Skip to content

Conversation

@tsterbak
Copy link
Member

@tsterbak tsterbak commented Feb 12, 2025

This release adds (actually, this release might be part of release 0.5.4 to fix the flatpak):

  • tracking of device codes of installation attempts in plausible.io.

New devices:

  • none yet

Chores:

  • fix flatpak creation
  • update flet to version 0.26.0
  • fix a bunch of bugs from the flet update

@tsterbak tsterbak self-assigned this Feb 12, 2025
@SirRGB
Copy link
Contributor

SirRGB commented Feb 13, 2025

Is this related to #623?

@tsterbak
Copy link
Member Author

Is this related to #623?

Yeah, I was hoping to fix it with release 0.5.4, but there seems to be an issue with flet version 0.24.1 (or I'm confused) and the flatpak fails to work. So I will try to go full flet 0.26.0 where it seems to be fixed.

@tsterbak
Copy link
Member Author

So this release will probably go out fast once I confirmed that the flatpak works this time.

New devices can go right in ;)

@tsterbak
Copy link
Member Author

tsterbak commented Feb 15, 2025

@SirRGB @MagicLike @freechelmi @BjornW I added some code to send the device code of started and successful installation attempts to an analytics service (plausible.io). This would be a direct way to collect usage information and what devices people use the installer with. And it would help my motivation to actually "see" people using the installer.

What do you think about it? Would you (strongly?) oppose this?

If your busy, a 👍 or 👎 would already help me 🙂

Edit: 👍 would indicate you are fine with collecting usage information, while 👎 would indicate the opposite (because it might be unclear from the text)

@freechelmi
Copy link
Contributor

freechelmi commented Feb 18, 2025

@SirRGB @MagicLike @freechelmi @BjornW I added some code to send the device code of started >and successful installation attempts to an analytics service (plausible.io).

This would be awesome, also even we could gather attempts on unsupported devices , this would help to focus on most popular devices to come.
Keeping the link to issue creation on unsupported device requests help keep in touch with testers for betas

@tsterbak tsterbak changed the title Release v0.5.5-beta Release v0.5.4-beta.1 Feb 19, 2025
@tsterbak tsterbak merged commit f052e6a into main Feb 19, 2025
@BjornW
Copy link
Contributor

BjornW commented Feb 19, 2025

A bit (too?) late: I'm not a big fan of telemetry even though I can see the usefulness.

Please make sure to clearly & explicitly explain this feature: why it has been added to the application, the purpose and the type of data collected.

@tsterbak has something like this already been added to the website?

@tsterbak
Copy link
Member Author

A bit (too?) late: I'm not a big fan of telemetry even though I can see the usefulness.
Please make sure to clearly & explicitly explain this feature: why it has been added to the application, the purpose and the type of data collected.

Yeah, I'm also not a big fan usually. But it would help a lot to see what devices people use the installer with and if they are successful (or not). And it would help my motivation to actually "see" people using the installer. 😁 But yeah, maybe it will be removed again, lets see how it goes.
I'm also thinking about making the collected data publicly available on the website or at least as regular updates on mastodon.

@tsterbak
Copy link
Member Author

Quick different question:
Can some of you (@freechelmi, @SirRGB or @BjornW) give this version of the installer a try? I have some issues with fastboot parts for some of my testing devices, but I'm not sure if its wear on the usb-cable/port or an issue with the installer.

@SirRGB
Copy link
Contributor

SirRGB commented Feb 22, 2025

Just ran the sargo test and didnt notice anything different.
Btw is there any way to see these stats pulbicly? Imo it would add to the transparency of the project...

@tsterbak
Copy link
Member Author

The test doesn't actually run anything its just for clicking through the frontend (if we are talking about the same thing). I noticed occasional issues with fastboot/adb and I'm not sure if it is my testing devices or the installer... (Points towards issues with the testing devices/cable/ports/drivers: in the state when the installer stops working, I can also not continue with non-installer adb/fastboot.)

Yeah, I could make the whole statistics of the website and the installer events publicly available. I think I will add a link to the readme and the website under "Statistics". Or do you have another idea?

@SirRGB
Copy link
Contributor

SirRGB commented Feb 22, 2025

Ic, Ive also ran through oai on my S5 Mini, to also close that weird issue I created at some point, so that should show up now ig. Maybe it is also possible to mention the "success rate" and "unlock/twrp" on the stats page, to see issues? I dont think we differentiate between "ended app" and "doesnt work" rn?

@tsterbak
Copy link
Member Author

tsterbak commented Mar 6, 2025

Ic, Ive also ran through oai on my S5 Mini, to also close that weird issue I created at some point, so that should show up now ig. Maybe it is also possible to mention the "success rate" and "unlock/twrp" on the stats page, to see issues? I dont think we differentiate between "ended app" and "doesnt work" rn?

@SirRGB I added a link to the collected statistics to the readme now. Have a look and maybe somethign is missing?

@SirRGB
Copy link
Contributor

SirRGB commented Mar 6, 2025

I didnt spot the actual devices and their finish rate...
Might overlook something, idk

@tsterbak
Copy link
Member Author

tsterbak commented Mar 6, 2025

I didnt spot the actual devices and their finish rate...
Might overlook something, idk

You have to scroll to the bottom and then click on the finished event for example. Then you see the device codes.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants