69 messages in com.selenic.mercurial-develRe: RFC: dealing with dead, anonymous...| From | Sent On | Attachments |
|---|---|---|
| Martin Geisler | May 7, 2010 10:33 am | |
| Dirkjan Ochtman | May 7, 2010 10:37 am | |
| Matt Mackall | May 7, 2010 10:59 am | |
| Martin Geisler | May 7, 2010 11:14 am | |
| Matt Mackall | May 7, 2010 11:46 am | |
| Dirkjan Ochtman | May 7, 2010 1:14 pm | |
| Gilles Moris | May 8, 2010 12:10 am | |
| Martin Geisler | May 8, 2010 1:31 am | |
| Martin Geisler | May 8, 2010 1:34 am | |
| Martin Geisler | May 8, 2010 1:48 am | |
| Martin Geisler | May 8, 2010 2:09 am | |
| Gilles Moris | May 8, 2010 3:50 am | |
| Martin Geisler | May 8, 2010 1:10 pm | |
| Gilles Moris | May 9, 2010 9:33 am | |
| Gilles Moris | May 9, 2010 1:49 pm | |
| Gilles Moris | May 9, 2010 1:49 pm | |
| Gilles Moris | May 9, 2010 1:49 pm | |
| Gilles Moris | May 9, 2010 1:49 pm | |
| Gilles Moris | May 9, 2010 1:49 pm | |
| Matt Mackall | May 9, 2010 2:36 pm | |
| Matt Mackall | May 9, 2010 2:43 pm | |
| Matt Mackall | May 9, 2010 2:45 pm | |
| Matt Mackall | May 9, 2010 2:47 pm | |
| Martin Geisler | May 9, 2010 4:11 pm | |
| Dirkjan Ochtman | May 9, 2010 10:50 pm | |
| Gilles Moris | May 9, 2010 11:31 pm | |
| Gilles Moris | May 9, 2010 11:49 pm | |
| Matt Mackall | May 10, 2010 12:04 am | |
| Gilles Moris | May 10, 2010 12:37 am | |
| Gilles Moris | May 10, 2010 12:37 am | |
| Matt Mackall | May 10, 2010 12:38 am | |
| Matt Mackall | May 10, 2010 12:58 am | |
| Sune Foldager | May 11, 2010 2:02 am | |
| Sune Foldager | May 11, 2010 2:08 am | |
| Sune Foldager | May 11, 2010 2:14 am | |
| Sune Foldager | May 11, 2010 2:20 am | |
| Matt Mackall | May 11, 2010 3:53 pm | |
| Matt Mackall | May 11, 2010 3:56 pm | |
| Gilles Moris | May 12, 2010 12:52 am | |
| Gilles Moris | May 12, 2010 1:15 am | |
| Gilles Moris | May 12, 2010 1:15 am | |
| Gilles Moris | May 12, 2010 1:15 am | |
| Gilles Moris | May 12, 2010 1:15 am | |
| Gilles Moris | May 12, 2010 1:15 am | |
| Gilles Moris | May 12, 2010 5:17 am | |
| Augie Fackler | May 12, 2010 7:35 am | |
| Matt Mackall | May 12, 2010 8:33 am | |
| Matt Mackall | May 12, 2010 8:34 am | |
| Matt Mackall | May 12, 2010 8:34 am | |
| Matt Mackall | May 12, 2010 8:55 am | |
| Gilles Moris | May 12, 2010 2:33 pm | |
| Gilles Moris | May 13, 2010 7:31 am | |
| Matt Mackall | May 13, 2010 1:48 pm | |
| Martin Geisler | May 17, 2010 1:09 am | |
| Martin Geisler | May 17, 2010 1:22 am | |
| Rafael Villar Burke (Pachi) | May 17, 2010 2:07 am | |
| Augie Fackler | May 17, 2010 5:57 am | |
| Martin Geisler | May 17, 2010 6:56 am | |
| Augie Fackler | May 17, 2010 7:46 am | |
| Martin Geisler | May 17, 2010 8:13 am | |
| Augie Fackler | May 17, 2010 8:41 am | |
| Gilles Moris | May 19, 2010 8:00 am | |
| Martin Geisler | May 19, 2010 3:53 pm | |
| Martin Geisler | May 19, 2010 3:58 pm | |
| Augie Fackler | May 24, 2010 5:15 am | |
| Peter Arrenbrecht | May 26, 2010 6:23 am | |
| Rafael Fernández López | Jun 3, 2010 5:35 am | |
| Peter Arrenbrecht | Jun 3, 2010 6:14 am | |
| Martin Geisler | Jun 4, 2010 12:18 am |
| Subject: | Re: RFC: dealing with dead, anonymous feature branches | |
|---|---|---|
| From: | Peter Arrenbrecht (pete...@gmail.com) | |
| Date: | May 26, 2010 6:23:05 am | |
| List: | com.selenic.mercurial-devel | |
On Fri, May 7, 2010 at 8:00 PM, Matt Mackall <mp...@selenic.com> wrote:
On Fri, 2010-05-07 at 19:37 +0200, Dirkjan Ochtman wrote:
On Fri, May 7, 2010 at 19:34, Martin Geisler <mg...@aragost.com> wrote:
I discussed this idea with Jan Sørensen from aragost and I had at some point mentioned that we're "afraid" of many heads because the number of heads is a term in the running time of some of our protocols. He suggested that one could make a "canonical dead head", that is, merge all dead heads until there is only a single dead head in the graph.
I think this is the important bit, it potentially alleviates a bunch of perf issues. We should get this before we argue about the other things.
So what happens when someone resurrects a dead head by committing to it? Answer: merge is confused.
There are three categories of algorithm that are affected by heads:
index-only: in memory, not a performance issue index-and-data (eg tags): mostly solved already by the tag cache[1] over-the-wire: probably addressable in the discovery protocol
IIRC, with the current favorite among the new discovery protocol ideas, wire traffic is not significantly affected by the number of heads. -parren
_______________________________________________ Mercurial-devel mailing list Merc...@selenic.com http://selenic.com/mailman/listinfo/mercurial-devel