Hi, On Tue, 9 Jan 2007, Jakub Narebski wrote: > Johannes Schindelin wrote: > > > On Sun, 7 Jan 2007, Junio C Hamano wrote: > > > >> Johannes Schindelin <Johannes.Schindelin@xxxxxx> writes: > >> > >>> Further, it makes no sense to force the parsing upon the helper > >>> functions; for_each_reflog_ent() now calls the helper function with > >>> old and new sha1, the email, the timestamp & timezone, and the message. > >> > >> Perhaps. I did it that way deliberately because all existing > >> users did not have to pay overhead of parsing. > > > > My reasoning is that invalid entries should rather be ignored than be > > taken into account. So, to verify that you are not walking invalid data, > > you have to parse it anyway. > > I think that Junio was talking about the fact, that if you for example > do need only refname and sha1, there is no need to parse object at all. > If you don't need to, don't parse. And it was exactly what _I_ was talking about, too: if there are invalid entries, you ignore them. So for example, if there is no timestamp and message, you don't want the osha1 or nsha1, because it is an _invalid_ record. Ciao, Dscho - To unsubscribe from this list: send the line "unsubscribe git" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html