>>I installed 5.0 and 5.1-beta on three different system (all are Fedora Core
>>4), and each time when I use the command-line client, I can not use the
>>familiar up-arrow feature that normally lets you scroll through queries
>issued
>>during previous command-line sessions. My history is completely blank.
>>
>>the $HOME/.mysql_history file does exist, but all it's got in it is:
>>
>>_HiStOrY_V2_
>>
>>in the first line.
>>
>>Please help! what am i doing wrong?
>
>to my surprise, i just discovered that connecting to the server over remote
>IP, with a command-line client on the remote machine, the command-line history
>DOES get preserved.
>
>But, that, of course, is due to the separate functionality of server and
>client (as I am now clued in by this), so that .mysql_history is a client-only
>thing, and, in the case of remoteness, it's the client machine that writes to
>its own local .mysql_history file.
>
>So, why is my client failing to write to .mysql_history?
>
>This appears to be an "out-of-the-box", default setting, or am I wrong?
>
>Does it work for anyone out there?
>
>Please help!
After giving up on this a few weeks ago (time scarcity) i returned to it today
and discovered this:
http://lists.mysql.com/commits/2515
It seems my problems stemmed from the above bug. If you build mysql yourself
without the ./configure option "--without-readline", which is what i had been
doing, your version will fail to copy stuff from .mysql_history.TMP to
.mysql_history after a session, in effect, leaving the history blank.
Currently there appear to be two solutions:
1) give ./configure the "--without-readline" option (not sure what the
tradeoffs are here, but there's no metion of any harm in he above post.)
2) apply msvensson's patch (see above link)
I tried applying the patch, but it failed for me; however, #1 restored my
command-line history functionality.
The bug exists in both 5.0 and 5.1.7-beta versions.
I am assuming the patch should work for both, since there is no diff-erence
between the files /<version-source-root>/cmd-line-utils/readline/readline.c ,
which is the file that needs to be patched.
I am surprised no one responded about this. Is there not anyone else out there
besides me doing their own build !? Surely they should have noticed
non-functioning command history.
I understand that the moral of the story is RTFBR (read the fine bug report),
but, still...
andrew
--
MySQL General Mailing List
For list archives: http://lists.mysql.com/mysql
To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
