I’m confused now, the post above says that there will be initial compatibility but after that (presumably as sublinks evolves) Lemmy apps will have to display as best they can.
That was my reason for asking about sublinks apps.
I know that devs like web interfaces but the truth is you need apps if anything social is going to become established.
It’s a sad fact in the fediverse that routinely apps or even front end GUI aren’t compatible. There’s only so many times this can happen (or instances shut down or not be maintained) before it all becomes a bit much.
Depending on the community (instance admin included) reception of Sublinks, several scenarios can happen
a lot of major instances switch to Sublinks
only a few instances switch to it
As Sublinks will be compatible with Lemmy apps at that time, there will be a time where both will coexist as options to use with the apps. Application developpers will probably cater to the most population solution, whatever that is. Ultimately, Lemmy and Sublinks won’t probably be fully compatible anymore, and apps will need to choose.
It’s actually kind of already the case with Mbin, which doesn’t have as many clients as Lemmy, while still allowing users to interact with the Lemmy communities.
Yes, my point is that an App or apps should be part of the design from the outset (see Pixelfed) Good planning rather than chance and potentially messing / confusing people.
I mean if you don’t secure it sure. The same goes for any site though and can be mitigated by having decent security. There was the thing a bit ago with Apple trying to discontinue PWAs in the EU due to the ruling that PWAs on apple devices shouldnt need to only be able to rely on webkit (their browser engine) which is why the voyager creator said to move over to native but them discontinuing PWAs got walked back
That’s not what I’m thinking about. Was hoping not to get too technical (I’ll try to keep thing simple) but there’s a general concern with PWAs that mean they can only support less secure authentication compared to a native app.
For example, if a device is used offline then a PWA can’t authenticate , so anyone who picks up the device can see the users content. There’s a way around this using what’s called a “service worker” which uses device storage but this would need to be different for each device type and some devices won’t persist between sessions. Also, if a user is in an online session then closes the PWA app, goes offline and then back into the app then the service worker will generate an error.
There are other Lemmy specific issues as well I don’t completely remember the detail but it relates to web clients not being able to set a custom cookie header to ensure compatibility (CORS limitation) There’s an explanation somewhere in the Voyager community but I couldn’t immediately scroll back enough to find it.
I’m confused now, the post above says that there will be initial compatibility but after that (presumably as sublinks evolves) Lemmy apps will have to display as best they can.
That’s indeed it: initial compatibility which might change as SL evolves
That was my reason for asking about sublinks apps.
I know that devs like web interfaces but the truth is you need apps if anything social is going to become established.
It’s a sad fact in the fediverse that routinely apps or even front end GUI aren’t compatible. There’s only so many times this can happen (or instances shut down or not be maintained) before it all becomes a bit much.
So I think relying on Lemmy apps is a mistake.
It’s really to early too know any of this.
Depending on the community (instance admin included) reception of Sublinks, several scenarios can happen
As Sublinks will be compatible with Lemmy apps at that time, there will be a time where both will coexist as options to use with the apps. Application developpers will probably cater to the most population solution, whatever that is. Ultimately, Lemmy and Sublinks won’t probably be fully compatible anymore, and apps will need to choose.
It’s actually kind of already the case with Mbin, which doesn’t have as many clients as Lemmy, while still allowing users to interact with the Lemmy communities.
Yes, my point is that an App or apps should be part of the design from the outset (see Pixelfed) Good planning rather than chance and potentially messing / confusing people.
The site will be a PWA. Not an app but something for people to use on mobile that could possibly be distributed on app stores
Aren’t there problems with authentication that cause security concerns when using a PWA?
Which is why the native apps are preferred for Voyager.
I mean if you don’t secure it sure. The same goes for any site though and can be mitigated by having decent security. There was the thing a bit ago with Apple trying to discontinue PWAs in the EU due to the ruling that PWAs on apple devices shouldnt need to only be able to rely on webkit (their browser engine) which is why the voyager creator said to move over to native but them discontinuing PWAs got walked back
That’s not what I’m thinking about. Was hoping not to get too technical (I’ll try to keep thing simple) but there’s a general concern with PWAs that mean they can only support less secure authentication compared to a native app.
For example, if a device is used offline then a PWA can’t authenticate , so anyone who picks up the device can see the users content. There’s a way around this using what’s called a “service worker” which uses device storage but this would need to be different for each device type and some devices won’t persist between sessions. Also, if a user is in an online session then closes the PWA app, goes offline and then back into the app then the service worker will generate an error.
There are other Lemmy specific issues as well I don’t completely remember the detail but it relates to web clients not being able to set a custom cookie header to ensure compatibility (CORS limitation) There’s an explanation somewhere in the Voyager community but I couldn’t immediately scroll back enough to find it.
Seems like a good idea, a la Voyager