On 12/27/2012 18:08, Daniel Colascione wrote:
Yes, we have to accept the minuses we can't easily fix, but this
one, we can can. Porting the Unix-mode temporary file logic to the
Windows build seems workable enough.
"Who is this 'we,' kemosabe?"
You think it's easy, try it.
You'll find that you can't simply make a one-line patch that swaps a
unixGetTempname() call in for an osGetTempPathW() call. It compiles and
then fails to link, because unixGetTempname() is ifdef'd out 5000 lines
away from the actual definition. You end up duplicating about 80 lines
of code, simply because that's far easier than doing the ifdef
gymnastics to reuse what's already there.
We aren't going to tell Mr. Hipp about this patch, lest he barf all over
his shoes. (Shhh.)
Here's the patch, for those with strong stomachs:
http://etr-usa.com/cygwin/sqlite3/src.patch
If you can come up with a cleaner patch, by all means, lay it on me.
Here are the packages built with it, for testing:
http://etr-usa.com/cygwin/sqlite3/libsqlite3-devel-3.7.15.1-1.tar.bz2
http://etr-usa.com/cygwin/sqlite3/libsqlite3_0-3.7.15.1-1.tar.bz2
http://etr-usa.com/cygwin/sqlite3/sqlite3-3.7.15.1-1-src.tar.bz2
http://etr-usa.com/cygwin/sqlite3/sqlite3-3.7.15.1-1.tar.bz2
Relative to the current 3.7.13-1 packages on the Cygwin repo mirrors,
this also changes:
- Upgrade to 3.7.15.1 upstream
- Enable a bunch of build options others wanted: column data, FTS...
_______________________________________________
sqlite-users mailing list
sqlite-users@sqlite.org
http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
