Re: [Suspend2-devel] Re: [ 00/10] [Suspend2] Modules support.

[Nigel Cunningham <ncunningham@cyclades.com>]

Hi.

On Saturday 04 February 2006 18:53, Pavel Machek wrote:
> This was personal email. It is pretty rude to post it to public lists.
>
> > But hey, you seem to be bent on not having it - and you seem to be the
> > one making that calls, so the rest of us that just want to use the
> > notebooks properly will have to patch until someone decides that
> > having something that works is more important than being right all the
> > time.
>
> In the end, it is important what is right, not what works. If you do
> not understand that -- bad for you.

True, but in something like putting code in the kernel vs in userspace, 
people apply different criteria in determining what is right. God hasn't 
written "You shall do suspend to disk in userspace" (and we'd probably 
rebel against Him anyway if He did :>), so we have to figure out what the 
best way is. Is userspace the 'right' solution? Well, yes, it does let you 
add features without adding to kernel code. But it also creates other 
problems. Putting it in kernel space has issues too - some things like 
userui are best left where they won't necessary take down the whole 
process if they don't work right. Personally, I think we're getting too 
polarised here. I've already accepted that there's a space for userspace 
code by merging (into Suspend2) code that puts the user interface there, 
along with management of storage. You agree that somethings, such as the 
atomic copy, simply can't be done in userspace. Without having looked 
seriously at the code yet, I'd be pretty sure that you're also leaving the 
calculation of what pages to store in the kernel. That just leaves how to 
store the image. Aren't we actually a lot closer than it has appeared?

Nigel

Attachment: pgp16ZR5IJDiV.pgp
Description: PGP signature