Re: SUBSURFACE update

[Dirk Hohndel]

Not sure why this requires reaching out to me personally instead of asking questions in the forum, but whatever.

Updates continue to be distributed, some of the details have changed (the biggest one being unable to update the Android version in the app store, but you can still download updated Windows versions, and you can still install a newer version on Android, you just need to deal with the uninstall / reinstall the first time you do it.

So, first step, ensure all of your device are synced with the cloud. Do so manually (by explicitly triggering a sync with the cloud) both on Android and on Windows.

Then uninstall the Android app from your phone.

Go to  https://subsurface-divelog.org/current-release/ and download the new versions - on Android that will require enabling installation from 'untrusted sources' and enabling the browser to install an app,

With that you should be ready to go.

Again, all of this you should have been able to ask on the forum - which is why I added the forum to this conversation,

/D

On Jan 30, 2024, at 04:08, Giulio wrote:

Hi Dirk
I'm sorry to disturb you personally, but I don't understand, even consulting the site or forum!
After some periods of absence from diving for personal reasons, during the last holidays I took a look at my subsurface installations to see what was again.
I therefore discovered that updates are no longer distributed, but it is necessary to download them. Well, but I am still at the 5.0.0.10 version of Win and at 3.4.7 (5.0.9.35) of Android, while it seems that for example for Win you moved on to the 6th ..
I can't find an Installer for win and/or an apk for Android on the repository! I only find some .java (I think to fill in, but this despite being a professional of computer science is not in my skills).
Where can I find the compiled to be installed? Can you give me a direct link?
I know you are a very helpful person, but at the same time decidedly determined in your choices, in the past we have had the opportunity to discuss it, but I allow myself to say that I believe this solution adopted very limiting for those who do not have enough IT skills.
I thank you, however, for your availability and I offer you a dear greeting.
Congratulations on the continuous work you are doing with Subsurface [compatibly to my gaps ;-)]
Giulio

[Joe Meyer]

Hi Dirk,
Does that mean it's not possible to publish rolling updates to Google Playstore? What are the major limitations in that?

I wonder how many users  will not be able to Subsurface because they cannot follow the sideloading thing. Also what happens to users that also have a company profile on android to does not allow installing APKs from untrusted sources due to company policy enforcement on the phone?

Regards

Joe

[Dirk Hohndel]

Correct, I can no longer post to Google Playstore, and the small number of users who still can access the app right now will lose that access, soon (Google keeps changing the timeline).

Sideloading works for the vast majority of users. If your employer is able to prevent you from loading an app on your phone, this is no longer your phone but your employer's phone - not something that I can do much about (and no, I would NEVER FOR A MOMENT consider allowing an employer to have such control over my personal device - if you want that much control, pay for a phone for me... and yes, I do travel with two phones because my employer very much demands that control - and there is not a single app on that device that wasn't installed by my employer). 

So yes, I assume there will be a few users using Subsurface on a work-device that will be frustrated by this solution, but that is the only one we have at our disposal at this point.

/D

[Joe Meyer]

Hi Dirk,

I definitively do not disagree about phones having work parts on them but sometimes there are situations where this might still be viable. But agree this is an exception within the userbase of Subsurface mobile users.

I wonder what are the Google Play Store restrictions about the app publishing and is there anything that could be done to still allow posting some versions to it? Is it technically just to cumbersome or what is the main reason? Maybe I could help looking at this if there are still options available?

Regards

Joe

[Dirk Hohndel]

You are welcome to help, of course.

Google is constantly changing rules, constantly tightening down what apps cannot do, and where they need to go through lengthy bureaucratic processes. The amount of times that I have personally wasted on yet another BS rule from Google... I don't want to contemplate. All of course in order to "protect the user" (i.e., to ensure that only Google gets to track you, spam you, exploit you) - but I digress.

The specific issue at hand is that in order to comply with the permission rules that would allow an app to scan for BLE devices, you need to follow certain rules - and Qt5 for Android isn't set up to do so. The Qt company has largely abandoned open source Qt5 (outside of the mandatory delayed GPL releases that they do). But of course Qt6 was a complete disaster when it was rolled out, missing half a dozen critical components that we needed -- and even now, years later, is still not fully compatible with the features that were available in Qt5. Up until fairly recently a couple of our core library components wouldn't work with Qt6 at all - so we couldn't migrate over. And that's really what triggered this change in strategy.

All that said, with the Plasma6 release I assume that the Kirigami port is far enough along that someone with sufficient motivation could port our UI (and the customizations that we did to Kirigami in the past) to the new Qt6 based Kirigami. And then deal with all the other changes to the way Android apps are built with Qt6, etc. To me, personally, that always seemed like an insane amount of work for essentially very little actual value - and so every time I started I gave up in disgust.

But that's the other part of the puzzle here. We have not a single Android developer left. Nor an active iOS developer. So for years and years this has always been me figuring things out, spending the time and the energy to chase both Apple and Google and (lately) the Qt Company and their ever shifting requirements. And I guess I have just gotten a little tired of that.

What we have now, Qt5 based, works just fine on iOS and works fine on Android with side loading. If my rough statistics are correct we have more than twenty thousand users between all of those platforms and we see about six thousand active users every month, almost all of which have at least one mobile device connected to the cloud. So clearly what we have today works for a lot of people. So in the tradeoff of spending my time on the infrastructure, on supporting users, and on making things better for all users, vs. chasing the handful of people who don't have a mobile device that is actually theirs that they could use... yeah... no.

/D

> On Mar 5, 2024, at 04:17, Joe Meyer wrote:
>
> Hi Dirk,

[rade...@gmail.com]

Hi Dirk,

I think I'm not the only one who appreciate the amount of efforts you're putting into Subsurface. So huge thanks for that!

For me, it's quite a surprise there are no active mobile developers anymore. So it looks like we desperately need someone who would be able to rebuild mobile applications on both platforms into a new version of the existing or a completely new framework.

Would it be possible to reach out to the existing community of users, even outside of those following this forum, to see if there is someone willing to do so?

Or to raise some kind of crowd funding to pay some professional to develop it while staying open source? I would gladly contribute a few $ for this purpose.

I think it would be very sad if such great app would gradually fade away just because everyone will expect such an app for free, with frequent updates and new features, but not willing to actually contribute somehow.

Regards,

Radek

Dne úterý 5. března 2024 v 13:35:48 UTC+1 uživatel Dirk Hohndel napsal:

[Joe Meyer]

I would definitively agree with Radek. Your work is extremely appreciated and Subsurface is one of the best dive softwares out there. THANK YOU for that!

I wonder how much it would take to get a community sponsored QT dev to port the existing parts into a new up to date framework that can handle the cross platform. I'm also willing to sponsor some $ for the work if that is what it takes.

Unfortunately I'm not at all into frontend development on mobile devices, otherwise I could  even give it a try myself. Alternatively I was wondering if it could be an option to run the existing code through an AI model and let it convert it to a supported framework on its own. I'm aware that this might not be as simple as it sounds.

Regards

Joe

[Michael Keller]

Hi Joe.

On Thu, 7 Mar 2024 at 06:21, Joe Meyer <linuxc...@gmail.com> wrote:

I wonder how much it would take to get a community sponsored QT dev to port the existing parts into a new up to date framework that can handle the cross platform. I'm also willing to sponsor some $ for the work if that is what it takes.

One problem here is that this isn't just a one-off situation. Google and Apple keep moving the bar for acceptance in their stores, so for any app to stay publishable in the stores a pretty much constant effort is required. And the work often requires specialist mobile development knowledge, something that is hard to source. So funding will probably have to be ongoing and not one-off-ish as is often the case for donations, so that a developer (or a team) can be retained on an ongoing basis to do the work to keep the app releasable.

Cheers

  Michael Keller

--

GCS$/CC/E/IT d- s+ a C++ UL+++/S++ P L++ E-
W++ N o? K? w O(++) M-- V+ PS+ PE+ Y? PGP+ t
5? X R tv b++ DI++ D++ G e+++ h---- r+++ y+++

[Dirk Hohndel]

On Mar 6, 2024, at 12:34, Michael Keller wrote:

Hi Joe.

On Thu, 7 Mar 2024 at 06:21, Joe Meyer wrote:

I wonder how much it would take to get a community sponsored QT dev to port the existing parts into a new up to date framework that can handle the cross platform. I'm also willing to sponsor some $ for the work if that is what it takes.

One problem here is that this isn't just a one-off situation. Google and Apple keep moving the bar for acceptance in their stores, so for any app to stay publishable in the stores a pretty much constant effort is required. And the work often requires specialist mobile development knowledge, something that is hard to source. So funding will probably have to be ongoing and not one-off-ish as is often the case for donations, so that a developer (or a team) can be retained on an ongoing basis to do the work to keep the app releasable.

I do share everyone's frustration and disappointment. And I keep hoping that someone will step up and work on the app - but it turns out that most Android developers only have experience with Java based tool chains - and that the total number of Qt based Android apps is minuscule, and even worse that even among those ours is a bit "special" since it also has a bunch of other complex dependencies that aren't easily satisfied. So even the theoretical pool of developers id tiny...

/D

[Joe Meyer]

Dear Dirk/Michael,

Seems like we are all quite on the same page. In the meantime I gave it a try to sideload the latest version to my FP5 manged mobile and I have to admit that the sideloading process has been simplified quite a bit since I last used it. It can be done by simply browsing and downloading the latest version from the Subsurface website directly. Worked flawless and hence I can agree with Dirk, that this option is actually a viable way for the time being, until there is hopefully better way to support Android. 

I also read a bit about QT6 and it seems with 6.2 it would probably support the necessary dependencies Subsurface mobile is requiring. Not that I have looked at the Subsurface Mobile code but 6.2 has a long list of additional features such as Bluetooth and QTReports that are currently missing and I assume are some of the dependencies Dirk is referring to. It shall be released in the September time frame, which is still a while from now but maybe at that time a real attempt to migrate to the newest QT version could be tried and hopefully this would allow to fulfill Googles requirements in the future.

Thanks a lot both for your reactions and insights. For further motivation, I provided a small donation to the team that are keeping things running! THANK YOU.

Regards

Albin

[Dirk Hohndel]

Qt 6.6 (!) appears to have everything that we need and I have started trying to create a working build.

That is made much harder by some of the changes made to the way Android binaries with Qt are built, and especially how Kirigami can be integrated (and of course all of our Kirigami changes need to be ported to the current versions).

But at least there appears to be light at the end of the tunnel.

That said, we really could use more help and especially a developer who is familiar with Qt and Android development. But those are very very very few it seems.

/D

[Joe Meyer]

If I come across a QT/Android dev, I will try to pitch for help on this.

Regards

Albin

[Joe Meyer]

I can offer my help for any kind of testing or other non coding work on this if needed. Just ping me if there is anything I could help with.