Re: [Etherboot-developers] Splashimage / Bootlogo
Brought to you by:
marty_connor,
stefanhajnoczi
|
From: Anselm M. H. <an...@ho...> - 2002-06-09 15:18:43
|
> I would prefer that effort was put into an external splash/menu program > instead of patching Etherboot. The stuff inside bootmenu.c and ansiesc.c > is ugly and hard to update as binaries are in boot ROMs. Correct. I'd like that too, but as my site anyway uses chaining (etherboot 5.05 loads etherboot 5.06) - we burnt the roms before exactly knowing what we wanted :-( - updating roms is not our problem. I wanted to tell you about a not-to-difficult solution. Just dump it, that's ok. It was a hint. > Within a few days I will release mknbi-1.2.8pre1, which will contain an > external menu program that has the code that does the menuing in > Etherboot. In tests, the menu loaded instantly and all the initial text > were cleared off. When control is handed back to Etherboot you get more > loading messages, What my patch could hinder Etherboot from. >but in my opinion that's not to great a drawback. It's an issue if you want the graphics to stay 'til X starts up; originally I think that was the task. > If it's an issue, some sort of VT switching might be devised so that the > splash/logo remains on the screen. The main thing is a user-friendly > screen and with an external program you are free of size limits on > Etherboot, you have all of the lower 512kB if you wish. At this point, it is probably much easier to user -DSILENTFORSPLASH (which keeps any output after loading e.g. the menu.nbi from appearing) than implementing VT switching - see, many old 386 consoles will happily run with 640x350x16, but there is not enough RAM on them to have a second console of that format at hand. VESA will be no question if you want to display boot graphics on these machines too. > So I will not be accepting this patch as it will be obsoleted or at > least be interfered with by developments on the external program. But please keep it in mind, just in case! > [...] and nested TFTP files [...] Are not used here. I don't use any RAM neither (except the buffer from tftp, but that is handled by the download code or ansiesc, not by my patch) as the graphics is immediately output' to ansiesc. No matter, if boot program does the image output, my patch would keep the screen clear afterwards. And that's what <devzero>, who originated that thread, asked for. Anselm |