On 2022-08-06, Scott Lurndal <scott@slp53.sl.home> wrote:
> Muttley@dastardlyhq.com writes:
>>On Sat, 06 Aug 2022 15:25:23 GMT
>>scott@slp53.sl.home (Scott Lurndal) wrote:
>>>Muttley@dastardlyhq.com writes:
>>>>On Thu, 04 Aug 2022 23:23:05 GMT
>>>>scott@slp53.sl.home (Scott Lurndal) wrote:
>>>>>That's why we always either use she-bang, or use the invoke the script
>>>>
>>>>Wtf is "she" for? Its either called hash or pound depending on what side of
>>>>the atlantic you're on.
>>>
>>>It is also known as "sharp" by musicians.
>>
>>True. Also MS with C# :)
>
> Which is the source for _SH_e bang.

The prevailing hypotheses are two: that it's a contraction of "sharp
bang" or "hash bang".

--
TXR Programming Language: http://nongnu.org/txr
Cygnal: Cygwin Native Application Library: http://kylheku.com/cygnal

On 2022-08-06, Keith Thompson <Keith.S.Thompson+u@gmail.com> wrote:
> Kaz Kylheku <480-992-1380@kylheku.com> writes:
>> Moreover, if we trace this again with -f to see what the child is doing:
>>
>> 1670 execve("./testscript", ["./testscript"], 0x1e00500 /* 24 vars */) = -1 ENOEXEC (Exec format error)
>> 1670 openat(AT_FDCWD, "./testscript", O_RDONLY|O_LARGEFILE) = 3
>> 1670 read(3, "FOO=abc\ncd /\n", 128) = 13
>> 1670 close(3) = 0
>>
>> See: execve happens.
>>
>> It cannot be that bash would read a command and just run it in the
>> current shell, if that command isn't a built-in or user-defined
>> function.
>
> Agreed.

Ah, I missed the -1 return value from execve(). It failed and returned
to the child shell.

By the way, what I didn't show from that log, is that a path search
takes place, looking for bash. It's not clear for what purpose!

1670 stat64(".", {st_mode=S_IFDIR|0755, st_size=12288, ...}) = 0
1670 stat64("/usr/local/sbin/bash", 0xbfd7534c) = -1 ENOENT (No such file or directory)
1670 stat64("/usr/local/bin/bash", 0xbfd7534c) = -1 ENOENT (No such file or directory)
1670 stat64("/usr/sbin/bash", 0xbfd7534c) = -1 ENOENT (No such file or directory)
1670 stat64("/usr/bin/bash", 0xbfd7534c) = -1 ENOENT (No such file or directory)
1670 stat64("/sbin/bash", 0xbfd7534c) = -1 ENOENT (No such file or directory)
1670 stat64("/bin/bash", {st_mode=S_IFREG|0755, st_size=1235608, ...}) = 0
1670 stat64("/bin/bash", {st_mode=S_IFREG|0755, st_size=1235608, ...}) = 0
1670 geteuid32() = 500
1670 getegid32() = 500
1670 getuid32() = 500
1670 getgid32() = 500
1670 access("/bin/bash", X_OK) = 0
1670 stat64("/bin/bash", {st_mode=S_IFREG|0755, st_size=1235608, ...}) = 0
1670 geteuid32() = 500
1670 getegid32() = 500
1670 getuid32() = 500
1670 getgid32() = 500
1670 access("/bin/bash", R_OK) = 0
1670 stat64("/bin/bash", {st_mode=S_IFREG|0755, st_size=1235608, ...}) = 0
1670 stat64("/bin/bash", {st_mode=S_IFREG|0755, st_size=1235608, ...}) = 0
1670 geteuid32() = 500
1670 getegid32() = 500
1670 getuid32() = 500
1670 getgid32() = 500
1670 access("/bin/bash", X_OK) = 0
1670 stat64("/bin/bash", {st_mode=S_IFREG|0755, st_size=1235608, ...}) = 0
1670 geteuid32() = 500
1670 getegid32() = 500
1670 getuid32() = 500
1670 getgid32() = 500
1670 access("/bin/bash", R_OK) = 0
1670 gettimeofday({tv_sec=1659771305, tv_usec=34961}, NULL) = 0
1670 getpid() = 1670
1670 getpgrp() = 1616

Because then then 1670 reads the script and runs it; nothing is done
with /bin/bash. I should try the experiment with bash not being found
in the path; does it make some kind of difference?

> (I never write scripts without a #! line, so I hadn't noticed this
> behavior until recently.)

But you can trade all the sane behavior for POSIX conformance,
so it may be worth it to reconsider? :)

Now here is what happens when I run the same script from a non-shell
process:

[pid 9814] execve("./testscript", ["./testscript"], 0xbfaaef14 /* 58 vars */) = -1 ENOEXEC (Exec format error)
[pid 9814] execve("/bin/sh", ["/bin/sh", "./testscript"], 0xbfaaef14 /* 58 vars */) = 0

The process made a single call to the execvp function in the C library.

This is documented in the Linux Programmer's Manual man page:

In addition, certain errors are treated specially [by the
functions execlp, execvp and execvpe].

If permission is denied for a file (the attempted execve(2)
failed with the error EACCES), these functions will continue
searching the rest of the search path. If no other file is
found, however, they will return with errno set to EACCES.

If the header of a file isn't recognized (the attempted execve(2)
failed with the error ENOEXEC), these functions will
execute the shell (/bin/sh) with the path of the file as its
first argument. (If this attempt fails, no further searching is
done.)

So, if you use execve() on an executable a shell script that has no hash
bang, it won't work; only the "p" functions emulate this. The script
looks executable only to applications that use execvp and friends.

--
TXR Programming Language: http://nongnu.org/txr
Cygnal: Cygwin Native Application Library: http://kylheku.com/cygnal

On 07/08/2022 01:03, Muttley@dastardlyhq.com wrote:
> On Sat, 06 Aug 2022 12:30:52 +0100
> Richard Kettlewell <invalid@invalid.invalid> wrote:
>> Muttley@dastardlyhq.com writes:
>>> scott@slp53.sl.home (Scott Lurndal) wrote:
>>>> That's why we always either use she-bang, or use the invoke the script
>>>
>>> Wtf is "she" for? Its either called hash or pound depending on what side of
>>> the atlantic you're on. Its never called "she" and I've never heard anyone
>>> anywhere ever refer to it like this.
>>
>> People have been calling it that for ages. I’m not a fan but it’s not
>> new.
>
> I've never heard it called that and I've been using *nix for 30 years. Perhaps
> its an american thing.
>
>
Been in use here in Oz since the 80s, to my knowledge.

Cheers,
Gary B-)

Eli the Bearded <*@eli.users.panix.com> writes:

> Note that I provided numbers for usage on NetBSD.

But what's the default shell on NetBSD and is it different on Panix?

On 07.08.2022 at 15:22, Anssi Saari scribbled:

> Eli the Bearded <*@eli.users.panix.com> writes:
>
> > Note that I provided numbers for usage on NetBSD.
>
> But what's the default shell on NetBSD and is it different on Panix?

NetBSD uses ash (Almquist SHell) by default.

--
With respect,
= Aragorn =

On Sat, 06 Aug 2022 17:45:19 -0400
Dan Espen <dan1espen@gmail.com> wrote:
>Muttley@dastardlyhq.com writes:
>
>> On Sat, 06 Aug 2022 08:13:34 -0400
>> Dan Espen <dan1espen@gmail.com> wrote:
>>>Muttley@dastardlyhq.com writes:
>>>Someone seems to be batting 1000.
>>
>> Is that some obscure sports analogy?
>
>Not obscure. Easily accessed by doing a search.

You posted it, you explain otherwise STFU.

On Sun, 7 Aug 2022 21:26:59 +1000
"Gary R. Schmidt" <grschmidt@acm.org> wrote:
>On 07/08/2022 01:03, Muttley@dastardlyhq.com wrote:
>> On Sat, 06 Aug 2022 12:30:52 +0100
>> Richard Kettlewell <invalid@invalid.invalid> wrote:
>>> Muttley@dastardlyhq.com writes:
>>>> scott@slp53.sl.home (Scott Lurndal) wrote:
>>>>> That's why we always either use she-bang, or use the invoke the script
>>>>
>>>> Wtf is "she" for? Its either called hash or pound depending on what side of
>
>>>> the atlantic you're on. Its never called "she" and I've never heard anyone
>>>> anywhere ever refer to it like this.
>>>
>>> People have been calling it that for ages. I’m not a fan but it’s not
>>> new.
>>
>> I've never heard it called that and I've been using *nix for 30 years.
>Perhaps
>> its an american thing.
>>
>>
>Been in use here in Oz since the 80s, to my knowledge.

I wonder how long a test match would have to be to bat 1000. A month?
I imagine he's talking about baseball aka the kids game rounders just like
american football is just rugby with body armour and lots of rest breaks for
the fat steroid whales who play it.

Muttley@dastardlyhq.com writes:

> On Sat, 06 Aug 2022 17:45:19 -0400
> Dan Espen <dan1espen@gmail.com> wrote:
>>Muttley@dastardlyhq.com writes:
>>
>>> On Sat, 06 Aug 2022 08:13:34 -0400
>>> Dan Espen <dan1espen@gmail.com> wrote:
>>>>Muttley@dastardlyhq.com writes:
>>>>Someone seems to be batting 1000.
>>>
>>> Is that some obscure sports analogy?
>>
>>Not obscure. Easily accessed by doing a search.
>
> You posted it, you explain otherwise STFU.

Oh sorry, didn't realize you were of limited abilities.

https://en.wiktionary.org/wiki/bat_a_thousand

--
Dan Espen

Aragorn <thorongil@telenet.be> writes:

> On 07.08.2022 at 15:22, Anssi Saari scribbled:
>
>> Eli the Bearded <*@eli.users.panix.com> writes:
>>
>> > Note that I provided numbers for usage on NetBSD.
>>
>> But what's the default shell on NetBSD and is it different on Panix?
>
> NetBSD uses ash (Almquist SHell) by default.

As users' default login shell? In case that wasn't clear.

On 07.08.2022 at 20:25, Anssi Saari scribbled:

> Aragorn <thorongil@telenet.be> writes:
>
> > On 07.08.2022 at 15:22, Anssi Saari scribbled:
> >
> >> Eli the Bearded <*@eli.users.panix.com> writes:
> >>
> >> > Note that I provided numbers for usage on NetBSD.
> >>
> >> But what's the default shell on NetBSD and is it different on
> >> Panix?
> >
> > NetBSD uses ash (Almquist SHell) by default.
>
> As users' default login shell? In case that wasn't clear.

I believe so, yes. NetBSD aims for full POSIX compliance and ash is a
POSIX shell.

--
With respect,
= Aragorn =

Muttley@dastardlyhq.com writes:
> On Sat, 06 Aug 2022 17:45:19 -0400
> Dan Espen <dan1espen@gmail.com> wrote:
>>Muttley@dastardlyhq.com writes:
>>
>>> On Sat, 06 Aug 2022 08:13:34 -0400
>>> Dan Espen <dan1espen@gmail.com> wrote:
>>>>Muttley@dastardlyhq.com writes:
>>>>Someone seems to be batting 1000.
>>>
>>> Is that some obscure sports analogy?
>>
>>Not obscure. Easily accessed by doing a search.
>
> You posted it, you explain otherwise STFU.

Be less rude, please. (You could start by not posting English to
de.comp.os.unix.shell, which I've dropped.)

--
Keith Thompson (The_Other_Keith) Keith.S.Thompson+u@gmail.com
Working, but not speaking, for Philips
void Void(void) { Void(); } /* The recursive call of the void */

Anssi Saari <as@sci.fi> wrote:
> Aragorn <thorongil@telenet.be> writes:
<snip>

>> NetBSD uses ash (Almquist SHell) by default.
>
> As users' default login shell? In case that wasn't clear.

When you install NetBSD, yor are prompted to choose the
shell for root. IIRC, the options are:

1. sh (ash)
2. ksh (PDKSH, v5.2.14 99/07/13.2 per man)
3. csh (the real csh, not tcsh)

I picked sh for root.

When creating users, you can choose any shell that is on
the system. I went with tcsh from pkgsrc for my User ID.

John

--
[t]csh(1) - "An elegant shell, for a more... civilized age."
- Paraphrasing Star Wars

On 2022-08-07, Muttley@dastardlyhq.com <Muttley@dastardlyhq.com> wrote:
> I wonder how long a test match would have to be to bat 1000. A month?

As short as a few seconds?

A rookie participant who has never swung a bat before, let alone played
ball, goes to bat, and happens to hit the one pitch that is thrown.

At that point, his or her batting average is 1000.

--
TXR Programming Language: http://nongnu.org/txr
Cygnal: Cygwin Native Application Library: http://kylheku.com/cygnal

On Sun, 07 Aug 2022 10:31:00 -0400
Dan Espen <dan1espen@gmail.com> wrote:
>Muttley@dastardlyhq.com writes:
>
>> On Sat, 06 Aug 2022 17:45:19 -0400
>> Dan Espen <dan1espen@gmail.com> wrote:
>>>Muttley@dastardlyhq.com writes:
>>>
>>>> On Sat, 06 Aug 2022 08:13:34 -0400
>>>> Dan Espen <dan1espen@gmail.com> wrote:
>>>>>Muttley@dastardlyhq.com writes:
>>>>>Someone seems to be batting 1000.
>>>>
>>>> Is that some obscure sports analogy?
>>>
>>>Not obscure. Easily accessed by doing a search.
>>
>> You posted it, you explain otherwise STFU.
>
>Oh sorry, didn't realize you were of limited abilities.
>
>https://en.wiktionary.org/wiki/bat_a_thousand

More a case of limited giving a shit.

On Mon, 8 Aug 2022 07:07:23 -0000 (UTC)
Kaz Kylheku <480-992-1380@kylheku.com> wrote:
>On 2022-08-07, Muttley@dastardlyhq.com <Muttley@dastardlyhq.com> wrote:
>> I wonder how long a test match would have to be to bat 1000. A month?
>
>As short as a few seconds?
>
>A rookie participant who has never swung a bat before, let alone played
>ball, goes to bat, and happens to hit the one pitch that is thrown.
>
>At that point, his or her batting average is 1000.

No idea what TF you're talking about. I was talking about cricket to an
australian, hence the "test match". In cricket the best you can do in one hit
is a 6. The highest single batsman score in its history is AFAIK 264.

On 08/08/2022 17:07, Kaz Kylheku wrote:
> On 2022-08-07, Muttley@dastardlyhq.com <Muttley@dastardlyhq.com> wrote:
>> I wonder how long a test match would have to be to bat 1000. A month?
>
> As short as a few seconds?
>
> A rookie participant who has never swung a bat before, let alone played
> ball, goes to bat, and happens to hit the one pitch that is thrown.
>
> At that point, his or her batting average is 1000.
>
Well, in cricket, rather than the jumped-up game of rounders that the
seppos play, having an average of "1000" would mean having averaged 1000
runs over every innings played without being "out":

"How do you calculate batsman average?
To calculate the batting average of any player, you need to divide the
total number of runs scored by that player by the number of times he or
she has got out (or been dismissed). We can calculate the number of
times the player has been dismissed by subtracting the number of not
outs from the number of innings batted."

Which is why Bradman's batting average is 99.94.

Cheers,
Gary B-)

Keith Thompson wrote:

> Since the #! feature is nearly universal for Unix-like systems and
> predates POSIX, I'm not sure why POSIX didn't standardize it.

I think it is mainly because POSIX doesn't specify pathnames for
utilities. It is very common to start a shell script with #! /bin/sh
but on Solaris 10 and earlier that would get you an old Bourne shell,
not a POSIX conforming shell (which would need #! /usr/xpg4/bin/sh).

So even if POSIX specified #!, portable applications wouldn't be
able to hard-code that line. They would need a config/install process
that finds the right version of sh (or whatever utility) and adds or
updates the #! line.

--
Geoff Clare <netnews@gclare.org.uk>

Geoff Clare <geoff@clare.See-My-Signature.invalid> writes:
> Keith Thompson wrote:
>> Since the #! feature is nearly universal for Unix-like systems and
>> predates POSIX, I'm not sure why POSIX didn't standardize it.
>
> I think it is mainly because POSIX doesn't specify pathnames for
> utilities. It is very common to start a shell script with #! /bin/sh
> but on Solaris 10 and earlier that would get you an old Bourne shell,
> not a POSIX conforming shell (which would need #! /usr/xpg4/bin/sh).
>
> So even if POSIX specified #!, portable applications wouldn't be
> able to hard-code that line. They would need a config/install process
> that finds the right version of sh (or whatever utility) and adds or
> updates the #! line.

Sure, but POSIX could have specified the behavior of #! and the meaning
of its argument(s) without depending on any particular pathnames.

--
Keith Thompson (The_Other_Keith) Keith.S.Thompson+u@gmail.com
Working, but not speaking, for Philips
void Void(void) { Void(); } /* The recursive call of the void */

In comp.unix.shell, Anssi Saari <as@sci.fi> wrote:
> Eli the Bearded <*@eli.users.panix.com> writes:
> > Note that I provided numbers for usage on NetBSD.
> But what's the default shell on NetBSD

Don't know, but probably not relevant.

> and is it different on Panix?

Users specify shell at time of sign up. The current web form defaults
to bash. When I signed up it was over the phone, and I'm not sure there
was a web form.

I will note that /bin/sh is currently available from chsh, but not from
the web form. That could be why it has such poor numbers.

Also, 'ksh' here is PD KSH, but nothing tells you that. Which is why I
expect real ksh has such low numbers by comparison, real ksh being
'ksh93' makes it sound like an obsolete version.

I don't remember how shell migration was handled when the switch to
NetBSD started.

Elijah
------
wasn't around for the migration from A/UX

David W. Hodgins <dwhodgins@nomail.afraid.org> wrote:
>
>Also, I was taught the ! is called an exclamation mark, not a bang.
>

Never heard the term "bang path"?

https://en.wikipedia.org/wiki/UUCP#Mail_routing

The Path: header line in your post to this newsgroup looked like this:

Path: reader2.panix.com!panix!3.eu.feeder.erje.net!feeder.erje.net!eternal-september.org!reader01.eternal-september.org!.POSTED!not-for-mail

though you likely don't care.

-Greg
--
::::::::::::: Greg Andrews ::::: gerg@panix.com :::::::::::::
I have a map of the United States that's actual size.
-- Steven Wright

On Mon, 08 Aug 2022 22:41:02 -0400, gerg <gerg@panix.com> wrote:

> David W. Hodgins <dwhodgins@nomail.afraid.org> wrote:
>>
>> Also, I was taught the ! is called an exclamation mark, not a bang.
>>
>
> Never heard the term "bang path"?
>
> https://en.wikipedia.org/wiki/UUCP#Mail_routing
>
> The Path: header line in your post to this newsgroup looked like this:
>
> Path: reader2.panix.com!panix!3.eu.feeder.erje.net!feeder.erje.net!eternal-september.org!reader01.eternal-september.org!.POSTED!not-for-mail
>
> though you likely don't care.

I have heard of it, but didn't associate it with the exclamation marks. I thought
it was just a term for explicitly choosing the path.

Regards, Dave Hodgins

On 2022-08-04, John D Groenveld <groenveld@acm.org> wrote:
> In article <2lprri-c25.ln1@paranoia.mcleod-schmidt.id.au>,
> Gary R. Schmidt <grschmidt@acm.org> wrote:
>>zsh is big in Japan, not just on Linux boxes but on Solaris systems as
>

ksh is the default shell on OpenBSD, PD KSH v5.2.14 99/07/13.2.
ksh93 is in ports, AJM 93u+m/1.0.3 2022-08-25

> I switched to zsh as my preferred interactive shell, but code
> in ksh on illumos, FreeBSD, Linux and Solaris.
>
> Kudos to Martijn for leading ksh's revival.

Martijn, keep up the good work!

--
XMPP tool@disroot.org
DB4CE71C8DEF84257AD0B23E4D5934D90226C7F3 00010011 00010011 ▞