Messages in this thread | | | Date | Sat, 19 Sep 1998 10:01:08 -0400 (EDT) | From | Gregory Maxwell <> | Subject | Re: Linux, UDI and SCO. |
| |
On Sat, 19 Sep 1998, Nathan Hand wrote:
> At least some Linux drivers have been created by reverse engineering > the binary-only drivers. In some extreme cases, people have sat down > with oscilloscopes and worked out what the hardware is doing.
Yep, some even used hacked DOSEMUs..
> So one immediate benefit of UDI is that it makes reverse engineering > much easier. The UDI layer can report a lot of what is going on, and > having an already working Linux driver reduces the reboot cycle.
Not nessassarly true. With the new funkey copyright laws that are being passed in the US, it may soon be illegial to reverse engineer drivers entirely.
It's clearly much more 'legal' to take the hardware and with only it, figure out how it works, they can't claim we hacked up their driver. Many reverse engineers laws say that it's okay for compatibility, with UDI we have 'compatibility' thus limiting further reverse engineering..
> The danger, as people have already pointed out, is that UDI can give > the hardware manufacturer an excuse not to release specs. I see that > the benefits of UDI outweigh this (pessimistic) possibility.
I see almost no beinfits. With curren't trends all hardware will have open specs within at least 10 years. UDI derails this.
> Other issues - technical, political, licensing, cost - are all of no > interest to the user who can't use their hardware under Linux.
The user can exchange their hardware. But if fundimental parts of most system's become tied to binary only drivers then they can't modify their entire system's software. That alone outweigs any possible UDI benifits. Futhermore, these drivers could cost money.
I'd rather have 1000 people exchangeing hardware, then 1000 people running binary drivers that cause random lockups, making Linux look bad.
Close drivers are bad, plain and simple. Were I Linux I would forbid them entirely.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
| |