[haiku-development] Re: R1A3: Image types to release?

  • From: "Michael Lotz" <mmlr@xxxxxxxx>
  • To: haiku-development@xxxxxxxxxxxxx
  • Date: Mon, 30 May 2011 19:23:43 +0200

> From: "Matt Madia" <mattmadia@xxxxxxxxx>
> > At the time of R1A2, anyboot was still fairly new.
> > Now, we know that anyboot works, but will be burned improperly by 
> > some
> > burning applications.
> Curently anyboot does not works (except for some few PC).
> IMHO : burning-applications are not guilty. however the MBR-code is.

Just for completeness sake: Anyboot worked on pretty much every single 
machine I tried it on. So from my perspective anyboot works on most 
machines (10+ systems from various generations with various chipsets I 
could get my hands on), except a few (the one originally reported in 
the quoted bug and the two you mentioned).

Please try to be less black/white, as such statements are hurtful (it 
implies that whoever implemented something didn't test his work at 
all). Also always remember that there simply are a huge number of 
completely different systems and configurations. To judge 
compatibility/support from the few systems one usually has access to is 
simply not a good idea. That's also why we absolutely require bug 
reports if something doesn't work. The assumption that it works for 
whoever implemented it, on his specific configuration(s), holds true 
more often (talking from my own experience at least).

> A fix exists
> http://dev.haiku-os.org/ticket/3441#comment:21
>   but it needs to be commited...

What would be really great is if you could point out what exactly it 
was that you fixed. The diff, since it converts the source between the 
assembler formats, isn't very helpful to see the actual change. You 
obviously figured out a specific issue and fixed it, so what was it? 
Knowing that might even allow fixing one of the existing MBR codes so 
that the GPT features don't need to be removed/recreated again.


Other related posts: