Re: Opening a thread in gmail



On Sun, Aug 30, 2009 at 8:40 PM, John
Ralls<jralls ceridwen fremont ca us> wrote:
>
> On Aug 30, 2009, at 6:21 PM, Chip Parker wrote:
>
> Well, the biggest reason why I didn't want to use the ML for this
> thread is because I'm still figuring things out and prefer not to look
> like a complete idiot due to entirely unexpected/unintended behavior.
> It also seems like _libiconv has been changed to _iconv in 10.6, which
> makes me scratch my head.
>
> http://developer.apple.com/mac/library/documentation/Darwin/Reference/ManPages/man3/iconv_open.3.html#//apple_ref/doc/man/3/iconv_open
>
> http://bugs.php.net/bug.php?id=49267
> <config.log>
>
> Chip,
> No reason to feel embarrassed about sticking your foot in it. Everyone does,
> just be glad you've provided an example for others to google and move  on.
> I don't see any difference between that man page and the one I get when I
> `man iconv_open` on Leopard. The PHP bug report doesn't make a lot of sense,
> either. Libiconv is a GNU project, you can look through their git repo
> at http://git.savannah.gnu.org/cgit/libiconv.git. I briefly browsed through
> the Changelog and didn't see anything that would match the problem being
> reported. "libiconv" is a weird thing to put in a source file anyway. The
> bug reporter was obviously not thinking very hard, either: Defining a macro
> to return it's name is silly.
> I'm sure there will be issues with 10.6, but I'm not likely to start testing
> on it for a couple of weeks; there's plenty of other stuff in the queue.
> Regards,
> John Ralls
>

A heads-up if you're planning on doing any 10.6sdk stuff at all:
hfs_format.h has...

761 typedef struct JournalInfoBlock JournalInfoBlock;
762 struct JournalInfoBlock {
763     u_int32_t   flags;
764     u_int32_t       device_signature[8];  // signature used to
locate our device.
765     u_int64_t       offset;               // byte offset to the
journal on the device
766     u_int64_t       size;                 // size in bytes of the journal
767     uuid_string_t   ext_jnl_uuid;
768     char            machine_serial_num[48];
769     char        reserved[JIB_RESERVED_SIZE];
770 } __attribute__((aligned(2), packed));
771
772
773 enum {
774     kJIJournalInFSMask          = 0x00000001,
775     kJIJournalOnOtherDeviceMask = 0x00000002,
776     kJIJournalNeedInitMask      = 0x00000004
777 };

Cairo fails to build, citing line 767 (probably line 765 on your box,
since I added a couple defines above trying to fix the problem.) I'm
no C coder, but when I see "error: expected specifier-qualifier-list
before 'uuid_string_t'" that means something's wrong. By copying in
_just_ include/hfs from 10.5, everything's golden.



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]