Hi > Is Haiku going to maintain two stacks (IDE and ATA)? If it will not > (which I suspect to be the case, but correct me if I am wrong), > providing an alternative stack to cater to those unlucky ones whose > hardware does not work with the default stack will not help the > purpose > of the alpha/beta releases, which is (should be) to iron out the bugs > of > whatever components you plan to include in the final R1 release. Absolutely. The only thing holding back the switch really is that there is known hardware that doesn't work with ATA. While this is also true for IDE, IDE has been tested for a long time and we know what type of hardware is affected. With ATA, since it hasn't seen such widespread usage, we have no real clue what hardware it will fail on. The fear basically boils down to that switching in the last minute will suddenly reveal that ATA fails on more hardware than IDE. But you are right that we should just finally get over that and switch to ATA, as that's what we want to include in R1 afterall. In that sense, I would vote for setting HAIKU_ATA_STACK to 1 for the alpha build profiles now so that the nigthlies will get it and the use becomes more widespread. I would further vote against producing separate images for both stacks, as I could imagine it to be rather confusing. If it turns out that ATA has too many issues we will find a way to include both stacks and make them switchable by the means of a safemode option. I agree on producing all types of images and publishing them, meaning raw ones, vmware and ISOs. Regards Michael