[PATCH 28/28] makes upload_pack void

[PATCH 28/28] makes upload_pack void

[David Rientjes]

Makes upload_pack void and removes conditional return.

		David

Signed-off-by: David Rientjes <rientjes@google.com>
---
 upload-pack.c |   11 +++++------
 1 files changed, 5 insertions(+), 6 deletions(-)

diff --git a/upload-pack.c b/upload-pack.c
index bbd6bd6..27e2abe 100644
--- a/upload-pack.c
+++ b/upload-pack.c
@@ -459,18 +459,17 @@ static int send_ref(const char *refname,
 	return 0;
 }
 
-static int upload_pack(void)
+static void upload_pack(void)
 {
 	reset_timeout();
 	head_ref(send_ref);
 	for_each_ref(send_ref);
 	packet_flush(1);
 	receive_needs();
-	if (!want_obj.nr)
-		return 0;
-	get_common_commits();
-	create_pack_file();
-	return 0;
+	if (want_obj.nr) {
+		get_common_commits();
+		create_pack_file();
+	}
 }
 
 int main(int argc, char **argv)
-- 
1.4.2.g89bb-dirty

Re: [PATCH 28/28] makes upload_pack void

[Nikolai Weibull]

On 8/14/06, David Rientjes <rientjes@google.com> wrote:
> Makes upload_pack void and removes conditional return.

> -static int upload_pack(void)
> +static void upload_pack(void)

I don't know for sure, but I'm guessing the intention was to be able
to return a failing code /if/ there ever was a condition where
upload_pack() would fail, e.g., if send_ref() would return a status
code instead of die():ing if it can't parse the given sha1.  In a
future libification, the change of return type may have to be
reverted.

  nikolai

Re: [PATCH 28/28] makes upload_pack void

[David Rientjes]

On Tue, 15 Aug 2006, Nikolai Weibull wrote:
> I don't know for sure, but I'm guessing the intention was to be able
> to return a failing code /if/ there ever was a condition where
> upload_pack() would fail, e.g., if send_ref() would return a status
> code instead of die():ing if it can't parse the given sha1.  In a
> future libification, the change of return type may have to be
> reverted.
> 

Of course.

If upload_pack were modified to return an error code based on a specific code 
path, I trust the implementer would know how to change void to int.

		David

Re: [PATCH 28/28] makes upload_pack void

[Nikolai Weibull]

On 8/15/06, David Rientjes <rientjes@google.com> wrote:
> On Tue, 15 Aug 2006, Nikolai Weibull wrote:
> > I don't know for sure, but I'm guessing the intention was to be able
> > to return a failing code /if/ there ever was a condition where
> > upload_pack() would fail, e.g., if send_ref() would return a status
> > code instead of die():ing if it can't parse the given sha1.  In a
> > future libification, the change of return type may have to be
> > reverted.

> Of course.
>
> If upload_pack were modified to return an error code based on a specific code
> path, I trust the implementer would know how to change void to int.

So do I.  However, I trust that whoever implemented send_ref() knew
about void.  (See how easy it was to do what you did but the other way
around?)

It was just a comment.  I don't have anything against the patch as such.

  nikolai