Jordan Crouse wrote:
LINUXBIOS_VER=2376
LINUXBIOS_CONFIG=Config.SPI.lb
Ok - but I'm setting those as the defaults in linuxbios/linuxbios.mk too.
Whats wrong with 2378?
If someone has confirmed it is good, then we can use it. But not until
we get positive confirmation (I guess richard confirmed it?).
I'm really paranoid about this version issue. We had a big problem,
never resolved, where somebody upgraded the epia chipset source for a
newer epia part and just totally broke the tree for epia (the changes
made should not have been a problem, according to everything we knew;
they nevertheless did not work). Yes, you read that right: a fix for
epia broke epia -- it was a chip rev. issue problem, I think. Sometimes
it can be something as minor as a "date fab'ed" problem -- bugs and bug
fixes sneak in to the chips, and a bug *fix* can cause as much trouble
as a new bug.
We've never recovered from that; it does not matter, as epia is no
longer made, but I was pretty unhappy about it. So, we definitely don't
want to bump the version numbers without a confirmed successful build.
As time goes by, we maybe ought to require two "success reports" for
OLPC, to be doubly sure.
ron
_______________________________________________
Devel mailing list
[email protected]
http://mailman.laptop.org/mailman/listinfo/devel
