Regarding the -A, I was planning to not support that at all. Since having a shell with autocompletion will be useless if I can't read the tables/columns from a database.
But this has come up more than once in talking to MySQL users. So I'm curious to know the scenario when this is useful. Does it really add that much overhead? Will it be reasonable to do it once when I startup the shell for the first time or when the user switches the database with a 'USE or \u'. On Mon, Apr 20, 2015 at 11:18 AM, Nicholas Leippe <[email protected]> wrote: > yes, \g is same as ;, \G is vertical. I'm guessing they are short for 'go'? > > I think allowing the pager to be modified at runtime is important. I > hate starting up a client, doing something, and then having to stop it > just to change a config file and start it again. Often this isn't > desireable when working on a client's system--you don't want to change > their admin's preferred settings just for your own session. > > Toggling -A from inside the client would be a nice addition also. > > Maybe a stretch of an idea, but you could add a "paranoid anti-paste > protection" toggle-mode that will ignore and warn when it experiences > too high of a data burst to prevent accidental pastes when on a > production login. I know I've accidentally pasted a query into the > wrong shell before. Sure, you could always type it into the wrong > shell, but typing is much slower, and you only risk hitting enter one > query at a time rather than pasting multiple queries with one click > possibly doing far more damage... > > /* > PLUG: http://plug.org, #utah on irc.freenode.net > Unsubscribe: http://plug.org/mailman/options/plug > Don't fear the penguin. > */ > /* PLUG: http://plug.org, #utah on irc.freenode.net Unsubscribe: http://plug.org/mailman/options/plug Don't fear the penguin. */
