ports/122865: astro/google-earth: wrong distfile checksum

[Juergen Lock]

In article <61974025 at bb.ipt.ru> you write:
>On Fri, 18 Apr 2008 23:09:38 +0200 Juergen Lock wrote:
>
>> Please test this update:
>> 	http://people.freebsd.org/~nox/google-earth-4.3.7191.6508.patch
>
>>  The new version seems to crash on 6.x and earlier because of missing
>> linux syscalls (I saw truncate64 in the logs), so it'll work for even
>> less ppl than before, and I got one report of
>> 	googleearth-bin: error while loading shared libraries: /usr/lib/librt.so.1: ELF file OS ABI invalid
>
>Though there is a PR about (similar) case, I'm sure it's a
>(mis)configuration case. The libruary /usr/lib/librt.so.1 is a native
>FreeBSD one. And under non (default) circumstances it should be found
>(used) before /compat/linux/lib/librt.so.1.
>
>> which may have to do with the linux_base version (the success report
>> I got was with fc4 and nvidia gfx; my only 7 box has intel gfx where
>> googleearth crashes the xserver on FreeBSD, but the old version did that
>> already...)
>
>I may report a partial success.
>
>While working at new linux-f8 ports I managed to run (current
>port version) googleearth at my nvidia card only with xorg nv driver
>and graphics/linux_dri-devel (actually it's a dri version base on
>xfree-6.5). Not fast but it works (with new f8 ports)!

Cool.  But you didn't get hardware gl that way, or did you?

> Didn't have
>time to test but seems that it should work with fc4 ports either.
>
 I have success reports for 7 using various linux_base ports now.
(Well, it still crashes once in a while but the old version did
that also...)

>Now I test your patch ant it partially works. Googleearth runs for
>some time and then freezes at state:
>-----
>load: 0.09  cmd: googleearth-bin 5616 [linuxfutex] 0.00u 0.00s 0% 86476k
>-----
>
 Hmm, are you running 7 or HEAD?

>And I get a console message:
>-----
>linux: pid 5554 (googleearth-bin): syscall truncate64 not implemented
>-----
>
 Oh, so that one is also missing in 7?

>Both cases seems nothing to do with the port system but a kernel one.

 True.  I guess we have no choice but updating anyway since google
only gives out the new distfile now...

 Thanx,
	Juergen