Mail Archives: cygwin/2005/12/30/00:08:30
On Thu, Dec 29, 2005 at 11:09:21PM -0500, Larry Hall (Cygwin) wrote:
>Tom Rodman wrote:
>><2 informal tests on my home PC, /etc/ssh_config is "OTTB">
>>
>>no problem, using PubkeyAuthentication:
>>
>> $ uname -a #login shell is bash, started by 'sshing in'
>> CYGWIN_NT-5.0 argon 1.5.19s(0.149/4/2) 20051227 16:45:51 i686 unknown
>> unknown Cygwin
>> $ ssh localhost date
>> Wed Dec 28 21:03:43 CST 2005
>> $ scp adm_tsr AT localhost:/etc/passwd /tmp/toss
>> passwd 100%
>> 704 0.7KB/s 00:00
>> $
>>
>>problems w/same tests using PasswordAuthentication:
>>
>> $ eval $(ssh-agent -k) #same bash session continuing
>> Agent pid 304 killed
>> $ uname -a
>> CYGWIN_NT-5.0 argon 1.5.19s(0.149/4/2) 20051227 16:45:51 i686 unknown
>> unknown Cygwin
>> $ ssh localhost date
>> Enter passphrase for key '/home/adm_tsr/.ssh/id_rsa':
>> adm_tsr AT localhost's password:
>> $ : skipped passphrase prompt, entered password; no date shows up
>> $ scp adm_tsr AT localhost:/etc/passwd /tmp/toss
>> Enter passphrase for key '/home/adm_tsr/.ssh/id_rsa':
>> adm_tsr AT localhost's password:
>> $ : skipped passphrase prompt, notice nothing echoed after password was
>> entered
>>
>>comments? can anyone else duplicate this?
>
>Confirmed. cygwin1-20051216.dll is the last snapshot that worked properly
>for this. Sorry, that's all the input I have for the moment.
Well, certainly if enough people say that 20051216 is the last snapshot that
worked properly eventually Corinna and I will believe them. I don't think we've
reached critical mass yet, though. After all you know how suspicious we are.
We don't really trust a bug report until it's been reported by at least 42
people.
We know, without testing, that this worked properly on B20 but what we
don't know is if it worked in 1.5.17, 1.5.16, etc. So, some
enterprising sould should probably step back in time before 1.5.18 and
provide a graph of operability for every past release. Once we have
that data, why then, it will probably only take a few reinstalls of
cygwin and, worse case scenario, a few movements of my computers in and
out of my office before this problem is fixed.
But, all kidding inside, I guess I really can't complain because I'm
selfishly holding back permission from Igor to debug this. We all know
that no one can debug anything unless they get permission from me first
and there's no way that I would ever want people to debug cygwin or
provide patches to fix problems or anything like that.
cgf
--
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/
- Raw text -