On Tue, 15 Aug 2023 23:32:22 +0300
Phil Carmody <pc+usenet@asdf.org> wrote:
>Muttley@dastardlyhq.com writes:
>> On Tue, 15 Aug 2023 16:12:19 -0000 (UTC)
>> kalevi@kolttonen.fi (Kalevi Kolttonen) wrote:
>>>Kenny McCormack <gazelle@shell.xmission.com> wrote:
>>>> But do they know that???
>>>
>>>"They" probably don't know it. But let's face it,
>>>nobody really wants to create a file having '-'
>>>filename on purpose. The filename is not descriptive
>>>at all, it would be just an insane choice for anything
>>>useful.
>>
>> I used to think the same thing about spaces in filenames. Then along came
>> Windows.
>
>Your memory if flawed. Spaces were always allowed in unix
>filenames. Often used to hide things in plain sight on FTP sites, for
>example.

Any character from 0-255 is allowed in unix filenames but only an idiot would
use spaces and non printing characters.

>MS Windows didn't allow spaces until Windows 95. I remember the Apple

DOS didn't like filenames more than 8 characters long and required a TLA.
So what?

It has become a standard in windows to use spaces in filenames no doubt due
to the hopeless command line interface so there was no necessity to play nice
with command line tools that by default use whitespace as field seperators.

>had it on my wall at work. Not because I was an Apple fanboi, far
>from it, but my disdain for all things MS Windows was already
>entrenched, and I was being forced to use it for the not-the-actual-work
>part of my job (which was on Sun workstations).

Apple arguably are even worse right now - case aware but case insensitive
unless you use wildcards on the command line when suddenly case matters again.
Hopeless.

eg:
fenris$ touch HELLO
fenris$ ls hello
hello
fenris$ ls H*
HELLO
fenris$ ls h*
ls: h*: No such file or directory

Lew Pitcher <lew.pitcher@digitalfreehold.ca> writes:
> On Wed, 16 Aug 2023 17:11:28 +0300, Phil Carmody wrote:
>> Lew Pitcher <lew.pitcher@digitalfreehold.ca> writes:
>>> On Tue, 15 Aug 2023 23:24:28 +0300, Phil Carmody wrote:
> [snip]
>>>>> I guess the stdin convention '-' is decades old.
>>>>> It was chosen because they had pick something that
>>>>> is short and not likely to exist as a real file.
>>>>
>>>> I propose that '|' might have been better - it's even mnemonic.
>>>
>>> But, '|' conflicts with the shell pipe character, and would need
>>> a workaround in some cases. For example,
>>
>> It's doing something special, why wouldn't you expect to do something
>> special to make use of it?
>
> Okay, then. How about a counter proposal:
> For Unix shells, the < character indicates redirection of stdin.
> I propose that '<' might have been a better choice as a standard
> program option to flag "read from stdin" than either '|' /or/ '-'
> were. :-)

Why *better* than '|', rather than *equally good as*, given that the
logic is absolutely equivalent?

Phil
--
We are no longer hunters and nomads. No longer awed and frightened, as we have
gained some understanding of the world in which we live. As such, we can cast
aside childish remnants from the dawn of our civilization.
-- NotSanguine on SoylentNews, after Eugen Weber in /The Western Tradition/

scott@slp53.sl.home (Scott Lurndal) writes:
> Richard Kettlewell <invalid@invalid.invalid> writes:
>>David Brown <david.brown@hesbynett.no> writes:
>>> On 15/08/2023 09:50, Richard Kettlewell wrote:
>>>> The cases I’ve encountered are where someone accidentally creates a
>>>> file with a weird name (starts with - or whatever) and then has
>>>> trouble deleting it. I don’t think they got as far as having
>>>> accidents with wildcards in the cases I remember, but the basic setup
>>>> of a file with a ridiculous name and a non-expert user does happen
>>>> occasionally.
>>>
>>> It's usually sufficient just to put the awkward name inside quotation marks.
>>
>>In the case we started with, quoting is not sufficient.
>
> Doesn't that depend on the type of quote? glob characters (*, ?) are
> treated as regular characters in single quotes, for example.

No, it depends on the program you intend to pass the filename to, and
how it treats its arguments.

Which we don't know. So we can't say anything more at the moment.

Phil
--
We are no longer hunters and nomads. No longer awed and frightened, as we have
gained some understanding of the world in which we live. As such, we can cast
aside childish remnants from the dawn of our civilization.
-- NotSanguine on SoylentNews, after Eugen Weber in /The Western Tradition/

Muttley@dastardlyhq.com writes:
> On Tue, 15 Aug 2023 23:32:22 +0300
> Phil Carmody <pc+usenet@asdf.org> wrote:
>>Muttley@dastardlyhq.com writes:
>>> On Tue, 15 Aug 2023 16:12:19 -0000 (UTC)
>>> kalevi@kolttonen.fi (Kalevi Kolttonen) wrote:
>>>>Kenny McCormack <gazelle@shell.xmission.com> wrote:
>>>>> But do they know that???
>>>>
>>>>"They" probably don't know it. But let's face it,
>>>>nobody really wants to create a file having '-'
>>>>filename on purpose. The filename is not descriptive
>>>>at all, it would be just an insane choice for anything
>>>>useful.
>>>
>>> I used to think the same thing about spaces in filenames. Then along came
>>> Windows.
>>
>>Your memory if flawed. Spaces were always allowed in unix
>>filenames. Often used to hide things in plain sight on FTP sites, for
>>example.
>
> Any character from 0-255 is allowed in unix filenames but only an idiot would
> use spaces and non printing characters.
>
>>MS Windows didn't allow spaces until Windows 95. I remember the Apple
>
> DOS didn't like filenames more than 8 characters long and required a TLA.
> So what?
>
> It has become a standard in windows to use spaces in filenames no doubt due
> to the hopeless command line interface so there was no necessity to play nice
> with command line tools that by default use whitespace as field seperators.
>
>>had it on my wall at work. Not because I was an Apple fanboi, far
>>from it, but my disdain for all things MS Windows was already
>>entrenched, and I was being forced to use it for the not-the-actual-work
>>part of my job (which was on Sun workstations).
>
> Apple arguably are even worse right now - case aware but case insensitive
> unless you use wildcards on the command line when suddenly case matters again.
> Hopeless.
>
> eg:
> fenris$ touch HELLO
> fenris$ ls hello
> hello
> fenris$ ls H*
> HELLO
> fenris$ ls h*
> ls: h*: No such file or directory

That I didn't know. Wow - if that's an actual interactive session being
quoted, that's frightening. Thanks for reminding me why my installation
of linux on my Apple G5 box wasn't moment too soon. How can people live
with such wrongthink? Stockholm Syndrome?

Mind geboggled,
Phil
--
We are no longer hunters and nomads. No longer awed and frightened, as we have
gained some understanding of the world in which we live. As such, we can cast
aside childish remnants from the dawn of our civilization.
-- NotSanguine on SoylentNews, after Eugen Weber in /The Western Tradition/

Keith Thompson wrote:

> If you have a file named -r, you can delete it with `rm ./-r`, or
> `rm -- -r` if your rm implementation supports that (POSIX doesn't
> specify it).

POSIX does specify that "rm -- -r" removes a file called -r.

Almost all the utilities have a statement under OPTIONS that they
conform to XBD 12.2 Utility Syntax Guidelines (in some cases with
exceptions for specific individual guidelines). The requirement to
support "--" is guideline 10.

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

Phil Carmody wrote:

> Muttley@dastardlyhq.com writes:
>>
>> Apple arguably are even worse right now - case aware but case insensitive
>> unless you use wildcards on the command line when suddenly case matters again.
>> Hopeless.
>>
>> eg:
>> fenris$ touch HELLO
>> fenris$ ls hello
>> hello
>> fenris$ ls H*
>> HELLO
>> fenris$ ls h*
>> ls: h*: No such file or directory
>
> That I didn't know. Wow - if that's an actual interactive session being
> quoted, that's frightening. Thanks for reminding me why my installation
> of linux on my Apple G5 box wasn't moment too soon. How can people live
> with such wrongthink?

They don't have to: they can configure the filesystem that contains
their home directory to be properly case sensitive.

I have access to an ancient MacBook which is configured that way:

$ uname -sr
Darwin 15.4.0
$ touch HELLO
$ ls hello
ls: hello: No such file or directory
$ ls H*
HELLO
$ ls h*
ls: h*: No such file or directory

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

On 2023-08-17, Muttley@dastardlyhq.com <Muttley@dastardlyhq.com> wrote:
> Any character from 0-255 is allowed in unix filenames but only an idiot would
> use spaces and non printing characters.

That's simply false. 0 is the null terminator and is not actually part
of the name. The path-component-separating slash cannot be contained in
a path component; there is no escape mechanism to include it.

POSIX defines a set of characters which are recommended for use in file
names for portability; it is wise for applications and users to stick to
that.

--
TXR Programming Language: http://nongnu.org/txr
Cygnal: Cygwin Native Application Library: http://kylheku.com/cygnal
Mastodon: @Kazinator@mstdn.ca

In article <87ttsxzxs7.fsf@fatphil.org>,
Phil Carmody <pc+usenet@asdf.org> wrote:
....
>> ls: h*: No such file or directory
>
>That I didn't know. Wow - if that's an actual interactive session being
>quoted, that's frightening. Thanks for reminding me why my installation
>of linux on my Apple G5 box wasn't moment too soon. How can people live
>with such wrongthink? Stockholm Syndrome?

Kinda hate to disagree with ya there, but running Linux on Apple hardware
is like buying a BMW, ripping out the engine, and changing it into a go-kart.

Now, before you think I'm some Apple Fanboy, let me assure you that just
the opposite is true. I run all Linux, but I consider it wrong to pay more
than $100 for the box upon which I run Linux. That's why I'm all Raspberry
Pi nowadays.

But if you've paid Apple's exhorbitant fee to buy their hardware, then you
might as well be running their OS. I don't want to have to dot all the I's
and cross all the T's on this, but I assume you get what I mean. If you
like Linux, you should not be paying Apple prices for your hardware.

Or, to put it another way, Apple stuff (hardware and software) is for
people who like that sort of thing. And they do indeed like it.

By the way, the car analogy is not accidental. I believe that the more you
pay for a car, the more problems you will have with it. You are literally
paying for the problems. And again, there are people who like that. So,
again, paying lots for a car is for people who like that sort of thing.
And they do indeed like it.

--
The randomly chosen signature file that would have appeared here is more than 4
lines long. As such, it violates one or more Usenet RFCs. In order to remain
in compliance with said RFCs, the actual sig can be found at the following URL:
http://user.xmission.com/~gazelle/Sigs/IceCream

In comp.unix.programmer Muttley@dastardlyhq.com wrote:
> They both cause problems.

So do nuclear bombs and fentynyl, yet they have
not enough in common to make a sensible comparison
between them.

Similar reasoning applies to filenames with spaces
and the '-' stdin convention. Many people like
the former and are justified in doing so, but
nobody has the desire to use '-' as a filename.

I am tired of making this obvious point over and
over, so if it does not sink in now, I have to
give up.

br,
KK

On Thu, 17 Aug 2023 14:57:44 +0300
Phil Carmody <pc+usenet@asdf.org> wrote:
>Muttley@dastardlyhq.com writes:
>> Apple arguably are even worse right now - case aware but case insensitive
>> unless you use wildcards on the command line when suddenly case matters
>again.
>> Hopeless.
>>
>> eg:
>> fenris$ touch HELLO
>> fenris$ ls hello
>> hello
>> fenris$ ls H*
>> HELLO
>> fenris$ ls h*
>> ls: h*: No such file or directory
>
>That I didn't know. Wow - if that's an actual interactive session being

It is, its a cut and paste from the terminal.

>quoted, that's frightening. Thanks for reminding me why my installation
>of linux on my Apple G5 box wasn't moment too soon. How can people live
>with such wrongthink? Stockholm Syndrome?

I bought a Mac because I wanted a unix box that I could just switch on and
get coding rather than mess about installing some linux distro and hope the
bits all work. Plus MacOS has a bonus of lots of high quality apps on the
App Store. Though tbh if Solaris workstations were still a thing I'd have bought
one of them instead.

On Thu, 17 Aug 2023 13:49:49 +0100
Geoff Clare <geoff@clare.See-My-Signature.invalid> wrote:
>Phil Carmody wrote:
>
>> Muttley@dastardlyhq.com writes:
>>>
>>> Apple arguably are even worse right now - case aware but case insensitive
>>> unless you use wildcards on the command line when suddenly case matters
>again.
>>> Hopeless.
>>>
>>> eg:
>>> fenris$ touch HELLO
>>> fenris$ ls hello
>>> hello
>>> fenris$ ls H*
>>> HELLO
>>> fenris$ ls h*
>>> ls: h*: No such file or directory
>>
>> That I didn't know. Wow - if that's an actual interactive session being
>> quoted, that's frightening. Thanks for reminding me why my installation
>> of linux on my Apple G5 box wasn't moment too soon. How can people live
>> with such wrongthink?
>
>They don't have to: they can configure the filesystem that contains
>their home directory to be properly case sensitive.

You can, but it can break a lot of programs that expect case insensitivity.

On Thu, 17 Aug 2023 13:52:23 -0000 (UTC)
Kaz Kylheku <864-117-4973@kylheku.com> wrote:
>On 2023-08-17, Muttley@dastardlyhq.com <Muttley@dastardlyhq.com> wrote:
>> Any character from 0-255 is allowed in unix filenames but only an idiot would
>
>> use spaces and non printing characters.
>
>That's simply false. 0 is the null terminator and is not actually part
>of the name. The path-component-separating slash cannot be contained in
>a path component; there is no escape mechanism to include it.

Ok, there are a couple of exceptions. Point still stands.

>POSIX defines a set of characters which are recommended for use in file
>names for portability; it is wise for applications and users to stick to
>that.

But they don't have to and when UTF8 filenames get used things can get messy
on non UTF8 terminals.

kalevi@kolttonen.fi (Kalevi Kolttonen) writes:
>In comp.unix.programmer Muttley@dastardlyhq.com wrote:
>> They both cause problems.
>
>So do nuclear bombs and fentynyl, yet they have
>not enough in common to make a sensible comparison
>between them.
>
>Similar reasoning applies to filenames with spaces
>and the '-' stdin convention. Many people like
>the former and are justified in doing so, but
>nobody has the desire to use '-' as a filename.
>
>I am tired of making this obvious point over and
>over, so if it does not sink in now, I have to
>give up.

It's like arguments about how many angels can dance
on the point of a pin - the use of a single hyphen
to denote stdin will _never_ change, it is pointless
to assume otherwise.

Kaz Kylheku <864-117-4973@kylheku.com> writes:
> On 2023-08-17, Muttley@dastardlyhq.com <Muttley@dastardlyhq.com> wrote:
>> Any character from 0-255 is allowed in unix filenames but only an idiot would
>> use spaces and non printing characters.
>
> That's simply false. 0 is the null terminator and is not actually part
> of the name. The path-component-separating slash cannot be contained in
> a path component; there is no escape mechanism to include it.
>
> POSIX defines a set of characters which are recommended for use in file
> names for portability; it is wise for applications and users to stick to
> that.

https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap03.html#tag_03_282

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
a b c d e f g h i j k l m n o p q r s t u v w x y z
0 1 2 3 4 5 6 7 8 9 . _ -

--
Keith Thompson (The_Other_Keith) Keith.S.Thompson+u@gmail.com
Will write code for food.
void Void(void) { Void(); } /* The recursive call of the void */

Muttley@dastardlyhq.com writes:

> On Tue, 15 Aug 2023 23:32:22 +0300
> Phil Carmody <pc+usenet@asdf.org> wrote:
>>Muttley@dastardlyhq.com writes:
>>> On Tue, 15 Aug 2023 16:12:19 -0000 (UTC)
>>> kalevi@kolttonen.fi (Kalevi Kolttonen) wrote:
>>>>Kenny McCormack <gazelle@shell.xmission.com> wrote:
>>>>> But do they know that???
>>>>
>>>>"They" probably don't know it. But let's face it,
>>>>nobody really wants to create a file having '-'
>>>>filename on purpose. The filename is not descriptive
>>>>at all, it would be just an insane choice for anything
>>>>useful.
>>>
>>> I used to think the same thing about spaces in filenames. Then along came
>>> Windows.
>>
>>Your memory if flawed. Spaces were always allowed in unix
>>filenames. Often used to hide things in plain sight on FTP sites, for
>>example.
>
> Any character from 0-255 is allowed in unix filenames but only an idiot would
> use spaces and non printing characters.

To some extent it depends on the file system, and to some extent it's
down to the kernel, but I don't know of any Unix/FS combination that
permits 0 (null) bytes in file names. I can't see how it could work.

And almost all file systems will prohibit the use of / in a file name,
though I've heard you can trick NFS into taking it.

Some file systems (depending on mount options) can have complex rules
such as rejecting any name with an invalid UTF-8 sequence.

--
Ben.

Geoff Clare <geoff@clare.See-My-Signature.invalid> writes:
> Keith Thompson wrote:
>
>> If you have a file named -r, you can delete it with `rm ./-r`, or
>> `rm -- -r` if your rm implementation supports that (POSIX doesn't
>> specify it).
>
> POSIX does specify that "rm -- -r" removes a file called -r.
>
> Almost all the utilities have a statement under OPTIONS that they
> conform to XBD 12.2 Utility Syntax Guidelines (in some cases with
> exceptions for specific individual guidelines). The requirement to
> support "--" is guideline 10.

Quite right, I sit corrected.

https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap12.html#tag_12_02

--
Keith Thompson (The_Other_Keith) Keith.S.Thompson+u@gmail.com
Will write code for food.
void Void(void) { Void(); } /* The recursive call of the void */

gazelle@shell.xmission.com (Kenny McCormack) writes:
> In article <87ttsxzxs7.fsf@fatphil.org>,
> Phil Carmody <pc+usenet@asdf.org> wrote:
> ...
>>> ls: h*: No such file or directory
>>
>>That I didn't know. Wow - if that's an actual interactive session being
>>quoted, that's frightening. Thanks for reminding me why my installation
>>of linux on my Apple G5 box wasn't moment too soon. How can people live
>>with such wrongthink? Stockholm Syndrome?
>
> Kinda hate to disagree with ya there, but running Linux on Apple hardware
> is like buying a BMW, ripping out the engine, and changing it into a go-kart.
>
> Now, before you think I'm some Apple Fanboy, let me assure you that just
> the opposite is true. I run all Linux, but I consider it wrong to pay more
> than $100 for the box upon which I run Linux. That's why I'm all Raspberry
> Pi nowadays.
>
> But if you've paid Apple's exhorbitant fee to buy their hardware, then you
> might as well be running their OS.

Hahahah. Nah, never paid a penny. It was a "sponsorship" from Apple.
Top of the range kit at the time too. All I had to do was write fast
code and not insult them in public. Managed for several years.

Oh - my code ran faster under linux than OSX. The apple compiler seemed
to never want to use all 32 of the FP registers (and, more bizarrely,
the register it would mysteriously forget existed wasn't always the same
one). I later worked for Freescale and had access to their internal
tools, and could see how good or bad a job the various compilers were at
keeping the pipelines full, and it was clear that Apple weren't as smart
as they thought they were.

Phil
--
We are no longer hunters and nomads. No longer awed and frightened, as we have
gained some understanding of the world in which we live. As such, we can cast
aside childish remnants from the dawn of our civilization.
-- NotSanguine on SoylentNews, after Eugen Weber in /The Western Tradition/

On 17/08/2023 16:09, Kenny McCormack wrote:
> In article <87ttsxzxs7.fsf@fatphil.org>,
> Phil Carmody <pc+usenet@asdf.org> wrote:
> ...
>>> ls: h*: No such file or directory
>>
>> That I didn't know. Wow - if that's an actual interactive session being
>> quoted, that's frightening. Thanks for reminding me why my installation
>> of linux on my Apple G5 box wasn't moment too soon. How can people live
>> with such wrongthink? Stockholm Syndrome?
>
> Kinda hate to disagree with ya there, but running Linux on Apple hardware
> is like buying a BMW, ripping out the engine, and changing it into a go-kart.
>
> Now, before you think I'm some Apple Fanboy, let me assure you that just
> the opposite is true. I run all Linux, but I consider it wrong to pay more
> than $100 for the box upon which I run Linux. That's why I'm all Raspberry
> Pi nowadays.
>
> But if you've paid Apple's exhorbitant fee to buy their hardware, then you
> might as well be running their OS. I don't want to have to dot all the I's
> and cross all the T's on this, but I assume you get what I mean. If you
> like Linux, you should not be paying Apple prices for your hardware.
>
> Or, to put it another way, Apple stuff (hardware and software) is for
> people who like that sort of thing. And they do indeed like it.
>

I disagree completely.

Apple made good quality hardware, with a style that many people like.
(I personally do not, and have never owned anything Apple.) Sure, you
pay a premium for it, but that's the case for a lot of hardware - if you
draw a graph of processor speed against processor price, you'll easily
see a point where you pay a lot of extra cash for a very small increase
in price. Yet for some people, that large increase is worth it.

So your Apple hardware might be twice the price for a 10% thinner and
lighter laptop. But if that's what you want, go for it - just as it is
fine to want far lower prices and accept slower or weaker hardware.

Then if someone decides that they want the Apple hardware, they've got
the money, they have no moral qualms about paying the high markup Apple
charges (some people have strong opinions about that kind of thing),
that's fine - they've bought the Apple hardware.

And if they decide that they prefer Linux (or Windows) to MacOS, then
installing their preferred system makes the system better - it increases
the value to them. It would be absurd to use an OS that is inferior (at
least in that person's view) just because they had paid for it! You
would not expect a Linux fan to use the Windows installation that comes
with most off-the-shelf PC's - why would you expect them to use MacOS ?

Case in point - Linus Torvalds used a MacBook running Linux as his main
PC for a while, because the hardware (the screen in particular) was
better than alternative laptops.

On 17/08/2023 15:52, Kaz Kylheku wrote:
> On 2023-08-17, Muttley@dastardlyhq.com <Muttley@dastardlyhq.com> wrote:
>> Any character from 0-255 is allowed in unix filenames but only an idiot would
>> use spaces and non printing characters.

Only an idiot would think one rule fits all use-cases.

>
> That's simply false. 0 is the null terminator and is not actually part
> of the name. The path-component-separating slash cannot be contained in
> a path component; there is no escape mechanism to include it.
>
> POSIX defines a set of characters which are recommended for use in file
> names for portability; it is wise for applications and users to stick to
> that.
>

It is even wiser to use characters that are appropriate for the task in
hand.

If you are making software that will be shared amongst a wide variety of
systems, stick to ASCII letters, digits, and underscore - that will work
on everything, including Windows.

If you are writing a document in Thai that will be read by Thai speakers
on Thai computers, name the document in Thai using Thai script.

It's not hard to handle "complicated" filenames from the command line,
outside a few pathological cases (like "-"). Sensible filenames are
fine, even with spaces, brackets, or non-ASCII letters. If those
features are helpful to the main use of the names for the files in
question, use them. If they are unhelpful, don't use them.

Thus it makes sense to avoid inconvenient characters in the names and
paths of programs, and you'll very rarely want them in program source
code names or directories. But they are common and useful for things
like document filenames.

Ben Bacarisse <ben.usenet@bsb.me.uk> writes:
> Some file systems (depending on mount options) can have complex rules
> such as rejecting any name with an invalid UTF-8 sequence.

Oooh, now I'm tempted to name files in Latin-1, to deliberately break
filesystems that make such assumptions!

Phil
--
We are no longer hunters and nomads. No longer awed and frightened, as we have
gained some understanding of the world in which we live. As such, we can cast
aside childish remnants from the dawn of our civilization.
-- NotSanguine on SoylentNews, after Eugen Weber in /The Western Tradition/

On 8/19/23 1:33 PM, Phil Carmody wrote:
> Ben Bacarisse <ben.usenet@bsb.me.uk> writes:
>> Some file systems (depending on mount options) can have complex rules
>> such as rejecting any name with an invalid UTF-8 sequence.
>
> Oooh, now I'm tempted to name files in Latin-1, to deliberately break
> filesystems that make such assumptions!
Be careful with that attitude. The "memory-safe" sudo replacement
doesn't support non-UTF-8 input either:

https://github.com/memorysafety/sudo-rs/issues/213

On Sat, 19 Aug 2023 16:15:32 +0300
=?UTF-8?B?T8SfdXo=?= <oguzismailuysal@gmail.com> wrote:
>On 8/19/23 1:33 PM, Phil Carmody wrote:
>> Ben Bacarisse <ben.usenet@bsb.me.uk> writes:
>>> Some file systems (depending on mount options) can have complex rules
>>> such as rejecting any name with an invalid UTF-8 sequence.
>>
>> Oooh, now I'm tempted to name files in Latin-1, to deliberately break
>> filesystems that make such assumptions!
>Be careful with that attitude. The "memory-safe" sudo replacement
>doesn't support non-UTF-8 input either:

Whats memory unsafe about sudo?

Phil Carmody <pc+usenet@asdf.org> writes:

> Ben Bacarisse <ben.usenet@bsb.me.uk> writes:
>> Some file systems (depending on mount options) can have complex rules
>> such as rejecting any name with an invalid UTF-8 sequence.
>
> Oooh, now I'm tempted to name files in Latin-1, to deliberately break
> filesystems that make such assumptions!

Why?

--
Ben.

In comp.unix.programmer Ben Bacarisse <ben.usenet@bsb.me.uk> wrote:
> Phil Carmody <pc+usenet@asdf.org> writes:
>
>> Ben Bacarisse <ben.usenet@bsb.me.uk> writes:
>>> Some file systems (depending on mount options) can have complex rules
>>> such as rejecting any name with an invalid UTF-8 sequence.
>>
>> Oooh, now I'm tempted to name files in Latin-1, to deliberately break
>> filesystems that make such assumptions!
>
> Why?

I suppose that was some kind of a joke.

Anyway, one cannot "break" those filesystems as the
invalid names are simply rejected. I am not 100%
sure but ZFS might be an instance of such a filesystem.

br,
KK

Muttley@dastardlyhq.com writes:
>On Sat, 19 Aug 2023 16:15:32 +0300
>=?UTF-8?B?T8SfdXo=?= <oguzismailuysal@gmail.com> wrote:
>>On 8/19/23 1:33 PM, Phil Carmody wrote:
>>> Ben Bacarisse <ben.usenet@bsb.me.uk> writes:
>>>> Some file systems (depending on mount options) can have complex rules
>>>> such as rejecting any name with an invalid UTF-8 sequence.
>>>
>>> Oooh, now I'm tempted to name files in Latin-1, to deliberately break
>>> filesystems that make such assumptions!
>>Be careful with that attitude. The "memory-safe" sudo replacement
>>doesn't support non-UTF-8 input either:
>
>Whats memory unsafe about sudo?
>
https://www.cisa.gov/news-events/alerts/2021/02/02/sudo-heap-based-buffer-overflow-vulnerability-cve-2021-3156