PCM Hammer - The Uno UI, Improvements, and priorities

[antus]

Interesting stuff! Sounds like it's truly some trial and error when it comes down to it.

You wouldn't believe how many times I've rebuilt kernels, tried again, as trial and error. This is what most people don't understand. They think someone (maybe even an AI these days) writes the code based on a spec and it's going to work. That could not be further from the truth with embedded hardware! The chips and the power supply and noise all have some kind of influence and minor changes can make all the difference. Even chip bugs... those early cpu cores in p04 were hard! And when it bricks and you have to move to BDM it sucks even more time. And then there are a lot of platforms to retest, on a global change, like DLC code.

[NSFW]

I do a test write at the start of every tuning session. A full read would take too long, but a test write gives me confidence that the connection is good - OBD port, interface device, USB cable.

It's true that it doesn't cover the most common failure mode with the 512k intel chips, but I still like to verify the link quality in my P59 / amd 1mb car.

It's a habit that started with the EcuFlash software that I use in my Subaru. Check out the screenshot at the top of this thread:

https://www.legacygt.com/topic/69752-ec ... 7-install/

The buttons are, from left to right: open, save, read, write, test-write, and one that I never use. Verify, maybe.

[NSFW]

For the countdown, I'd like to remove it completely and improve the unlock failure handling (if we even need to change it). The response should tell us rejected due to too many attempts or not ready, and I dont think we sleep but we could put a 5 second delay on it so that we dont use all the 3 attempts before hitting the 10 second too many attempts error.

That's a really good idea.

[crowbarmaster]

Damn it, I fell asleep

New builds are available here:

https://github.com/crowbarmaster/PcmHam ... 4275634860

[crowbarmaster]

Addressing the top bar issues:

Over on the left, I'm suggesting we remove the labels, and just run with:

A newly installed application starts with "Select a device." This will immediately tell the user at least once what the label is used for.

After selection is proper, Change to "Detecting ({Port or connection type})" Shortened Bluetooth to BT to save space.

Rearranged the labels, and did some spanning trickery to get to this:

Screenshot 2026-04-11 151633.png

The screen now has to get quite small before it hinders the back button now, and the labels are nice and readable in this small space.
Even if labels collide, they still are discernable between each other.

Edit: Sitting back and looking at this, the twin "Connected" labels are going to confuse. Since I already drive the labels back to Detecting on fault, would that be enough to signify error?

Update: It doesn't look to bad on my Pixel 9:

Screenshot_20260411-192330.png

Screenshot_20260411-192523.png

[julespatch]

hey, ive got a pile of bricked p01/p59s under my desk if we need some more test units

[crowbarmaster]

Well, come on in then!

I've done a lot of testing with a P01 now, but the more the merrier. We still have loads of different variables to account for yet when it comes to the app itself, Android, Bluetooth, Devices etc.

I have an OBDLink MX+ and an EX, the ODBX Pro GT, and a VCX nano used for testing, so devices beyond that really help too.

Any details you can report on issues will be helpful!

[NSFW]

I'd like to think that we can just remove the back-button on Android (set that grid column to zero width) and just rely on the native back-button instead. Google says that all Android devices are required to have one.

Apple phones don't have a standard back-button, but I think we could use left-swipe for that instead. I've never had an iphone but it seems to be pretty common. Google says:

"A left swipe on iOS is not a universal equivalent to the Android back button. While swiping right from the left edge acts as a "back" action in many native Apple and some third-party apps, it is not enforced system-wide by Apple, leading to inconsistent behavior across different apps."

Sounds like Apple needs to get their shit together, but left-swipe should be fairly easy for iOS people to discover and use.

[antus]

I have noticed on android that hitting back takes me out of the app instead of back a page. Not sure if that is normal. I'd expect back to go back a page until your at the app root, then back out of the app (but leaving it running so you can swap back to it). But im not sure exactly how that should work, say, if you are in the middle of a flash operation. Then I guess you cant back out of the page. So does it detect this and leave the app running in the background? If it is running in the background does that need BT permissions that it can always talk to the BT not just when the app is in the foreground?

I tested the lastest crowbarmaster windows build with AVT on the P11 that happens to be on my bench and was able to read and write it successfully.

[crowbarmaster]

Yes, right now that is entirely expected. Android requires we place a handler on an event tied to the Page object. I think I know what needs to happen for that yet, and it's on the top of the list.
This goes hand in hand with a task handler that can run as a Foreground Service, which will prevent the app from being suspended/killed while the service is active.

Stay tuned, it's all coming!