Notices

Indeed, just QML is insufficient for full compatibility. You have to redesign to make it cross-platform or use some cross-platform approach from the start (as modRana) or "translate" the app. Not sure whether translating that UBPorts app would be better than continuation of the available code. Have you tried to make some comparisons between them?

I agree with @r0kk3rz assessment - we are currently lacking people with skills taking the time for development. And its not clear how that could be resolved.

As for Matrix clients, lots of them are discontinued. Not sure why, but it looks to be all over the place when I checked at matrix.org. The client maintained developed by @r0kk3rz isn't the simplest piece of software, but I haven't looked deep due to commitments in navigation apps and will probably not be able to do so in a while.

But if you are interested in development of the client, even investing 0.5h per day could make a huge difference.

totally agree with @rinigus about: 0.5h. = )

@r0kk3rz and @rinigus, thank you so much for all the detailed information. It is a little sad to see that all the efforts made by the Sailfish and ubports communities cannot be suitable combined to achieve some synergy for both OSes. But that probably can't be changed any time soon.

@rinigus, I noticed all the discontinued Matrix clients as well. This may be a key problem as to why "people" choose certain commercial chat protocols over XMPP, Matrix etc. Too many clients, each with different strengths and weaknesses. Ironically the best XMPP client I experienced was the native one that came with the N900. From that point of view, it might actually be best to get Riot.im working on Sailfish (i.e. hosting your own copy of riot-web on the Sailfish device). But I feel the foundations for that may first need to be laid by Jolla (?).

Alas, I wish I were able to code. At present, I'd need hundreds of half-an-hours to just cover the basics of QML, JS etc.

@dubliner why not to donate at few RedBull to: @r0kk3rz, for coding matrix.?

The matrix guys recently did a talk at CCC about the current state of the matrix platform and where they are going next.

https://media.ccc.de/v/35c3-9400-mat...d_year_to_date

The project lead also had a talk on FOSDEM:

https://fosdem.org/2019/schedule/eve..._french_state/

This talk also goes over the current state and future plans besides they work they did for the French government. With the 1.0 version and specs out of the door, this seems to be going somewhere now?

I followed that, too. It's quite amazing to see lots of things happening there.

I was also delighted to see that r0kk3rz has already commented on the issues listed on Github. My hopes are high for some further progress there now that some of the factors have been pinpointed.

As for "all we need is everybody spending some time on development", I am sorry to have learned that time is not the only constricting factor. I don't want to go OT too much, but I spent two afternoons installing the SDK, getting the examples running, only to find out that I cannot seem to manage to make harbour-matrix start in the VM. So, apart from time, one also needs lots and lots of background knowledge and presumably a ton of talent - neither of which I do seem to sport.

There was a bit of good news, albeit well hidden, on TJC: cy8aer has been working on a Matrix client (and he posted it yesterday, i.e. it does not seem to be an April fools' joke ). Yay!!!

Here is the link:

https://gitlab.com/cy8aer/bluepill

Now, an RPM would be nice for testing.

@cy8aer you made my day!!!

yep my post was done on march 31... As I wrote: that is early stage. You can compile it with an actual sdk and if you have an account somewhere you can already chat. But there are many things missing:

• register
• upload of any pictures music data
• no e2e - I need to know how to get this done with matrix-python-sdk - and how to pack the libraries to make them harbour compatible
• no room manipulation
• no m.read receiving by now (sending works but cannot be switched off - prefs missing)
• many preferences stuff (with up/download to server)
• does anyone used the gstreamer webrtc stuff? No voip.
• no unread status
• no notifications
• no logo. Need to convince Daniel to make a new one with blue pills...

And there must be much caching stuff done. When you first login it will be waaaay slow because I pre cache the room data.

Of course: rpm when it is useful ready. Probably there will be a non crypto version first on harbour and e2e will follow on openrepos lateron because I guess they never will accept the binary python cffi stuff.

Much stuff to do...

I investigated the gstreamer webrtc stuff when I was getting the plugin ready for inclusion into mer.

https://github.com/r0kk3rz/gstwebrtc-demos

Audio should work fine, video would need to be H264 as the hardware encoders for VP8 arent enabled in gst-droid yet