On Sun, Jun 06, 2004 at 02:44:32PM +0100, Paul Duffy wrote: > NTFS read capability compiled in by default. Won't happen anytime soon due to the concerns mentioned elsewhere in this thread. > Make Wine work again dammit. The problem here is prelinking/address space randomisation. Wine wants to load Windows libraries and such at absolute addresses and prelink might have already reloaded stuff there. If you disable prelink, it'll likely work again. > An actual option (ie: one which doesn't break a gazillion dependencies) to > make a non-preemptable kernel so kmail stops hogging my damn CPU. It already is non-preemptable. Assuming you meant the opposite, and want us to enable CONFIG_PREEMPT, it's very unlikely to happen. It's a marginal gain for a handful of uses, but brings a whole world of pain for a lot of other users. > 'Probe all SCSI LUNs' compiled in by default so my 6-in-1 card reader can read > something other than Sony Memory Stick without a recompile. Won't happen. Having this as default will break on some devices. It locks up the SCSI bus of badly designed devices. If you send the relevant info from dmesg, I'll add it to the list of devices recognised, and it'll automatically be whitelisted. Longterm, things may change for these devices so the USB layer recognises the type of device, and does the necessary magic automatically. > I feel these are mostly minor fixes (like the kernel defaults). A lot of people don't realise, it's very easy to pick a set of kernel defaults that make everything work 'out of the box' for one particular system. Trying to make a kernel that works out of the box for every possible system we target, is a very different problem space. For every reason to enable an option, there's usually a downside, and a reason not to enable it. Sometimes it's a question of balance, sometimes there are hard and fast rules (like 'it breaks vendor x's gizmo') Dave
