You could package it to substitute the original binary as it's done by the enhanced busybox version.

I think it is not good idea to create package which substitute package. Then we will have package which substitute package which substitute package which substitute package...

No Debian system is doing this

And as I wrote I need to fix some code...

Recodes of the other things will come. I have a vested interest in seeing an /sbin/preinit replacement (or licensing change - not that I'm hoping that'll go through) that permits for proper OSS-style redistribution.

As for the other pieces, I honestly don't see what's wrong with packaging open-getbootstate seperately at first, and the making a 'proper' package replacement once all the components are available in properly open form.

Up to you and the CSSU people of course, but I don't think an open replacement of a closed part should be kept from the CSSU, especially when we don't know when or if everything that belonged to the same Nokia stock package will get recoded.

Oh, I never addressed this point - what's the code you think needs fixing up?

correct saving boot state for DSME. Now it save in harmattan format, this need to rewrite for fremantle version. DSME is open, so I need only time to fix that. You can see in my getbootstate commented code, which force to always write boot state to DSME file.

Next problem is that my version (patched harmattan) use system time to detect frequently boot counts. But on Maemo 5 system time is read from HW in later boot. So getbootstate still gets < 60 seconds timestamps.

Fixed problem with time. Now it only use reset boot count and watchdog reset count (in fremantle watchdog is not used, but it does not break anything). Updated sources are on: https://code.launchpad.net/~pali/+ju...o_getbootstate

I got the opportunity to test your getbootstate on a second, completely newly reflashed N900 (fiasco and eMMC), without any other modifications involved in the boot.

I just wanted to say that it works perfectly on this N900 too, with almost completely stock settings (this is not, however, your latest update that fixed the problem with time, but the one I compiled earlier - but that shouldn't make a difference).

So as I understand it you only have to fix the format in which it saves the bootstate for DSME and then it's ready for CSSU inclusion?

By reading the whole post (and better the rest of the thread) not only that sentence

Now that a "stable" CSSU has been created, is there any more chance of this being moved into the "testing/sid" CSSU?