I am planning to release SYSLINUX 2.08 in the next day or so; basically 2.08-pre9 without any further changes. So, I would appreciate any and all testers for this... -hpa
maybe the bug I told about a few days ago. However there are too many programs involved to be sure that this is a Memdisk bug.. (solely, or at all) the following is an image of a diskette I've made as testcase. diskette contains: *Syslinux 2.08pre9 *corresponding Memdisk *FreeDOS kernel 2032a *Himem64.exe (xms-driver up to a few GB) *dos=high this disk works (presented option 2). However when loading the same using Memdisk (option 1) it fails. http://www.darklogic.org/fdos/beta9/minicd/memdisk.bmp (36KB) shows option1 results within Vmware (and on my dad's computer) diskette: http://www.darklogic.org/fdos/beta9/minicd/boot.img [1.44MB] any feedback/suggestions welcome regards, Bernd Blaauw
sorry for the private msg - was unintentional and a consequence of my school's exchange-server don't let this issue stop you from releasing anything. I'm having more and more problems in trying to reproduce this buggy behaviour. Strange.. Bernd
>Does it ever happen with the traditional HIMEM.SYS?yes (himem.exe version 2.07 of Feb 13 2003). But the problem is that the freedos project has been using 2 drivers (fdxms.sys versions, and himem.exe versions). I'll have to do some extensive testing. So many variables.. *fdos-kernel *isolinux/memdisk version. *himem that's at least 2^3 testing paths. my guess is I need to find the oldest working combination and then update components. Not any combination is guaranteed to work, unfortunately. a "hey, it WAS good and now it's broken" message is even more useful. One thing I can really eliminate is not using Memdisk. It seems to be a key factor in exposing the behaviour I've been experiencing, although it probably isn't the source of it. oh well. Shall we abandon this issue? I've gone back to using the FDXMS driver. any suggestions still welcome though. goodnight (3:37 here), Bernd Blaauw