On Sun, Mar 02, 2025 at 09:36:24PM +0100, Volker Schlecht wrote:
> On 2025-03-02 17:47, Jeremie Courreges-Anglas wrote:
> > Confirmed.  I checked that the same -D/#defines set was used before
> > and after.  In the new version some are passed on the cli and some end
> > up in sqlite_cfg.h.  Looks mostly good, fts3 is indeed implied by
> > fts4, but geopoly is no longer implied by --enable-rtree.  So unless
> > you have a good reason to leave it out, I suggest you add
> > --enable-geopoly to match the current feature set.
> 
> I don't have a good reason to leave geopoly out, but in that case it would be
> altogether simpler to just --enable-all, no?

I didn't know about --enable-all.  Explicitely listing what we use is
a way to play it safe. Since --enable-all is explicitely documented in
https://sqlite.org/src/doc/trunk/doc/compile-for-unix.md one can
hope it's safe... I'm fine with you replacing
                --enable-rtree \
                --enable-geopoly \
                --enable-fts4 \
                --enable-fts5 \
                --enable-session
with --enable-all.  The result:

Feature flags...
  + fts4
  + fts5
  + geopoly
  + rtree
  + session
  - update-limit
  - memsys5
  - memsys3
  - scanstatus
  + json
Library feature flags: -DSQLITE_ENABLE_COLUMN_METADATA -DSQLITE_ENABLE_DBSTAT_VTAB -DSQLITE_ENABLE_FTS4 -DSQLITE_ENABLE_FTS5 -DSQLITE_ENABLE_GEOPOLY -DSQLITE_ENABLE_MATH_FUNCTIONS -DSQLITE_ENABLE_PREUPDATE_HOOK -DSQLITE_ENABLE_RTREE -DSQLITE_ENABLE_SESSION -DSQLITE_ENABLE_UNLOCK_NOTIFY -DSQLITE_THREADSAFE=1
Shell options: -DHAVE_READLINE=1 -DSQLITE_HAVE_ZLIB=1


My 2 cents,
-- 
jca