Linux programming and its programming
interfaces, rather Solaris issues.
UNIX admin
> Oh, don't give me that advocacy crap. ISTR that TOG was considering
> waiving the license fee for Linux if it could meet the compliance test,
> but it still hasn't happened. Maybe the Linux groupies have too big egos?
Definitely!!! I'm with you in this one!
David Combs
Another Tom <sg7...@snet.net> wrote:
>On Fri, 20 Feb 2004 01:52:54 +0000, Richard L. Hamilton wrote:
>
>> In article <z78Zb.83360$dP1.2...@newsc.telia.net>,
>> Michael Laajanen <michael.laajanen.no-spam.@telia.com> writes:
>>> HI,
>>>
>>> replace Solaris print system with CUPS or ESP, you will understand why
>>> if you try both and compare againts Solaris printer system for easy of
>>> use if you are no expert :))
>>>
>>> http://www.easysw.com/printpro/
>>>
>>
Please clue me in: what are some of the things you can
do with those two (CUPS, ESP, etc) that you find so
much harder with Solaris?
No, I'm not challenging you are anything, just wondering
(especially if those other ones are free, open). Myself,
my most complicated job thus far is like:
lp -n 3 -o nobanner foo.txt (or foo.ps)
or
lp -t "my title" foo.txt
Pretty simple stuff -- for a single-user machine.
What are some *other* things that I might want to do?
Thanks,
David
Richard B. Gilbert
David Combs wrote:
I just have to butt in here!
I have made many fruitless attempts to get Solaris 8 /X86 to print to an
HP LaserJet equipped with a JetDirect card. I followed the directions
in the manual (System Administration Guide, Volume 2). The best I could
get was "stair step" printing. I tried HP's software for installing a
printer. It works great on the Sparc platform but it uses a Sparc
excutable and does not work on Intel.
I know it needs a "filter" somewhere but the manual is silent on exactly
which filter it needs or even where to find the available filters.
(I'll be damned if I'll write my own filter for what has to be one of
the most common printers on the planet!)
Almost anything would be better than the existing software and
documentation!
Greg Andrews
>
>I just have to butt in here!
>
>I have made many fruitless attempts to get Solaris 8 /X86 to print to an
>HP LaserJet equipped with a JetDirect card. I followed the directions
>in the manual (System Administration Guide, Volume 2). The best I could
>get was "stair step" printing. I tried HP's software for installing a
>printer. It works great on the Sparc platform but it uses a Sparc
>excutable and does not work on Intel.
>
>I know it needs a "filter" somewhere but the manual is silent on exactly
>which filter it needs or even where to find the available filters.
>(I'll be damned if I'll write my own filter for what has to be one of
>the most common printers on the planet!)
>
>Almost anything would be better than the existing software and
>documentation!
>
That question has been answered before in this newsgroup.
http://groups.google.com/groups?as_umsgid=9te211%24k02%241%40news.panix.com
You've already tried the fourth option listed in that article (HP's
software), and found it won't work on your Solaris x86 machine.
The first three options in the article use functions and filters that
are present in the sparc and x86 versions of Solaris, so they will work
equally well on your x86 machine.
The manual that has the information about the magic "auto" and "text"
printer names is the Jetdirect manual. Those names are magic to the
Jetdirect card, not to Solaris, so you won't find them in any Solaris
manual.
If your printer is one that lacks Postscript support, but you want to
print graphics on it (which are usually in Postscript format on Unix
machines), then you can use the FAQ answer:
http://www.sun.drydog.com/faq/6.html#6.4
which has also been mentioned in this newsgroup before, and therefore
can be found via google.
Alan Hargreaves - Product Technical Support (APAC)
examples (it's on docs.sun.com). You want to configure a network printer
to use protocol "tcp" to talk to port 9100 (the AppSocket port) on the
printer.
I used this to great effect ( around a hundred printers) in my last job.
alan.
Greg Andrews
>Look at Volume 2 of the system administrators guide. It's got some great
>examples (it's on docs.sun.com). You want to configure a network printer
>to use protocol "tcp" to talk to port 9100 (the AppSocket port) on the
>printer.
>
>I used this to great effect ( around a hundred printers) in my last job.
>
See infodoc 27128 for descriptions of the options to do that.
(Then peek at the author of that doc!)
UNIX admin
news:v_ednbx-n5b...@comcast.com...
> I just have to butt in here!
>
> I have made many fruitless attempts to get Solaris 8 /X86 to print to an
> HP LaserJet equipped with a JetDirect card. I followed the directions
> in the manual (System Administration Guide, Volume 2). The best I could
> get was "stair step" printing. I tried HP's software for installing a
> printer. It works great on the Sparc platform but it uses a Sparc
> excutable and does not work on Intel.
>
> I know it needs a "filter" somewhere but the manual is silent on exactly
> which filter it needs or even where to find the available filters.
> (I'll be damned if I'll write my own filter for what has to be one of
> the most common printers on the planet!)
>
> Almost anything would be better than the existing software and
> documentation!
If you have a LaserJet with a JetDirect card, then you should be able to set
the printer up as a network printer -- with its own IP address, in which
case Solaris should be set up to print to a network print server, which
would be the LaserJet. HP's software is for admining the LaserJet via the
JetDirect card, but you can probably download it for Windows and set the
printer up in it. Then just print to a network print server from Solaris,
and have Solaris queue the requests for the printer via lp/lpr/Samba.
Alan Hargreaves - Product Technical Support (APAC)
No, don't set it up as a remote printer that way.
1. It is not a supported configuration (if you are going to do lpd, use
netpr and protocol=bsd, that way solaris handles the queueing).
2. I have yet to see a printer implement lpd in a multithreaded fashion.
ie if the printer is busy, expect your commands like lpq to hang.
The filter you want would be unix2dos (1).
alan.
UNIX admin
news:c3928b$pbu$1...@new-usenet.uk.sun.com...
> > If you have a LaserJet with a JetDirect card, then you should be able to
set
> > the printer up as a network printer -- with its own IP address, in which
> > case Solaris should be set up to print to a network print server, which
> > would be the LaserJet. HP's software is for admining the LaserJet via
the
> > JetDirect card, but you can probably download it for Windows and set the
> > printer up in it. Then just print to a network print server from
Solaris,
> > and have Solaris queue the requests for the printer via lp/lpr/Samba.
> >
> >
>
> No, don't set it up as a remote printer that way.
>
> 1. It is not a supported configuration (if you are going to do lpd, use
> netpr and protocol=bsd, that way solaris handles the queueing).
> 2. I have yet to see a printer implement lpd in a multithreaded fashion.
> ie if the printer is busy, expect your commands like lpq to hang.
>
> The filter you want would be unix2dos (1).
Wouldn't Solaris handle queueing anyway? Can't one set up a local queue to
a remote printer?
Greg Andrews
>UNIX admin wrote:
>>
>> If you have a LaserJet with a JetDirect card, then you should be able to set
>> the printer up as a network printer -- with its own IP address, in which
>> case Solaris should be set up to print to a network print server, which
>> would be the LaserJet. HP's software is for admining the LaserJet via the
>> JetDirect card, but you can probably download it for Windows and set the
>> printer up in it. Then just print to a network print server from Solaris,
>> and have Solaris queue the requests for the printer via lp/lpr/Samba.
>>
>>
>
>No, don't set it up as a remote printer that way.
>
Alan, "UNIX Admin's" description is so vague that you can't tell
what kind of setup is being described.
>
>1. It is not a supported configuration (if you are going to do lpd, use
>netpr and protocol=bsd, that way solaris handles the queueing).
>2. I have yet to see a printer implement lpd in a multithreaded fashion.
>ie if the printer is busy, expect your commands like lpq to hang.
>
>The filter you want would be unix2dos (1).
>
In most cases, that's worse than using the Jetdirect card's "text"
printer name. The reason is unix2dos badly mangles PCL files, because
it does more data conversions than just adding a Carriage Return to
Line Feeds. The Jetdirect card's "text" printer name doesn't do that.
"lpadmin -o protocol=bsd,dest=printer:text" is the better way to go
unless you know PCL files are never sent to the printer.
Greg Andrews
>
>Wouldn't Solaris handle queueing anyway? Can't one set up a local queue to
>a remote printer?
>
Nope. A "remote printer" configuration means the local machine is
merely a print client - not a print server. The machine that is
designated as the print server in the remote printer config is the
print server. It's the print server's responsibility to queue the
print jobs for feeding to the printer one at a time.
Since this is Unix, the print clients don't need to queue the jobs
to feed to the print server one at a time. The clients can feed
print jobs to the server with multiple simultaneous connections, so
the jobs are queued on the server faster.
Starting with Solaris 2.6, the print client configuration does not
perform queueing of print jobs. The lp/lpr binary saves the job in
the /var/spool/print directory and forks a background process to feed
the print job to the server. If you invoke lp several times quickly
to print to the same printer, you'll have several jobs in /var/spool/print,
each with a background process feeding the job to the print server.
When you review the evolution of Unix printing, you'll see that the
basic print system had queueing because printers were attached to
parallel or serial ports. The Berkeley lpd simply attached the network
printing protocol (RFC 1179) to the existing code that had queueing
and filtering. So print clients ended up with FIFO queueing and
filtering, even though it made no sense. AT&T did the same thing with
the SVR4 "s5" printing protocol - bolted it onto the side of the
existing code. On Solaris 2.6 and higher, Sun broke away from the
SVR4 mold (and a ton of architectural bugs) and the result is that
the print client does not queue and does not filter. Those functions
belong on the print server - especially filtering.
If your Solaris machine must feed print jobs directly to a network
printer, then it needs queueing and filtering. Set it up with the
"network printer" config instead of the "remote printer" config.
Only use the "remote printer" config when your Solaris machine feeds
the print jobs to a multitasking computer that's a print server
(Unix machines, and even Win NT or 2000 servers). Very few network
print servers like Jetdirect cards/boxes have queueing and filtering,
so even though people call them "print servers", you should use the
"network printer" config rather than the "remote printer" config.
Mark Valery
No, don't set it up as a remote printer that way.
>
> 1. It is not a supported configuration (if you are going to do lpd, use
> netpr and protocol=bsd, that way solaris handles the queueing).
> 2. I have yet to see a printer implement lpd in a multithreaded fashion.
> ie if the printer is busy, expect your commands like lpq to hang.
>
> The filter you want would be unix2dos (1).
>
> alan.
work, but this is what I have done at home to a jetDirect server
connected to a HP deskJet and at work to a jetDirect server connected
to a HP laserJet postscript printer.
lpadmin -p lj8150dn_dp -s 10.1.10.48
This results in an entry in /etc/printers.conf
lj8150dn_dp:\
:bsdaddr=10.1.10.48,lj8150dn_dp,Solaris:
I pipe my postscript files to lpr -Plj8150dn_dp
At home, the command for creating the printer queue is:
lpadmin -p oj -s 100.0.1.100
I convert my text files to postscript using mpage, send the result
to ghostscript to create a PCL file for my specific deskjet and then
pipe the PCL file to lpr -Poj
I have also used hpijs to create color PCL output, but I use the same
printer queue setup.
Hope this helps. I have never had any queuing problems, even when
other machines have sent jobs to the same printer I'm trying to
print to.
Alan Hargreaves - Product Technical Support (APAC)
this one more times than I care to remember.
If you use netpr and protocol=bsd, *then* you do the queueing on the
solaris host. You can then use remote printers to talk to the single
solaris host that talks to the printer.
The other problem that you get by having multiple machines using remote
printing directly to a printer is that TCP doesn't queue connections
requests. If you have one machine talking to the printer, any otehrs
will time out.
alan.
Greg Andrews
>
>lpadmin -p lj8150dn_dp -s 10.1.10.48
>This results in an entry in /etc/printers.conf
>lj8150dn_dp:\
> :bsdaddr=10.1.10.48,lj8150dn_dp,Solaris:
>
(i.e. a remote printer configuration)
>
>Hope this helps. I have never had any queuing problems, even when
>other machines have sent jobs to the same printer I'm trying to
>print to.
>
Try issuing 100 lp commands for that printer very quickly, such as
in a script. You'll find that a few of the jobs become stuck on
the Solaris machine because the printer's Jetdirect card or box
accepted all the simultaneous TCP connections but never returned
the printing protocol responses on some of them. The Jetdirect
responded on all the other connections, but lost track of those few.
Then switch the Solaris queue to the network printer type:
touch /dev/lj8150dn_dp
chmod 600 /dev/lj8150dn_dp
chown lp:lp /dev/lj8150dn_dp
lpadmin -p lj8150dn_dp -v /dev/lj8150dn_dp -m netstandard
lpadmin -p lj8150dn_dp -o dest=10.1.10.48 -I any
enable lj8150dn_dp
accept lj8150dn_dp
Note that "-I any" uses an uppercase i as the option letter, not
a lowercase L. Some newsreaders make it hard to tell the difference.
With the network printer configuration, you can issue as many lp
commands as you like in a quick burst, and none of them will be
stranded on the Solaris machine. Solaris queues the jobs and makes
just one connection at a time to the printer's Jetdirect card/box,
and never sees the problem where the Jetdirect loses track of the
connection.
Mark Valery
> Try issuing 100 lp commands for that printer very quickly, such as
> in a script. You'll find that a few of the jobs become stuck on
> the Solaris machine because the printer's Jetdirect card or box
> accepted all the simultaneous TCP connections but never returned
> the printing protocol responses on some of them. The Jetdirect
> responded on all the other connections, but lost track of those few.
Thanks, that explains why I never had a problem at my former
companies. We never used scripts to fire off a bunch of jobs,
just interactive users. I'll give your setup a try the next
time I add a printer.
Greg Andrews
There are some software packages that let print jobs build up
in an internal queue, then issue dozens of rapid-fire lp commands
to print them. Certain configurations of the SAS software will
do that. Also, once in a while on large servers, you'll get a
bunch of people who all decide to print at the same time.
It's not a universal problem (i.e. not everyone sees it), but it
is seen by sites that have bursty patterns of printing.
Richard B. Gilbert
Greg Andrews wrote:
Where would I find "infodoc 27128"? Google came up with a single
reference containing a broken link and Groups.Google came up with a
pointer to your message!
Alan Hargreaves - Product Technical Support (APAC)
It's in the public collection. If mozilla doesn't line wrap on me, here
is a pointer to it:
http://sunsolve.sun.com/pub-cgi/retrieve.pl?doc=finfodoc%2F27128&zone_110=27128&wholewords=on
alan.
--
Alan Hargreaves
Senior Technical Support Specialist/VOSJEC Engineer
Product Technical Support (APAC)
Sun Microsystems
Greg Andrews
>reference containing a broken link and Groups.Google came up with a
>pointer to your message!
>