[Samba] regpatch writing to local registry hive with -F not working (registery-utils 4.0.0~alpha15~git20110124.dfsg1-2ubuntu1)

[RiCH]

On Thu, 2011-05-19 at 09:08 +0200, denis.bonnenfant wrote:

> Wilco Baan Hofman a écrit :
> > On Wed, 2011-05-18 at 16:01 +0200, denis bonnenfant wrote:
> >   
> >> Le mercredi 18 mai 2011 à 15:07 +0200, Wilco Baan Hofman a écrit :
> >>     
> >>> On Wed, 2011-05-18 at 12:33 +0200, Michael Wood wrote:
> >>>
> >>>       
> >>>> Then it seems the -F option should be removed from regpatch.  Or
> >>>> should regpatch be replaced with something similar to the Python script
> >>>> you included below?
> >>>>         
> >> My patch adds a -K option to regpatch for specifying the predef key
> >> where -F registry should be mounted. I will submit it for review soon,
> >> I'm currently experimenting a little bit, and there are still some bugs
> >> in .reg parsing
> >>     
> >
> >   
> 
> In fact it doesn't make sense to mount regf files to anything but HKCU , 
> so -K option is not necessary, and HKCU can be hardcoded as mountpoint 
> for files specified by -F option.
> 


Sorry if I am misunderstanding something here, but if the .reg file is
modifying a software or system key shouldn't regpatch
 mount the appropriate hive,  mounting under HKLM/Software & system
under HKLM/system etc. ?

(I currently am using Ghost's linux ghregedit to do this, but I am
hoping for a license free version, ... I have been searching for a very
long time for something that actually works!)

I was envisaging a usage along these lines;
regpatch -F  /mnt/xp/WINDOWS/ myregpatch.reg #regpatch loads the hives
needed itself based on the WNIDOWS target dir (auto filling
system32/config), this approximates to how ghrededit works
..or more in keeping with how its setup at the moment
regpatch -F  /mnt/something/software,/mnt/something/system
myregpatch.reg #tell it what hives to load

I appreciate all the effort everyone is putting in. Thanks.


> 
> > I use this quite a bit and I'm aware of two bugs, which is not directly
> > related to the parsing, but that on windows unicode is implicit for
> > certain data types even when the data is given in binary format, it's
> > still converted. The other 'bug' is that unicode .reg files are not yet
> > supported. I'd be very interested to know what other bugs there are.
> >   
> I found some problems with value deletion, sometimes values are not 
> deleted, reg_expand_sz data not correctly saved,  but i'm not sure that 
> the problem is in parsing. I'm going to experiment a little bit more.
> 
> Denis