Re: make install

At 10:34 AM -0400 10/4/04, Andy Dougherty wrote:
>On Mon, 4 Oct 2004, Dan Sugalski wrote:
>
>>  At 10:08 PM +0100 10/3/04, Nicholas Clark wrote:
>>  >On Sat, Oct 02, 2004 at 10:11:34PM -0400, Andy Dougherty wrote:
>>  >>  On Sat, 2 Oct 2004, Nicholas Clark wrote:
>>  >>
>>  >>  > parrot has had a make install target for quite a while, which
>>  >>seems to work.
>>  >>
>>  >>  That's surprising -- At one time I thought I had hardwired the ICU data
>>  >>  directory to the build directory as a "temporary" measure to 
>>get things at
>>  >>  least sort-of working.  Perhaps someone's done it correctly since.
>
>>  >>  Does it really work even if the build directory is gone?
>
>>  >ponie really worked if installed, but ponie is currently only using
>>  >parrot's data structure code. I doubt that it's going anywhere near ICU
>
>>  If Parrot even started then it worked out. If the data directory's
>>  not there, things die on startup.
>
>I really don't see how it could possibly work.  Are you *sure* you moved
>the build directory out of the way?  Are you *sure* you don't have the
>environment variable PARROT_ICU_DATA_DIR set pointing somewhere?  Can run
>truss (or trace or whatever it's called on your system) and verify it
>really is picking up the installed icu directories?

It's also possible that the build used a system-installed ICU, in 
which case you're fine no matter what.
-- 
				Dan

--------------------------------------it's like this-------------------
Dan Sugalski                          even samurai
[email protected]                         have teddy bears and even
                                       teddy bears get drunk