51killer.tcl on 11.x?
This is more of a curiosity question than a request for help - I've already resorted to doing a C&DE to fix the Message 51 error on a HD I swapped between TiVo HDs.
I restored an old image on a 160GB HD for my TiVo HD (PROM'd) and had the usual Message #51 error when trying to do anything. So I fired up telnet and ran 51killer.tcl. Said successful, rebooted.. nothin'. Then I found the 51killer.tcl for 8.x. No go.
So then I looked at /State/Media and saw the DiskConfiguration and DiskConfiguration2 items were deleted successfully after running the updated 51killer.tcl. However, when I rebooted the TiVo, they were being recreated and not updated to match the new TiVo (I can't say exactly if the entries were exactly the same, but I still had Message 51). There was an "UNKNOWN" entry that was the same before and after the restart, but I could not remove it.
Anybody know if there's an extra step to getting 51killer.tcl ported to 11.x or did I just have a very stubborn installation?
Yes, the script just needed a tweak (which evidently I had made back in 2009 but never posted for some reason).
See this thread (link) for the updated script.
Thanks! Yep.. that was the "UNKNOWN" entry that I was talking about. I wonder why mfs_dumpobj shows UNKNOWN but TiVoWebPlus shows it correctly. I guess those names are hardcoded?
Wish I posted 1 day earlier. Would've saved 2.5 hours
Yes, effectively they are hardcoded. mfs_dumpobj was compiled with what is now, an older (fixed) version of the MFS schema, so it doesn't recognize newer MFS types. I forget if TiVoWebPlus is the same way, but merely has a newer version of the schema or if it dynamically queries the schema.
There is a compiled in schema, but it can be overwritten with a schema read from a file, via the TIVO_SCHEMA environment variable. See README.Jamie in the mfs-utils distribution for more info.
There is also a schema.tcl script that can dump the current schema, although I've found that it doesn't return correct results anymore. It makes some assumptions about consecutive ordering of object and attribute id's that are no longer true in some of the later tivo releases. I also think that the dtivo and sativo schema's may have diverged now so that a single schema may no longer work for both platforms.
Ah, thanks for the clarification regarding the TIVO_SCHEMA environment variable. I had encountered the script to dump the current schema, but didn't realize that it could be read at runtime (and had noticed issues with its output).