Regid Ichira wrote: > I had difficulties getting ssh(1)'s ESCAPE CHARACTERS to be recognized > from within a login shell over ssh. In particular, sometimes the escape > character was not recognized as such.
Can you provide an example? I have never had the ssh escape character not be recognized. I often use it. > I was able to find in gmane a similar issue for a Gentoo user from a > few years ago. I don't have that gmane URL handy. > Do you find the below patch acceptable? > +With a login shell, one might issue a single new line character to prepare > +the correct conditions for ssh to catch the escape character. I find the existing sentence "The escape character must always follow a newline to be interpreted as special." to be concise and descriptive. It will be hard to improve upon it. Your addition here seems to be "muddy". Plus third party phrasing "one might issue" isn't in the same voice as the original. But what is special about a login shell? There isn't anything special about it that I can tell. Therefore I think the statement is wrong. > +If the escape character is cought by the remote application, perhaps > +it is echoed back by a login shell, then it will not affect the > +underline ssh channel. s/cought/caught/ Uhm, what? This is also incorrect. The escape character is consumed by the ssh program and is not transmitted to the remote application. The remote application won't read it. If the remote application does read a tilde then it is only that the tilde was passed through escaped in which case it was not an escape character. > +Pressing the return key twice will close the command line after it > +was started from a login shell. Second, uhm, what? Please say a few more words about what you are talking about here. I cannot understand it and it does not match how ssh operates. I think there must be some misunderstanding happening. Please describe what you are seeing so that we can help work through it. Bob
signature.asc
Description: Digital signature
