Slashdot Mirror
Free/Open Source Software Hardware Requirements?
Bender asks: "Most on Slashdot seem to be concerned with getting Free/Open Source software to be compatible with hardware (firmware, register sets, etc). My question is from the other side of the table: I'm in the hardware business and I'm wondering if there are any central guidelines to better guarantee compatibility with Linux/*BSD. As an example, to guarantee that our hardware runs Microsoft Windows, we have to conform to the Windows Logo Program Requirements. These requirements dictate (among other things) firmware interfaces, debug ports, and DRM. Some of these requirements, if not implemented carefully, could trigger incompatibilities with non-Microsoft operating systems. Is there a Linux/*BSD equivalent to the Microsoft requirements to allow hardware designers to build OS agnostic systems?"
... send a free piece of the hardware to every major kernel-programmer.
If you mod this up, your slashdot background will turn into a beautiful sunset!
Documentation available under NDA is useless for open source (publishing the source itself will likely break the NDA).
Stephan
maybe GNU could launch a documentation database with docs from the manufacturers.
I fail to see why office or home applications should dictate a particular architecture. Gaming and lab work are probably the only things which may be picky, due to bus speeds. The AMD64 is a nice start, but when can we exepct some of the other housecleaning of PC design? All I've got on my desk at home is very souped up PC-XT. Meanwhile some really good architecture has died along the way as everyone fought to support quite possibly the most exasperating legacy beast, just like everyone else. Moo.
A feeling of having made the same mistake before: Deja Foobar
Insightful, this? People, the o.p. asked a sensible question which shows they are serious about adding support for Linux, and all you can poke out of your ass is a dig at Microsoft. If you can't help, then leave alone!
POSIX.1 ... nuff said.
As for hardware specs... on an i386 class box [which includes pretty much everything even the x86_64] the PCI bus is accessible through MMIO and I/O ports.
This doesn't change just because you're in BSD or Linux.
Sure the actual code may change due to the organization of the respective kernels but the hardware manual which explains how the device works is applicable to both.
The biggest problem with most hardware is
1. Undocummented hardware
2. "pointless revisions" [*]
[*] A typical ploy is to change the mapping on a wifi card so that version X-1 drivers don't work. It's not that the developer actually fix things or improve it otherwise just to make sure the OSS crowd is always at least one step behind.
Tom
Someday, I'll have a real sig.
Maybe, but considering that Linux and friends have been accelerating in marketshare growth and visibility, the compatibility issues will become a greater priority for the OEMs and it will be the companies' employees who will take care of what would otherwise have been the dirty hippies' project. Just another instance of supply catching up with demand.
I can say that the Logo certification requires a significant amount of technical work, not just some buy off. So yes, the logo does actually mean something on a solid, technical level with respect to accomodating Windows and working with the Windows environment.
I do not do the work, but I have had products I'm working on impacted by some pretty low-level technical changes on the product required to meet WHQL from other groups.
XML is like violence. If it doesn't solve the problem, use more.
Diversity does not "hold open source back." It only does so in the minds (and I use the term "mind" rather loosely, here) of chimps like yourself, who've been brought up on Microsoft's "Eine Reich, Eine Volk, Eine OS," rhetoric. The thing that makes this worse, however is the fact that people don't advocate a monoculture because it's in any way a good thing...it most certainly isn't. They advocate it purely because they've been brainwashed by Microsoft to do so.
If diversity holds open source back in any manner, it does so only because those who desire a monoculture desire one precisely for the reason that a monoculture allows them to avoid having to think...which as we all know, is the one thing most human beings will do virtually anything in order to avoid having to do. A monoculture means that people who actually deliberately and consciously desire to be stupid lemmings are empowered to do so. If there's only one choice, you don't need to put any thought into making said choice. It is only those people that either have an active desire to avoid using their brains, or a fear of personal responsibility, who do not want choice. Unfortunately, I'm aware that that constitutes 98%-99% of the human population. You yourself however have a choice as to whether you also wish to be a lemming, and join the others in their journey over the cliff, or whether you choose to be self-determining. It's a case of the eternal red pill vs blue pill question again.
My answer to this hardware developer would be that Linux (or not just Linux - operating systems in general) primarily needs peripherals which talk to the rest of the hardware in a relatively straightfoward and sane way...unlike winmodems as probably the best example which slave off the CPU, and do so in an undiscoverable and intentionally proprietary/closed manner. Hardware shouldn't be designed to keep secrets...its purpose and means of performing its tasks should be as easily visible as possible. The more people who can figure out how the hardware works, or at the very least how to relatively easily adapt it for their particular operating system, the wider the potential adoption of said hardware will be, and the more unit sales and money you as a hardware developer will make.
Strange, I dual boot both XP and Slackware, and I have never had a problem with either of them that was not directly related to a mistake of my own.
I'm not easy on my machine either - I'm both an avid developer and gamer.
Here's how I see it, NDA is bad, nothing done under an NDA is worth using. Because if you have work done under the NDA and you stop working on it, noone else has the documentation you signed an NDA for and therefore cannot maintain the code.
NDA work being released is almost as bad as a binary.
I'm sick of following my dreams - I'm just going to ask them where they're going and hook up with them later.