Not all input bits getting sent

[steve057917]

Hi, I was trying to set up an Arduino to send 20 bits of info (button presses) to JMRI. But only 17 were getting sent.
After head scratching and testing with a bare minimum sketch that just toggles the inBits every 5 seconds, I saw the comments about issue #10. I think the update made under issue #10 , line 115 is not correct. I changed the comparison to > (greater than) and I am now getting the full 24 bits of input data. Please double check the comparisons and logic for testing for out of bounds conditions.

[wizel10]

Hi,

I can confirm this bug. I've been struggling with the same issue - inputs above bit 16 never reaching JMRI.

The fix is on line 115 of CMRI.cpp. The bounds check:

if ((pos + 7) / 8 >= _tx_length)

should be:

if (pos / 8 >= _tx_length)

With the original code, bit 17 evaluates as (17+7)/8 = 3, which equals _tx_length (3 for 24 inputs),
so it returns false and the bit never gets sent.

With the fix, bit 17 evaluates as 17/8 = 2, which is less than 3, so it works correctly.

I tested this fix and can confirm bits 17-23 now work correctly.

Hope this helps and that the fix gets merged soon.

[steve057917]

Hi, thanks for looking at this. I have made your suggested change to my copy of CMRI.cpp and it appears to be working. Thanks again.

…

On Mon, 4 May 2026 at 20:24, wizel10 ***@***.***> wrote: *wizel10* left a comment (madleech/ArduinoCMRI#17) <#17 (comment)> Hi, I can confirm this bug. I've been struggling with the same issue - inputs above bit 16 never reaching JMRI. The fix is on line 115 of CMRI.cpp. The bounds check: if ((pos + 7) / 8 >= _tx_length) should be: if (pos / 8 >= _tx_length) With the original code, bit 17 evaluates as (17+7)/8 = 3, which equals _tx_length (3 for 24 inputs), so it returns false and the bit never gets sent. With the fix, bit 17 evaluates as 17/8 = 2, which is less than 3, so it works correctly. I tested this fix and can confirm bits 17-23 now work correctly. Hope this helps and that the fix gets merged soon. — Reply to this email directly, view it on GitHub <#17 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/CDEUW7LUMRG2ZCEJ3UXZZR34ZDU5FAVCNFSM6AAAAACYQC7XK6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DGNZTHA2TKNZRGU> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[wizel10]

Glad it helped. You raised the issue, so thanks to you.
Didn't thought it was any issue with CMRI after so many years. Seems that not many people uses more than 16 sensors.
I've subnmitted a code request change, but don't think will get updated as looks as an abandoned project.
Regards

[Erik84750]

I hope madleech will in due time also update the cmri.cpp

Thanks for pointing this out.

[brocci]

Hi. I have proposed a fix last week regarding this. See #16 . Waiting for approval and merge.

[madleech]

Thanks for this, I'll look into it this weekend. I've been slowly catching up on some of my repo maintenance, adding tests, etc.

[madleech]

It took a little longer than expected, but I've released v1.6.0 with this change. Thank you very much for your patience and debugging, I really appreciate it.

As a side note, I no longer have a functional model railway, and haven't used JMRI in years, so my ability to test changes is very limited. Since I'm not daily-driving this code, I have been hesitant to touch it for fear of breaking it for people who do use it. However as part of v1.6.0 I've added unit tests and a CI harness to run them automatically, which should make changes safer.

If someone could test v1.6.0 then I'll be happy to close this issue.

[Erik84750]

Many thanks for your update Michael.
I will be able to do testing but that will be only towards year's end. Maybe others will be faster? I keep you posted here.
Erik