Pi missing keypresses

There is nothing stopping a user choosing a specific date version of 5.25 and remaining on that as their own “5.24+a bit” or perhaps “5.24u” (for USB fix)

I have already done that for my Pi which was particularly troublesome with missing keypresses (and key releases). It is my understanding that the ‘ROM’ linked at the beginning of this thread was produced specifically as a potential “5.24+a bit” specifically for this problem and it works for me.

I will be using it unless and until a more formal ‘fix’ is available.

Does *rmreinit usbdriver in predesk fix the problem?

It changes the behaviour slightly in that a menu open by itself now when the desktop starts. More interesting a typo in my first attempt which caused the boot to be interrupted allowed me to notice a “No keyboard no present – autobooting”.

Actually I use my old RPC keyboard with a PS/2 keyboard+mouse adapter, so I removed it before booting and my mouse issues disappeared. Seems that for some reason 5.24 as an issue with this adapter.

Is there any specific reason why RISC OS cannot use semantic versioning ? It seems to me that 5.24.1 version, incorporating Colin’s fix only, would solve the problem of “I don’t want to use 5.25 just to have my keyboard working”. It could also help in other future cases similar to this one.

Is there any specific reason why RISC OS cannot use semantic versioning ? It seems to me that 5.24.1 version, incorporating Colin’s fix only, would solve the problem

Probably not. A question for the ROOL guys I suppose.

Little known fact 1. The only way of reliably reading the OS version number is OS_Module,20,0,-1, which reads the BCD version number of the UtilityModule into R6.

Little known fact 2. Module version numbers have (up to) four integer and four decimal digits.

Therefore one could build an OS where UtilityModule is version 5.241 (or indeed 5.2401)

Therefore one could build an OS where UtilityModule is version 5.241

That should work well when the next version (5.26) comes alone with a version number smaller by 214 !

There have been so many important architectural changes squeezed into minor version numbers (eg 4.29→4.33, 5.22→5.23) that I’m using the full resolution of the four decimals for threshold testing.

It would also be a solution to the decision by somebody to release different module builds with the same version number

Like the ‘Debugger’ module switching to 2.xx – and getting mixed up with 26bit ‘Debugger Plus’ :-|

Like the ‘Debugger’ module switching to 2.xx – and getting mixed up with 26bit ‘Debugger Plus’

Not switching, incrementing.
It was a strange idea, to use merely a higher number to tell it apart (smacks of the same logic that gave us SIX), I guess we ought to be glad it took over two decades and the demise of Acorn before it became a problem.

Hi Colin,

Would anyone who has a problem with missing keypresses/ pointer problems like to try out this rom on a pi.

I have been using your ‘rom’ on a Ras Pi almost since you made it available with good results.

But the extra good news is that particular Pi uses Aemulor and Easysockets with a socket client program of mine written in BBC Basic. I am observing an improved easysocket error rate from that program. (with nobody touching the keyboard all day today and just under 2000 http requests).

I now have your ‘rom’ installed on this Pi and the same easysockets program running. Early indications are that the error rate is improved compared with 5.24 (although maybe not quite the same improvement as the original Pi).

The change has been added to ROOL’s CVS so should be in todays Beta RPi ROM.

So I think the crucial question (as it isn’t a problem for everyone and how bad it is seems to vary) is: Has anyone who was getting missing/autorepeating keys on a Pi still getting the problem with builds after June 2nd?