On Tue, Sep 11, 2007 at 04:26:14PM +0000, Johannes Schindelin wrote:
> Hi,
>
> On Tue, 11 Sep 2007, Pierre Habouzit wrote:
>
> > On Tue, Sep 11, 2007 at 03:03:21AM +0000, Johannes Schindelin wrote:
> > > static const struct transport_ops bundle_transport = {
> > > + /* set_option */ NULL,
> > > + /* get_refs_list */ get_refs_from_bundle,
> > > + /* fetch_refs */ fetch_refs_from_bundle,
> > > + /* fetch_objs */ NULL,
> > > + /* push */ NULL,
> > > + /* disconnect */ close_bundle
> >
> > Not that I care that much, and I don't know what is done in git
> > usually. But aren't C99 initializer safer ? (wrt struct transport_ops
> > API possible changes in the future).
>
> No, they are not, since we do not allow such backwards incompatibilifiers
> to creep into git's code.
>
> Yes, I know some would prefer the latest and greatest, and another
> language altogether, but keep this in mind: what if somebody made git
> incompatible with _your_ setup? So really, don't do it to others that
> lightly.
It has nothing to do with latest. I feel that if for some reason
transport_ops need a new function, it will break silentely, whereas if
you use:
static const struct transport_ops bundle_transport = {
+ .set_option = NULL,
+ .get_refs_list = &get_refs_from_bundle,
+ .fetch_refs = &fetch_refs_from_bundle,
+ .fetch_objs = NULL,
+ .push = NULL,
+ .disconnect = &close_bundle
Or even:
static const struct transport_ops bundle_transport = {
+ .get_refs_list = &get_refs_from_bundle,
+ .fetch_refs = &fetch_refs_from_bundle,
+ .disconnect = &close_bundle
It's pretty straightforward to extend the transport_ops API, afaict
it's what the kernel does e.g.. I wasn't suggesting anything else.
--
·O· Pierre Habouzit
··O madcoder@xxxxxxxxxx
OOO http://www.madism.org
Attachment:
pgpsXXZUy2pdy.pgp
Description: PGP signature
