[ECOS] CONFLICTS that SHOULD be!

[ECOS] CONFLICTS that SHOULD be!

[Trenton D. Adams]

The following two items were required by the File I/O package.  When I
removed that package they were not disabled.  As a result I got an error
saying that the file or directory didn't exist.  In particular,
<cyg/posix/signal.h>.
CYGBLD_ISO_SIGNAL_NUMBERS_HEADER
CYGBLD_ISO_SIGNAL_IMPL_HEADER

The following two items were required by another package which I think
was File I/O as well.  When I removed the package they were not
disabled.  As a result I got an error saying that the file or directory
didn't exist.  In particular, <cyg/posix/limits.h> and/or
<cyg/fileio/limits.h>
CYGBLD_ISO_POSIX_LIMITS_HEADER
CYGBLD_ISO_OPEN_MAX_HEADER

To be or not to be, that is the question!

Should they not be hooked to a conflict?  If so, maybe using the old
1.3.1.2 tool caused this?  If so, no big deal, I can deal with disabling
them manually! :)  I don't use the new tool because it has that package
removing bug where it crashes.

Trenton D. Adams
Extreme Engineering
#17, 6025 - 12 St. SE
Calgary, Alberta, Canada
T2H 2K1

Phone: 403 640 9494 ext-208
Fax: 403 640 9599

http://www.extremeeng.com

Re: [ECOS] CONFLICTS that SHOULD be!

[Jonathan Larmour]

"Trenton D. Adams" wrote:
> 
> The following two items were required by the File I/O package.  When I
> removed that package they were not disabled.  As a result I got an error
> saying that the file or directory didn't exist.  In particular,
> <cyg/posix/signal.h>.
> CYGBLD_ISO_SIGNAL_NUMBERS_HEADER
> CYGBLD_ISO_SIGNAL_IMPL_HEADER
> 
> The following two items were required by another package which I think
> was File I/O as well.  When I removed the package they were not
> disabled.  As a result I got an error saying that the file or directory
> didn't exist.  In particular, <cyg/posix/limits.h> and/or
> <cyg/fileio/limits.h>
> CYGBLD_ISO_POSIX_LIMITS_HEADER
> CYGBLD_ISO_OPEN_MAX_HEADER
> 
> To be or not to be, that is the question!

Hmmm... yes, it would do that. This is an old chestnut that has caught us
out even internally before: should values be reinferred every time
something changes. Or perhaps even just when the user explicitly wants.
Right now we do neither :-|.

I think we should at least allow the user to force a reinference. However
customers tend not to pay for host tools, so very little development gets
done except in "spare time", so I don't know when that would appear. Sorry.
Of course, you have the source :).

Jifl
-- 
Red Hat, Rustat House, Clifton Road, Cambridge, UK. Tel: +44 (1223) 271062
Maybe this world is another planet's Hell -Aldous Huxley || Opinions==mine