Linked by Thom Holwerda on Fri 25th Jul 2008 22:55 UTC, submitted by Chavez
Thread beginning with comment 324643
To view parent comment, click here.
To read all comments associated with this story, please click here.
To view parent comment, click here.
To read all comments associated with this story, please click here.
ACPI is an example of a terrible specification. It's 600 pages long and contains all sorts of features it probably doesn't need.
Last time I heard, the Microsoft DSDT compiler was also very good at generating DSDTs that could only be interpreted properly by Microsoft operating systems (unlike the Intel one).
Also, if I remember correctly, the DSDTs can specify that certain things should only work on certain operating systems. Since many manufacturers seem to have the idea that Windows is the only OS that supports modern computers, they mark all the ACPI features as requring the OS name to be Windows.
Really, the ACPI spec should be simple and OS independent.
Last time I heard, the Microsoft DSDT compiler was also very good at generating DSDTs that could only be interpreted properly by Microsoft operating systems (unlike the Intel one).
Also, if I remember correctly, the DSDTs can specify that certain things should only work on certain operating systems. Since many manufacturers seem to have the idea that Windows is the only OS that supports modern computers, they mark all the ACPI features as requring the OS name to be Windows.
Really, the ACPI spec should be simple and OS independent.
True, which is the great thing with Apple - they do their own firmware, and they seem to get the stuff working correctly without too many problems - which the benefit of controlling the whole widget.
Quite frankly, I'd sooner see each distribution owned by an OEM, and seeing the OEM customising their hardware and distribution so that they work together seamlessly. A single focus on their hardware alone so that no compromises are made for the 'greater compatibility' outside their own hardware line.
anyone else getting flashbacks to stories about unix fragmenting into 1001 slightly incompatible variants?
sure, the gpl licence should stop most of that, but if so, why the hell even bother...
and on the topic of ACPI:
http://linux.slashdot.org/article.pl?sid=08/07/25/1150218
Quite frankly, I'd sooner see each distribution owned by an OEM, and seeing the OEM customising their hardware and distribution so that they work together seamlessly.
Yes, let's go back to the eighties and make incompatible variations for feature differentiation! How does that help anyone? (Arguably, the eee has already gone this path,and this is why many people eventually replace the severely limited customized Xandros distribution.)
Edited 2008-07-26 10:19 UTC






Member since:
2006-01-04
ACPI is an example of a terrible specification. It's 600 pages long and contains all sorts of features it probably doesn't need.
Last time I heard, the Microsoft DSDT compiler was also very good at generating DSDTs that could only be interpreted properly by Microsoft operating systems (unlike the Intel one).
Also, if I remember correctly, the DSDTs can specify that certain things should only work on certain operating systems. Since many manufacturers seem to have the idea that Windows is the only OS that supports modern computers, they mark all the ACPI features as requring the OS name to be Windows.
Really, the ACPI spec should be simple and OS independent.