Re: [Libcdio-devel] Fedora bug report concerning CD-TEXT

[Thomas Schmitt]

Hi,

i wrote:
> i tested my favorite MNC / ISRC retriever. It needed some corrections.

Rocky Bernstein wrote:
> You think there is a bugi n that macro definition?

No, There was a bug in my initial untested proposal.
buf had become a char * (from char[24]) but its sizeof() was
still used as input for CDIO_MMC_SET_READ_LENGTH8().


>  #define CDIO_MMC_SET_READ_LENGTH8(cdb, len) \
>  cdb[8] = (len      ) & 0xff

Actually we'd need to set a big endian 16 bit number at cdb[7] and cdb[8].
But since we only use small numbers here, it is ok.


> Again, what I'd suggest is to create a branch off of master and commit your
> changes.

But don't you want to re-arrange it into mmc_[hl]l_cmds.c ?
If i did this myself, i'd have to ask you about every single decision.


> I find working off of diffs in the mailing list archive a little bit
> wonky.

That's why i posted the whole functions.

They are tested on MMC level in a provisory contraption within libcdio.
What is missing is their appropriate place and wiring. Plus possibly
corrections where i don't meet the coding standards.

Take them and make them fit into your plans. Adapt the three calls of
the old functions in mmc_get_mcn(), mmc_get_track_isrc(), and
mmc_isrc_track_read_subchannel() to what you finally decide.
This will save us a lot of forth and back.
(... and me the plight to find my cheat sheet about git.)


Have a nice day :)

Thomas