Call for testing: generaldelta

[Haszlakiewicz, Eric]

> -----Original Message-----
> From: mercurial-bounces at selenic.com [mailto:mercurial-
> 
> Mercurial 1.9 includes an experimental feature called 'generaldelta'
> 
> == Evaluating compression: ==
> 
> Do two clones:

Here's my results using a project with enough branching to make the hgweb graph fail to render, and the somewhat unusual feature of having more than one root.

> $ hg clone -U --pull proj proj-normal
> $ hg clone -U --pull --config format.generaldelta=1 proj proj-gdelta
requesting all changes
adding changesets
adding manifests
adding file changes
added 1643 changesets with 6512 changes to 1909 files> 

> Then compare their sizes:
> 
> (Unix) $ du -sh proj-normal proj-gdelta
11M     proj-normal
11M     proj-gdelta

> And compare their manifest sizes:
> $ ls -l proj-normal/.hg/store/00manifest.*
-rwx------+ 1 ehaszla Domain Users 725959 Jul 27 12:20 proj-normal/.hg/store/00manifest.d
-rwx------+ 1 ehaszla Domain Users 105088 Jul 27 12:20 proj-normal/.hg/store/00manifest.i
> $ ls -l proj-gdelta/.hg/store/00manifest.*
-rwx------+ 1 ehaszla Domain Users 548705 Jul 27 12:21 proj-gdelta/.hg/store/00manifest.d
-rwx------+ 1 ehaszla Domain Users 105088 Jul 27 12:21 proj-gdelta/.hg/store/00manifest.i> 

> This data may also be valuable:
> 
> $ hg debugrevlog -m
format : 1
flags  : generaldelta

revisions     :   1642
    merges    :    197 (12.00%)
    normal    :   1445 (88.00%)
revisions     :   1642
    full      :      9 ( 0.55%)
    deltas    :   1633 (99.45%)
revision size : 548705
    full      :  98529 (17.96%)
    deltas    : 450176 (82.04%)

avg chain length  : 231
compression ratio : 354

uncompressed data size (min/max/avg) : 730 / 136104 / 118489
full revision size (min/max/avg)     : 318 / 50728 / 10947
delta size (min/max/avg)             : 0 / 43210 / 275

deltas against prev  : 1332 (81.57%)
    where prev = p1  : 1327     (99.62%)
    where prev = p2  :    1     ( 0.08%)
    other            :    4     ( 0.30%)
deltas against p1    :  301 (18.43%)
deltas against p2    :    0 ( 0.00%)
deltas against other :    0 ( 0.00%)

I didn't do any of the other tests.  (sorry, ENOTIME)

eric