managed mounts. hindsight sadness at their departure

managed mounts. hindsight sadness at their departure

[Shaddy Baddah]

Hi,

I apologise if this is inappropriate for this list, or even at all, but
I'd like to record a little bit of sadness that managed mounts are no
longer available.

I was in agreement with the rationale behind deprecating them, partly
described in the 1.7 "what's new" section:

<snip>
- File names are case sensitive if the OS and the underlying file system
   supports it.  Works on NTFS and NFS.  Does not work on FAT and Samba
   shares.  Requires to change a registry key (see the user's guide).
   Can be switched off on a per-mount base.

- Due to the above changes, managed mounts have been removed.
</snip>

I would add that also the unicode support was reason to no longer
persist with managed mounts.

But now in hindsight, one feature of the managed mounts that I counted
as not significant enough to keep it included, seems to me to have a
kind of valid use case. That is allowing multiple files of the same name
differing only in case.

This feature of the managed mount would have provided a handy workaround
to a minor issue with cloning the Linux kernel through Cygwin. For
whatever reason, the kernel source includes certain duplications of
filenames (that is, if you discount their case). This leads to the
ignorable, but irritating assertion of a git status that files are
modified. This because they have been incorrectly compared with a
different file, of the same name (but for case), that has been checked
out over the original:

shaddy@***-w7 ~/workarea/linux-2.6
$ git status
# On branch wireless-testing
# Changed but not updated:
#   (use "git add <file>..." to update what will be committed)
#   (use "git checkout -- <file>..." to discard changes in working 
directory)
#
#       modified:   Documentation/IO-mapping.txt
#       modified:   include/linux/netfilter/xt_CONNMARK.h
#       modified:   include/linux/netfilter/xt_DSCP.h
#       modified:   include/linux/netfilter/xt_MARK.h
#       modified:   include/linux/netfilter/xt_RATEEST.h
#       modified:   include/linux/netfilter/xt_TCPMSS.h
#       modified:   include/linux/netfilter_ipv4/ipt_CONNMARK.h
#       modified:   include/linux/netfilter_ipv4/ipt_DSCP.h
#       modified:   include/linux/netfilter_ipv4/ipt_ECN.h
#       modified:   include/linux/netfilter_ipv4/ipt_MARK.h
#       modified:   include/linux/netfilter_ipv4/ipt_TCPMSS.h
#       modified:   include/linux/netfilter_ipv4/ipt_TOS.h
#       modified:   include/linux/netfilter_ipv4/ipt_TTL.h
#       modified:   include/linux/netfilter_ipv6/ip6t_HL.h
#       modified:   include/linux/netfilter_ipv6/ip6t_MARK.h
#       modified:   net/ipv4/netfilter/ipt_ECN.c
#       modified:   net/ipv4/netfilter/ipt_TTL.c
#       modified:   net/ipv6/netfilter/ip6t_HL.c
#       modified:   net/netfilter/xt_CONNMARK.c
#       modified:   net/netfilter/xt_DSCP.c
#       modified:   net/netfilter/xt_MARK.c
#       modified:   net/netfilter/xt_RATEEST.c
#       modified:   net/netfilter/xt_TCPMSS.c
#
no changes added to commit (use "git add" and/or "git commit -a")

Not that I'd assert that this is reason enough to have kept the managed
mounts, it is just an unfortunate limitation that 1.7 now has over 1.5.

If I get the time one day to understand why these duplicates exists,
I'll make a patch and suggest it to the Linux kernel project (I've
dipped my toe in there a while back).

Regards,
Shaddy

Re: managed mounts. hindsight sadness at their departure

[Eric Blake]

[-- Attachment #1: Type: text/plain, Size: 862 bytes --]

According to Shaddy Baddah on 3/4/2010 8:58 PM:
> <snip>
> - File names are case sensitive if the OS and the underlying file system
>   supports it.  Works on NTFS and NFS.  Does not work on FAT and Samba
>   shares.  Requires to change a registry key (see the user's guide).
> But now in hindsight, one feature of the managed mounts that I counted
> as not significant enough to keep it included, seems to me to have a
> kind of valid use case. That is allowing multiple files of the same name
> differing only in case.

Request case sensitivity from windows, and you have that feature back.

> 
> Not that I'd assert that this is reason enough to have kept the managed
> mounts, it is just an unfortunate limitation that 1.7 now has over 1.5.

No, it is not a limitation, provided you follow the instructions.

$ touch a A
$ ls ?
A  a



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 320 bytes --]

Re: managed mounts. hindsight sadness at their departure

[Shaddy Baddah]

Hi,

On 5/03/2010 4:33 AM, Eric Blake wrote:
> Request case sensitivity from windows, and you have that feature back.

How embarrassing. I recalled fleetingly seeing something on case
sensitivity in the 1.7 user manual, and checked back. But only the
CYGWIN variable section. Seeing that it was restricted to globbing
behaviour, I missed 
http://www.cygwin.com/cygwin-ug-net/using-specialnames.html#pathnames-casesensitive

>> Not that I'd assert that this is reason enough to have kept the managed
>> mounts, it is just an unfortunate limitation that 1.7 now has over 1.5.
>
> No, it is not a limitation, provided you follow the instructions.
>
> $ touch a A
> $ ls ?
> A  a

Many thanks for pointing it out. I was completely unaware that Windows
was capable of case sensitive file naming. I must admit that my
entrenched views of Windows capabilities in this area make me a little
bit apprehensive of possible side-effects to turning this on. In your
experience, are there any?

Of course, I will still try it anyway.

Regards,
Shaddy

Re: managed mounts. hindsight sadness at their departure

[Corinna Vinschen]

On Mar  5 05:02, Shaddy Baddah wrote:
> On 5/03/2010 4:33 AM, Eric Blake wrote:
> >Request case sensitivity from windows, and you have that feature back.
> 
> http://www.cygwin.com/cygwin-ug-net/using-specialnames.html#pathnames-casesensitive
> [...]
> Many thanks for pointing it out. I was completely unaware that Windows
> was capable of case sensitive file naming. I must admit that my
> entrenched views of Windows capabilities in this area make me a little
> bit apprehensive of possible side-effects to turning this on. In your
> experience, are there any?

I'm using it all the time and have no problems, but, still, Win32
applications have no idea what to do with 2 files only differing
by case.  Best you keep directories with such files separate.

> Of course, I will still try it anyway.

That's the spirit!


Corinna

Re: managed mounts. hindsight sadness at their departure

[Shaddy Baddah]

Hi,

On 5/03/2010 9:35 AM, Corinna Vinschen wrote:
> On Mar  5 05:02, Shaddy Baddah wrote:
>> On 5/03/2010 4:33 AM, Eric Blake wrote:
>>> Request case sensitivity from windows, and you have that feature back.
>>
>> http://www.cygwin.com/cygwin-ug-net/using-specialnames.html#pathnames-casesensitive
>> [...]
>> Many thanks for pointing it out. I was completely unaware that Windows
>> was capable of case sensitive file naming. I must admit that my
>> entrenched views of Windows capabilities in this area make me a little
>> bit apprehensive of possible side-effects to turning this on. In your
>> experience, are there any?
>
> I'm using it all the time and have no problems, but, still, Win32
> applications have no idea what to do with 2 files only differing
> by case.  Best you keep directories with such files separate.
>
>> Of course, I will still try it anyway.
>
> That's the spirit!

Of course, this has worked a treat. Thanks again Eric and Corinna and
all the patient cygwin-talk'ers.

Regards,
Shaddy