[9fans] man -t cat|lp takes 4min

[9fans] man -t cat|lp takes 4min

[west9]

With, for example, man -t cat|lp the printer light starts blinking right after
 <return>. It then takes about 4 minutes before printing begins; the output is proper.

The action is the same with or without the change to lpdspool suggested in the 
 *printer configuration* document.

 This problem appeared about 6 months ago, possibly after a replica/pull.
 Before that I noticed no delay in printing man pages this way.

 The printer is an NEC 1800N with 12MB memory, driven via lpt. It's a postscript printer.

  Suggestions, please.

Thanks,
 -Tom

Re: [9fans] man -t cat|lp takes 4min

[boyd, rounin]

> With, for example, man -t cat|lp the printer light starts blinking right after
> <return>. It then takes about 4 minutes before printing begins; the output is proper.

it's not in that silly standby state is it?
--
MGRS 31U DQ 52572 12604

Re: [9fans] man -t cat|lp takes 4min

[Russ Cox]

Try man -t cat | lp -dstdout >/tmp/cat.ps
cat /tmp/cat.ps >/dev/lpt1data

and see which is taking the long time.
Presumably it's the cat to lpt1data,
but devlpt hasn't changed in at least 3 years.

Russ

On 5/11/05, west9@worldnet.att.net <west9@worldnet.att.net> wrote:
> 
> With, for example, man -t cat|lp the printer light starts blinking right after
>  <return>. It then takes about 4 minutes before printing begins; the output is proper.
> 
> The action is the same with or without the change to lpdspool suggested in the
>  *printer configuration* document.
> 
>  This problem appeared about 6 months ago, possibly after a replica/pull.
>  Before that I noticed no delay in printing man pages this way.
> 
>  The printer is an NEC 1800N with 12MB memory, driven via lpt. It's a postscript printer.
> 
>   Suggestions, please.
> 
> Thanks,
>  -Tom
>

Re: [9fans] man -t cat|lp takes 4min

[geoff]

I wonder if lp is waiting for something to time out.  You might try
`lp -D' and see if that tells you anything.

Re: [9fans] man -t cat|lp takes 4min

[Bruce Ellis]

only a guess but i think gs is to blame.  15 years ago i spent a lot
of time constructing a postscript page that would take 5 minutes.
and that was a 68020 dinosaur engine.

use an older gs or just wait ... and wait ....

brucee

On 5/12/05, geoff@collyer.net <geoff@collyer.net> wrote:
> I wonder if lp is waiting for something to time out.  You might try
> `lp -D' and see if that tells you anything.

Re: [9fans] man -t cat|lp takes 4min

[Russ Cox]

while gs is often to blame, i doubt it is to blame
in this case -- he's using a postscript printer, so
gs shouldn't even be involved.

russ

On 5/12/05, Bruce Ellis <bruce.ellis@gmail.com> wrote:
> only a guess but i think gs is to blame.  15 years ago i spent a lot
> of time constructing a postscript page that would take 5 minutes.
> and that was a 68020 dinosaur engine.
> 
> use an older gs or just wait ... and wait ....
> 
> brucee
> 
> On 5/12/05, geoff@collyer.net <geoff@collyer.net> wrote:
> > I wonder if lp is waiting for something to time out.  You might try
> > `lp -D' and see if that tells you anything.
>

Re: [9fans] man -t cat|lp takes 4min

[Bruce Ellis]

*shouldn't* is not a scientific term.

did we try something older that WORKED.

brucee

On 5/12/05, Russ Cox <russcox@gmail.com> wrote:
> while gs is often to blame, i doubt it is to blame
> in this case -- he's using a postscript printer, so
> gs shouldn't even be involved.
> 
> russ
> 
> On 5/12/05, Bruce Ellis <bruce.ellis@gmail.com> wrote:
> > only a guess but i think gs is to blame.  15 years ago i spent a lot
> > of time constructing a postscript page that would take 5 minutes.
> > and that was a 68020 dinosaur engine.
> >
> > use an older gs or just wait ... and wait ....
> >
> > brucee
> >
> > On 5/12/05, geoff@collyer.net <geoff@collyer.net> wrote:
> > > I wonder if lp is waiting for something to time out.  You might try
> > > `lp -D' and see if that tells you anything.
> >
>

Re: [9fans] man -t cat|lp takes 4min

[Jack Johnson]

On 5/11/05, Bruce Ellis <bruce.ellis@gmail.com> wrote:
> *shouldn't* is not a scientific term.

There might be one or two statisticians out there that disagree. ;)

-Jack

Re: [9fans] man -t cat|lp takes 4min

[Bruce Ellis]

probability is reasonably well understood.
statistics?  give me a break.

i recent poll of 3000 idiots showed that they were idiots.

brucee

On 5/13/05, Jack Johnson <knapjack@gmail.com> wrote:
> On 5/11/05, Bruce Ellis <bruce.ellis@gmail.com> wrote:
> > *shouldn't* is not a scientific term.
> 
> There might be one or two statisticians out there that disagree. ;)
> 
> -Jack

Re: [9fans] man -t cat|lp takes 4min

[C H Forsyth]

>>Alt case:  man -t cat>/tmp/cat.ps;cat /tmp/cat.ps>/dev/lpt1data: 40 sec., but

surely that produces a copy of the ditroff not postscript?
did you leave out `|lp -d stdout' in that?

>>Could it be that the fonts are being sent/requested unnecessarily?

lp -d stdout won't include the Lucida fonts, but
lp direct to a printer will add them if needed (which they probably are).
they look to be about 740k bytes.
devlpt sets the parallel port to its original 1980s mode.
it will be slow.
given all that, 4 minutes seems not too surprising.
short of revamping devlpt, i'd suggest you print several pages at once,
make sure it has plenty of paper, and take yourself out for lunch.

(it's not too bad with usb.)

Re: [9fans] man -t cat|lp takes 4min

[west9]

>Try  -dstdout >/tmp/cat.ps
>cat /tmp/cat.ps >/dev/lpt1data

Here are the print times for some cases:

Base case, about 500 char plain text: 25 sec

Problem case:  man -t cat|lp : 4 minutes ; But printer light starts blinking a
 few secs after <cr>

Alt case:  man -t cat>/tmp/cat.ps;cat /tmp/cat.ps>/dev/lpt1data: 40 sec., but
 printing is in times font, which is native to the printer. 
 cat /tmp/cat.ps|page yields the man page in the usual man font.

>I wonder if lp is waiting for something to time out.  You might try
>`lp -D' and see if that tells you anything.

 I used the -D option and didn't see any trouble, but that might not mean much,
 given my level. I don't even know how to capture it easily, given the number of apostrophes it contains.

 Could it be that the fonts are being sent/requested unnecessarily?

Thanks to all,

 -Tom