CYGWIN / (no)glob documentation bug
Hi,
http://cygwin.com/cygwin-ug-net/using-cygwinenv.html says:
(no)glob[:ignorecase] - if set, command line arguments containing
UNIX-style file wildcard characters (brackets, question mark,
asterisk, escaped with \) are expanded into lists of files that
match those wildcards. This is applicable only to programs running
from a DOS command line prompt. Default is set.
It seems that globbing is applicable not only to programs running
from a DOS command line prompt, but to programs run from any native
Windows (non-Cygwin) application.
FWIW, that behaviour may cause somewhat mysterious problems. For
example, I just had a non-Cygwin git front-end attempting to run
git with command line parameters cat-file, -e and
123456789^{commit} and the result was a failure because of brace
expansion. But I guess there isn't much that can be done about
this (except fix the documentation and have users use noglob) since
the cygwin dll can't really know when globbing is desired and when
it isn't.
Also, braces aren't listed as wildcard characters in the quoted
documentation even though GLOB_BRACE is passed to glob() in the
argument globbing code.
http://www.cygwin.com/faq/faq-nochunks.html#faq.api.globbing should
probably be updated as well.
--
Hannu
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Re: AltGr sometimes generates Control_L
Jon TURNEY jon.tur...@dronecode.org.uk writes: Hannu Koivisto wrote: Hmm... can I see the first few lines of your /var/log/Xwin.0.log, please, so I can see what keyboard layout windows is reporting and X is using. The only keyboard related lines I can see are around line 36 and they are: (--) 16 mouse buttons found (--) Setting autorepeat to delay=250, rate=31 (--) winConfigKeyboard - Layout: 040B (040b) (--) Using preset keyboard for Finnish (40b), type 4 But a bigger problem is that AltGr sometimes generates Control_L in addition to ISO_Level3_Shift, which makes Emacs (which I run from a Linux machine remotely over X) quite unhappy. For example, when I'm trying to enter }, which in Finnish layout is done by AltGr+0, Emacs complains that C-} is undefined. Hmm.. in fact, I think a Ctrl-L is generated for every AltGr keypress by Windows. The X server contains special code to eat these keypresses, but it seems this is not working as reliably as it should be. Interesting. And odd :) It would be useful to know if you have TweakUI installed, as comments in the code indicate and FAQ 5.1.7 indicate that did cause problems with this AltGr handling which are thought to be fixed.. I don't have TweakUI installed. I've seen earlier complaints about the same problem dating back to 2005 but nothing that would help me solve this problem. Links, please I actually simply read the mails using my news reader (I'm reading this list via gmane) and I don't have them around anymore, but gmane searchs finds something: http://search.gmane.org/?query=altgrgroup=gmane.os.cygwin.xfree -- Hannu -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://x.cygwin.com/docs/ FAQ: http://x.cygwin.com/docs/faq/
AltGr sometimes generates Control_L
Greetings, I run Cygwin/X with XWin.exe -multiwindow -clipboard -silent-dup-error. I start xev from shell and start pressing(releasing) a and AltGr keys consecutively: a AltGr a AltGr ... If I'm lucky, pressing and releasing a results to... ---8-- KeyPress event, serial 27, synthetic NO, window 0xe1, root 0x7c, subw 0x0, time 1906663234, (595,1269), root:(771,1501), state 0x10, keycode 38 (keysym 0x61, a), same_screen YES, XLookupString gives 1 bytes: (61) a XmbLookupString gives 1 bytes: (61) a XFilterEvent returns: False KeyRelease event, serial 27, synthetic NO, window 0xe1, root 0x7c, subw 0x0, time 1906663312, (595,1269), root:(771,1501), state 0x10, keycode 38 (keysym 0x61, a), same_screen YES, XLookupString gives 1 bytes: (61) a XFilterEvent returns: False ---8-- ...and pressing and releasing AltGr results to... ---8-- KeyPress event, serial 27, synthetic NO, window 0xe1, root 0x7c, subw 0x0, time 1906710515, (595,1269), root:(771,1501), state 0x10, keycode 113 (keysym 0xfe03, ISO_Level3_Shift), same_screen YES, XLookupString gives 0 bytes: XmbLookupString gives 0 bytes: XFilterEvent returns: False KeyRelease event, serial 27, synthetic NO, window 0xe1, root 0x7c, subw 0x0, time 1906710609, (595,1269), root:(771,1501), state 0x90, keycode 113 (keysym 0xfe03, ISO_Level3_Shift), same_screen YES, XLookupString gives 0 bytes: XFilterEvent returns: False ---8-- Unfortunately, sometimes pressing and releasing a resulted to... ---8-- KeyPress event, serial 27, synthetic NO, window 0xe1, root 0x7c, subw 0x0, time 1906670500, (595,1269), root:(771,1501), state 0x14, keycode 38 (keysym 0x61, a), same_screen YES, XLookupString gives 1 bytes: (01) ☺ XmbLookupString gives 1 bytes: (01) ☺ XFilterEvent returns: False KeyRelease event, serial 27, synthetic NO, window 0xe1, root 0x7c, subw 0x0, time 1906670593, (595,1269), root:(771,1501), state 0x14, keycode 38 (keysym 0x61, a), same_screen YES, XLookupString gives 1 bytes: (01) ☺ XFilterEvent returns: False ---8-- ...where the string looked like a happy face. But a bigger problem is that AltGr sometimes generates Control_L in addition to ISO_Level3_Shift, which makes Emacs (which I run from a Linux machine remotely over X) quite unhappy. For example, when I'm trying to enter }, which in Finnish layout is done by AltGr+0, Emacs complains that C-} is undefined. ---8-- KeyPress event, serial 27, synthetic NO, window 0xe1, root 0x7c, subw 0x0, time 1907120437, (478,827), root:(676,1088), state 0x10, keycode 37 (keysym 0xffe3, Control_L), same_screen YES, XLookupString gives 0 bytes: XmbLookupString gives 0 bytes: XFilterEvent returns: False KeyPress event, serial 27, synthetic NO, window 0xe1, root 0x7c, subw 0x0, time 1907120437, (478,827), root:(676,1088), state 0x14, keycode 113 (keysym 0xfe03, ISO_Level3_Shift), same_screen YES, XLookupString gives 0 bytes: XmbLookupString gives 0 bytes: XFilterEvent returns: False KeyRelease event, serial 27, synthetic NO, window 0xe1, root 0x7c, subw 0x0, time 1907120531, (478,827), root:(676,1088), state 0x94, keycode 37 (keysym 0xffe3, Control_L), same_screen YES, XLookupString gives 0 bytes: XFilterEvent returns: False KeyRelease event, serial 27, synthetic NO, window 0xe1, root 0x7c, subw 0x0, time 1907120531, (478,827), root:(676,1088), state 0x90, keycode 113 (keysym 0xfe03, ISO_Level3_Shift), same_screen YES, XLookupString gives 0 bytes: XFilterEvent returns: False EnterNotify event, serial 27, synthetic NO, window 0xe1, root 0x7c, subw 0x0, time 1907125468, (167,87), root:(365,348), mode NotifyNormal, detail NotifyNonlinear, same_screen YES, focus YES, state 16 ---8-- I've seen earlier complaints about the same problem dating back to 2005 but nothing that would help me solve this problem. I'm using cygwin in Windows XP SP3. Package versions: cygwin 1.5.25-15 libX11_61.1.5-2 xorg-server 1.5.3-7 xorg-x11-base 7.4-1 xorg-x11-bin7.4-1 xorg-x11-bin-dlls 7.4-1 xorg-x11-bin-lndir 7.4-1 xorg-x11-etc7.4-1 xorg-x11-fenc 7.4-1 xorg-x11-fnts 7.4-1 xorg-x11-libs-data 7.4-1 xorg-x11-xwin 7.4-1 -- Hannu -- Unsubscribe
Raw device access to an SD card
Greetings, The goal is to create a bootable Linux installation on an SD card by writing an existing image using dd in Cygwin. The card is inserted to an SD card slot of a laptop and apparently appears as \\.\physicaldrive2 to the OS. Now, dd if=.\\physicaldrive2 reads something (hopefully from that card, hard to say for sure) but an attempt to write with dd if=image of=.\\physicaldrive2 fails with Invalid argument error. Should this work? Is there something that one must do first? I tried mount -f -b //./physicaldrive2 /dev/foo as suggested by someone but according to the Cygwin user's guide that is deprecated and besides, /dev/foo behaved just like .\\physicaldrive2 directly. FWIW, I tried to find if the card device gets mapped to some /dev/sd? but that didn't seem to be the case. Only /dev/sda was available (hard drive). Is it possible to list that kind of mappings somehow? -- Hannu -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: ssh/pubkey authentication and use of subst
Corinna Vinschen [EMAIL PROTECTED] writes: On Oct 30 12:44, Hannu Koivisto wrote: Based on earlier discussions on this list, it's apparently a known problem that when you use public key authentication, you are not authenticated through windows, which means that you cannot map network shares, for example. That's not right. The problem is that you didn't logon using a password and you are running in a foreign logon session. The result is that you have to use explicit identification when connecting to a share. Assuming you are on machine or in domain BRAIN, user name PINKY. When you logged on using password authentication, everything is known to identify and authorize you automatically to a server, so the following works (assuming you *have* permissions to access the share): Ok. $ net use '\\server\share' However, this doesn't work with pubkey authentication because your authorization information is incomplete. Therefore you have to identify and authorize explicitely: $ net use '\\server\share' /user:'BRAIN\PINKY' your-password or $ net use x: '\\server\share' /user:'BRAIN\PINKY' your-password Unfortunately the explicit form doesn't work for me via pubkey authentication either, I get System error 5 has occurred. Access is denied. (return code is 2). Precisely the same command works if I log in using password authentication. Both the client and the server machines run Windows XP SP2, openssh is 4.7p1-2, cygwin 1.5.24-2. sshd was set up with ssh-host-config. I don't need shares, just subst, but I'd be happy to provide more information and test things to help to figure this out. I have no idea why subst fails, though. Must have something to do with the below as well. subst also says Access denied - path (return code is 1). You are running as the user you have logged in as. However, since no Windows authentication took place, you don't get your own logon session. You're running in the logon session of the user running sshd. This situation is wrongly evaluated by Windows, so that functions returning a user name from a SID return the name of the user running sshd. But the application token does *not* grant you the permissions of the user running sshd. The token is still correct and only grant you the rights your user account has. The user and owner SIDs in the token are correctly set to the SID of your own account. Only the Windows functions returning the user name do return the wrong name. Thanks for the explanation. -- Hannu -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
ssh/pubkey authentication and use of subst
Greetings, Based on earlier discussions on this list, it's apparently a known problem that when you use public key authentication, you are not authenticated through windows, which means that you cannot map network shares, for example. I was surprised to find that I cannot use subst either. It works fine if I log in using password authentication, just like mapping of network shares works. Unfortunately I have a situation where I'd really need to combine passwordless ssh logins and use of subst. It would be ok if I needed to give the password once as a separate step and then further use of ssh didn't require giving a password. Are there any workarounds to this problem? I was wondering if fsh (http://www.lysator.liu.se/fsh/) could be of help here: I haven't used it before but apparently with it you create a tunnel once (and give the password) and further logins reuse the existing tunnel. Unfortunately I couldn't find fsh as a Cygwin package, which makes me wonder if there is a technical reason it is not available. If I understood correctly, when you log in using pubkey authentication, you basically are the user sshd runs as. I have set up sshd without privilege separation so that would be the system account. One would expect system account to be able to use subst but if it indeed can't, would it be possible to extend the capabilities of system account to that effect? Any other ideas? -- Hannu -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
openssh - cygcrypto dependency problem
Greetings, I upgraded openssh to 4.7p1-2, hoping to get ssh-add working. I also upgraded all other packages to their latest versions. openssl is 0.9.8e-3 and openssl097 is 0.9.7l-1 (these weren't changed during the upgrade). ssh binary can no longer be executed, ssh whatever does nothing and strace ssh reveals the reason: The procedure entry point EVP_CIPHER_CTX_get_app_data could not be located in the dynamic link library cygcrypto-0.9.8.dll. Apparently cygcrypto is part of openssl of which I have the latest version. Even before openssh upgrade I had the same problem with ssh-add but ssh itself worked fine. Unfortunately I can't remember which version of openssh package I had before upgrade. Downgrading to 4.6p1-1 didn't help. Given that no one else seems to have reported this problem on this list, I'm wondering if I could be doing something wrong here... -- Hannu -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: Sharing zsh history in cygwin.
Andrew Markebo [EMAIL PROTECTED] writes: Anyone managed to share command line history between running zsh's in latest cygwin's? I don't use SHARE_HISTORY (just inc_append_history, hist_ignore_all_dups, hist_no_store and my history sizes are 100 times larger than yours), but... setopt HISTIGNOREDUPS HISTIGNORESPACE EXTENDED_HISTORY setopt INC_APPEND_HISTORY SHARE_HISTORY HISTSIZE=300 SAVEHIST=300 HISTFILE=~/.history The last line, HISTFILE, freezes the cygwin-distributed zsh (4.2.0) just after reading the config-files. Removing just that line ...after some cygwin update (a few months ago) my zshs also started freezing during startup. I tracked the problem down to HISTFILE being assigned a file that resides on a network drive (a Linux samba share in this case, I didn't try anything else). If I changed my history file to reside on a local hard drive, zsh no longer froze. Is your home directory on a local drive? -- Hannu -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
Re: How to make Cygwin map backspace to ASCII DEL?
Igor Pechtchanski [EMAIL PROTECTED] writes: man stty I'm well aware of the stty program. I was going to add please, don't suggest stty to my original mail, but thought it might be somewhat impolite. If you are trying to imply that 'stty erase $(tput kbs)' (where tput kbs outputs ASCII DEL) would do what I want, you must have misread the question. It doesn't and I think it shouldn't. Or is there some Cygwin specific extension in the stty program that does what I want? I don't see anything in stty(1). -- Hannu -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
How to make Cygwin map backspace to ASCII DEL?
Greetings, I just upgraded all Cygwin packages to their latest versions. Although I don't use Cygwin a lot, I'm fairly sure that I would have noticed if backspace in an ordinary Windows console window had ended up as ^H to a Cygwin application (zsh in this case) running in that console window, which is what happens now. Anyway, whatever the old behaviour was, I want backspace to generate ASCII DEL (^?) and not ^H --- how can I make Cygwin do that? -- Hannu -- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Problem reports: http://cygwin.com/problems.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
