Re: Re: [RFC] Introduce session.lock, session.lazy_write and session.lazy_destory

On Sat, Jan 4, 2014 at 8:10 AM, Yasuo Ohgaki <yohgaki@ohgaki.net> wrote: > On Fri, Jan 3, 2014 at 11:23 AM, Rowan Collins <rowan.collins@gmail.com>wrote: > >> On 02/01/2014 20:46, Yasuo Ohgaki wrote: >> >>> I've added session_discard() to RFC that discards changes made to >>> session data and close session. >>> >>> Since there is session_commit(), >>> >>> >>> http://www.php.net/manual/en/function.session-commit.php >>> >>> session_abort() may be better name for it. >>> Any comment? >>> >> >> Since session_commit() is an alias for session_write_close(), the new >> function should probably also have "close" in its name (although oddly we >> "start" a session, rather than "open" it). >> >> Both "discard" and "abort" also sound a bit like destroying the session >> itself to me, rather than just ignoring the current changes to it. >> >> Perhaps something like session_revert_close() - "revert" being a closely >> analagous action in version control systems. > > > session_revert_close() sounds reasonable if session_write_close() is main > function. > Rather than session_start(), session_open() would be better name, since > there is session_write_close(). > > IIRC, session_write_close() was considered odd name and session_commit() > is made as alias of it later. Expect-able/consistent function names may be > better as main function and document them for new release. We should be > careful not to have too many aliases, though. > > Main functions > > session_start() (or session_open()?) > session_write_close() > session_revert_close() > > Aliases > > session_open() (or session_start()?) > session_commit() > session_abort() > > Introducing session_revert_close() only seems good to me, but I don't mind > to have session_open()/session_abort() at all. Any comments? > Choosing good names is difficult ;-) > BTW, I'm not the one who named session_write_close()/session_commit(). Regards, -- Yasuo Ohgaki yohgaki@ohgaki.net