On Mon, Sep 30, 2013 at 09:59:30AM -0500, Felipe Contreras wrote: >I switched away from offlineimap when I got a synchronization issue >and had to re-download all my mail, and it took forever. Then mbsync >was much faster, but today offlineimap seems quite fast, maybe even >faster.
I am being just curious now (-: When was the last time you compared offlineimap and mbsync? Do you remember rough numbers? My story is: I started off using offlineimap and used it for around half a year until I discovered mbsync. I was very impressed by the design of mbsync, its speed and that it was in C. I started migrating away from offlineimap and one morning made a simple comparison on a "cold" mailbox: mbsync was 5 times faster than offlineimap¹. I don't recall exact configuration of my mailbox, but it was something about 30 maildirs, 500 e-mails/day, local network IMAP server. Almost a year on mbsync now, very happy (-: Wonder how much offlineimap improved since that time. But doubt I'd switch back. > >Either way, even if it was slower, and buggy, it is more useful, and >that's what really matters. > >There's no point in using mbsync when I cannot use most of my Gmail >labels because they have too many messages, and to forever have the >unread flag out of sync. I'd really like to understand what is the underlying problem and why offlineimap can handle it better (-: You are saying you experience a slowness when pulling message statuses from Gmail back to local Maildir, correct? Essentially it is a "FETCH 1:N (UID FLAGS)"-like IMAP query, where N is a highest known message ID that results in multiple rename(2) syscalls. (Unless I'm missing something.) Now via a Gmail web interface I mark all my 7870 read messages in a test folder as unread and run mbsync --verbose. It took 3.323s on my 2 years old HDD laptop connected to Internet via a home WiFi router. When you are saying mbsync takes way long and you have too many messages, what exact numbers do you mean? ¹ https://twitter.com/mvuets/status/273332162890190848 ------------------------------------------------------------------------------ October Webinars: Code for Performance Free Intel webinars can help you accelerate application performance. Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from the latest Intel processors and coprocessors. See abstracts and register > http://pubads.g.doubleclick.net/gampad/clk?id=60134791&iu=/4140/ostg.clktrk _______________________________________________ isync-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/isync-devel
