From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jody Bruchon Subject: EMS support will probably never happen Date: Fri, 17 Feb 2012 09:48:01 -0500 Message-ID: <4F3E68A1.6080300@jodybruchon.com> References: <4F3E65B3.3070401@cowlark.com> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <4F3E65B3.3070401@cowlark.com> Sender: linux-8086-owner@vger.kernel.org List-ID: Content-Type: text/plain; charset="us-ascii"; format="flowed" To: linux-8086@vger.kernel.org Unfortunately, the LIM EMS specification seems to be an API spec for programs and drivers, not a hardware interface standardization. IBM, Intel, AST, and others apparently all use a totally different driver with different interfaces. Also, based on the extreme price of EMS cards on eBay and the very small number of them I am finding, EMS support is probably a waste of time to even think about. There is no standard hardware interface, and if the hardware is so hard to find that I can't get it to test with anyway, there's little point in planning to add support for it. Unless someone has a valid objection, tonight I will remove the placeholder code for EMS support and the EMS support option which does nothing anyway. Just to be clear about what I'm doing, I'm trying to shrink the ELKS code base by stripping out vaporware and useless code, particularly since one direction we are interested in taking is porting to a different compiler, and every single line of code that's not necessary when such work begins will only be an additional burden at that time. Plus, novices who look at ELKS and happen to enable those options will see options for drivers that simply don't exist or aren't functional, and then be confused when it doesn't work. Jody Bruchon