On Thu, 4 Oct 2012, Ulrich Wilkens wrote:
> Maybe I should describe what I've done so far in a separate mail thread
> and maybe I should also commit an intermediate code level which compiles
> but doesn't run completely. Any suggestions?
Definitely, yes! I guess you have copied over .cf files f
On Thu, 4 Oct 2012, Ulrich Wilkens wrote:
> On 10/04/12 10:17, Marcin Cieslak wrote:
>
>>> #define BuildDtInfo NO or YES, and handle it like other
>>
>> Nice.
>>
>> Of course a proper solution should be for other systems as well
>> if there is a need. I just wonder whether it will not be faster
>>
On 10/04/12 10:17, Marcin Cieslak wrote:
>> #define BuildDtInfo NO or YES, and handle it like other
>
> Nice.
>
> Of course a proper solution should be for other systems as well
> if there is a need. I just wonder whether it will not be faster
> just to make dtinfo work - I didn't give it a try ye
On Wed, 3 Oct 2012, Jon Trulson wrote:
> On Wed, 3 Oct 2012, Marcin Cieslak wrote:
>
> I guess my question on this one is - why do it differently for FBSD?
It's just a quick way to get correct packing list which we might
need to register binary packages with the OS at the end of the install.
>
On Wed, 3 Oct 2012, Marcin Cieslak wrote:
I guess my question on this one is - why do it differently for FBSD?
Why not define, perhaps in the 'cf's a
#define BuildDtInfo NO or YES, and handle it like other
flags... Default it to YES in Imake.tmpl maybe.
Linux and the BSD's can declare it to NO
- Add new CPP flag HAVE_DTINFO for
all architectures except FreeBSD
- CDE-INFOLIB-LOC.src, CDE-LOC.src, CDE-MSG-LOC.src
CDE-SHLIBS.src have dtinfo related portions
excluded by #ifdef HAVE_DTINFO
- Implement new file type "ignore" in the UDB
to allow the installer to skip some files.
Fo
