At Wed, 18 Jan 2012 14:16:08 -0500, Ludovic Marcotte wrote: > On 18/01/12 01:39, André Schild wrote: > > There already exists a sogo backend for Z-Push, a few posts ago is a > > link for download. > > I will try to make it better available, so we can work on improving it. > We could also import it on our source repository. We did this a while > ago from the connector developed by Philipp Kewisch, but he soon after > abandoned its development.
One of the problems with the existing SOGo backend is that it uses an old version of Z-Push that has a different license. The old version was GPLv2 only, newer versions are AGPLv3 only. Because the SOGo backend code doesn't have an explicit different license, we can only assume that it's GPLv2 only and can't use this code with newer Z-Push versions. I don't think we want to maintain our own Z-Push fork forever and not be able to use newer Z-Push features. Because of that and that I also read that it was not production quality, I tried to see how hard it would be to implement a new backend (without reading any of the old code of course) that uses caldav-client-2.php from davical to talk to the caldav server. I spend a few days working on it and got as far as that I could sync my calendar items from the server to the phone. I've just pushed the code to github in the case anyone would want to continue with the effort: https://github.com/dekkers/s-push/tree/caldav The problem is that I guess that 2-3 weeks of development time is needed to get two-way syncing of both calendar and contacts working and most bugs have been ironed out. I currently don't have the free time to do it and don't think there are enough people/companies willing to donate money to fund the development (but if you are, speak up here or in private, maybe I'm wrong and something like a kickstarter crowdfunding of this feature is possible). > > The one big advantage I see of ActiveSync is, that you have only one > > setup to do for Email,Calendar,Contacts and Tasks and all traffic > > flows via HTTP(S) > > > > There are several disadvantages of ActiveSync compared to native > > Cardav/Caldav support: > > - Only ONE Calendar, Contacts and Tasks can be synched This is actually a limitation a lot of clients have, but not a protocol limitation. The protocol supports multiple calendars/addressbooks/etc. > > - Only one ActiveSync account is possible This of course also depends on the client and some clients support multiple accounts. > > - Mail handling is very limited > Indeed but it should be compared with SyncML/Funambol, not CalDAV/CardDAV. > > I would still very much like to open a discussion about this - ie., > Funambol (SyncML) backend vs. Z-Push (ActiveSync) backend. > > To offer a very good mobile experience, I feel we'll have to make a > choice and focus our efforts into one or the other. We chose Funambol > many years ago. It might have been a good choice back then but it > doesn't mean it's still the best. I think the advantage of ActiveSync is that almost every phone has ActiveSync built-in and that it's well tested. Having ActiveSync is pretty much a requirement for a mobile phone, compared to SyncML/Funambol that also needs to implement the client application. Another advantage of a caldav/carddav backend for Z-Push is that it can also be shared by other projects implementing caldav and/or carddav, such as for example davical. > On my end (and speaking for myself only), there are some things I > dislike about Funambol such as a worthless bug tracker, patches taking > decades to get accepted (or even bugs getting acknowledged), Java-based > (enough said here), requires more resources than all other SOGo > components and requirements combined together and multiplied by your > favorite factor of the day (must be greater than 2), almost impossible > to package together with SOGo and more. I agree about Funambol, I ditched it completely completely and currently only support iOS and Android for the very small group of customers I currently have. With regards to packaging, Z-Push is already packaged in Debian under the name D-Push because of trademarks. > What I mostly fear about Z-Push is that we'll revisit all the device > synchronization bugs we had over the past few years with Funambol/SyncML > and we'll have to hack around like we did. This is very valuable time > and know-how here. Getting the Z-Push connector up and running can be > done in a snap, but refining it will take some time. Also, having the > Sword of Damocles on top of my head regarding the licensing agreement > doesn't make me a happy, happy, joy man. If you're concerned about patents on ActiveSync then you should also be concerned about the patents on the Exchange-Outlook protocol, because MS also lists quite a lot of patents in that area. At least they have been forced by the European Commission to disclose those patents, so it is possible to check whether any patents are infringed. Kind regards, Jeroen Dekkers -- users@sogo.nu https://inverse.ca/sogo/lists
