On Tue, Sep 9, 2008 at 3:11 PM, Theodore Tso <tytso@xxxxxxx> wrote: > On Tue, Sep 09, 2008 at 01:53:40PM -0600, Mark wrote: >> >> Anyway, I didn't say that IMAP was a "resource hog" in the grand >> scheme of computing, only in comparison with POP3. >> > > It really depends on how the IMAP/POP3 server was implemented. Some > POP3 servers are implemented optimizing for the case where all of the > mail messages are downloaded, and then deleted; so if you download all > of the messages, but only delete a few, it could be quite expensive if > the POP3 server stores the mailbox as a single file, and deletes > require copying all of the messages which are _not_ deleted. > > In contrast, if your IMAP server is implemented to use some kind of > back-end database, or a Maildir directory, a particular implementation > of IMAP could be more performance than a particular implementation of > POP3 protocol. > > It's also true that IMAP has some extra operations that allow for > partial download of just the header information, for example. If a > POP3 client implements this by downloading *all* of the mail messages > each time you connect, without caching the downloaded mail messages > due to local space reasons, in theory POP3 could be less efficient and > end up impacting the server more. > > On the other hand, IMAP has some optional SEARCH capabilities which > could impact the server; but the server doesn't have to implement said > search extensions. > > So it's a lot more complicated than saying categorically that IMAP > always uses more server resources than POP3, or vice versa. A lot > depends on how the client was implemented, how the server was > implemented, and what sort of user functionality is desired. > > Regards, > > - Ted I've said many times that the protocol is pretty much irrelevent; it's the implementation and features of the software that really matters. > > P.S. Something I very much like is a program called "mbsync", > available at http://isync.sourceforge.net. This will use IMAP4 > download a local copy of all of your mail message to a local laptop or > tablet, in a Maildir directory on local disk (or flash storage). This > allows you to read your e-mail messages while disconnected, without > needing access to the network. After you mark messages as deleted, > the next time you synchronize your local Maildir store with your IMAP4 > mailbox, it will delete any messages marked as deleted, and then > download any new messages. Since the messages aren't deleted on the > IMAP server, if your tablet gets lost or destroyed, the messages > aren't lost, since there is a copy safely on the IMAP server. This sounds like going to a whole lot of work to do something that I've been doing for years with POP3 without any additional server or software... > > Yes, you can in theory do this using POP, Not "in theory", in reality, and easily... > but this particular > implementation only does it for IMAP. It's open source --- people who > want to implement this for POP are free to do so. In general, I find > the attitude of the original requester who seemed to expect that other > people implement software for his own convenience to be a little > annoying. I'm sorry that you find it annoying that somebody asks for basic functionality... > If there's something you want that isn't yet implemented in > open source, either implement it yourself, or gently request that > someone who can implement do so --- or perhaps you can hire them or > otherwise give them some kind of incentive to implement the feature > for you. Open source developers do *not* owe anything to their user > base; they implement feature requests out of the goodness of their > hearts, or because they need the feature as well. > In other words, f*** off! This is exactly the attitude that will always keep open source from gaining significant market share. Only the projects that actually listen to such concerns ever make any real impact (Mozilla, anyone?). > Or, of course, you can use another solution. I still carry around a > Palm Pilot because none of the opensource PIM applications are good > enough yet. And I use a Blackberry because its e-mail solution is > still far superior to what is available on the Palm or N800 platform. > But you don't see me whining about how people *must* make improvements > to the N800's mail client. I hope it will happen, but unfortunately I > don't have time to help improve the N800 mail client. I did however, > help improve the isync project until it was good enough to meet my > needs. That's the way open source works... you scratch your own > itch. > > Bull hockey! The Nokia tablets are sold as consumer devices, but are severely lacking in software functionality. I didn't lay out several hundred in cash just so I can carry yet another device around. The whole point was to replace several devices with one. This amounts to false advertising, and ensures that the the tablets will never take off or make any real impact with consumers, especially since Nokia has been rumbling for some time about orphaning the devices completely and leaving them to their fate with the community. In other words, resigning them to a fate of never being anything more than developers' playthings. Whether installable applications are open source or not is irrelevant. Open source applications exist for pretty much every platform and OS out there. But when the built-in apps don't measure up to consumers' needs or expectations, it's a guarantee of failure. I've got a deal for you: if you want to develop open source (or for that matter, *any*) software and don't want to hear any complaints, then keep it to yourself. Don't make it available for anyone else to download. On the other hand, if you want to create software that other people will actually *want*, then when somebody mentions that it lacks specific functionality that they need, just implement it! Or, if it is something that is truly impossible, explain why. Don't just whine and bitch and moan that they didn't suck up to you in the right way. Put up or shut up! If you want anyone to use your software and are using them for alpha and beta testers, then you *have* to expect some feedback. If you don't want feedback, don't distribute your software. It's that simple! Mark _______________________________________________ maemo-users mailing list maemo-users@xxxxxxxxx https://lists.maemo.org/mailman/listinfo/maemo-users
