63
60
content plus the size of the compressed text. Related to bug #109114.
64
61
(John Arbash Meinel)
66
* Repositories using CHK pages (which includes the new 2a format) will no
67
longer error during commit or push operations when an autopack operation
68
is triggered. (Robert Collins, #365615)
70
63
* Stacking will no longer accept requests to stack on the same
71
64
branch/repository. Existing branches that incorrectly reference the same
72
65
repository in a stacking configuration will now raise
80
73
* Progress bars no longer show the network transport scheme or direction.
86
* ``chk_map.iter_interesting_nodes`` now properly uses the *intersection*
87
of referenced nodes rather than the *union* to determine what
88
uninteresting pages we still need to look at. Prior to this,
89
incrementally pushing to stacked branch would push the minimal data, but
90
fetching everything would request extra texts. There are some unhandled
91
cases wrt trees of different depths, but this fixes the common cases.
92
(Robert Collins, John Arbash Meinel, #390563)
94
80
* Command lookup has had hooks added. ``bzrlib.Command.hooks`` has
95
81
three new hook points: ``get_command``, ``get_missing_command`` and
96
82
``list_commands``, which allow just-in-time command name provision
104
90
for files with long ancestry and 'cherrypicked' changes.)
105
91
(John Arbash Meinel, Vincent Ladeuil)
107
* ``GroupCompress`` repositories now take advantage of the pack hints
108
parameter to permit cross-format fetching to incrementally pack the
109
converted data. (Robert Collins)
111
93
* OS file locks are now taken out using ``CreateFile`` rather than
112
94
``LockFileEx`` on Windows. The locking remains exclusive with
113
95
``LockFileEx`` but now it also works on older versions of Windows (such
118
100
properly fetch the minimum number of texts for non-smart fetching.
119
101
(John Arbash Meinel)
121
* ``Repository.commit_write_group`` now returns opaque data about what
122
was committed, for passing to the ``Repository.pack``. Repositories
123
without atomic commits will still return None. (Robert Collins)
125
* ``Repository.pack`` now takes an optional ``hint`` parameter
126
which will support doing partial packs for repositories that can do
127
that. (Robert Collins)
129
* RepositoryFormat has a new attribute 'pack_compresses' which is True
130
when doing a pack operation changes the compression of content in the
131
repository. (Robert Collins)
133
* ``StreamSink`` and ``InterDifferingSerialiser`` will call
134
``Repository.pack`` with the hint returned by
135
``Repository.commit_write_group`` if the formats were different and the
136
repository can increase compression by doing a pack operation.
137
(Robert Collins, #376748)
139
104
* ``VersionedFiles._add_text`` is a new api that lets us insert text into
140
105
the repository as a single string, rather than a list of lines. This can
177
142
this class and the UI. (Martin Pool)
145
bzr 1.16.1 2009-06-26
146
#####################
148
End user testing of the 2a format revealed two serious bugs. The first,
149
#365615, caused bzr to raise AbsentContentFactory errors when autopacking.
150
This meant that commits or pushes to 2a-format repositories failed
153
The second bug, #390563, caused the smart server to raise AbsentContentFactory
154
when streaming 2a stacked 2a-format branches. This particularly affected
155
branches stored on Launchpad in the 2a format.
157
Both of these bugs cause command failures only, neither of them cause data
158
corruption or data loss. And, of course, both of these bugs are now fixed.
163
* We now properly request a more minimal set of file texts when fetching
164
multiple revisions. (Robert Collins, John Arbash Meinel, #390563)
166
* Repositories using CHK pages (which includes the new 2a format) will no
167
longer error during commit or push operations when an autopack operation
168
is triggered. (Robert Collins, #365615)
170
* ``chk_map.iter_interesting_nodes`` now properly uses the *intersection*
171
of referenced nodes rather than the *union* to determine what
172
uninteresting pages we still need to look at. Prior to this,
173
incrementally pushing to stacked branch would push the minimal data, but
174
fetching everything would request extra texts. There are some unhandled
175
cases wrt trees of different depths, but this fixes the common cases.
176
(Robert Collins, John Arbash Meinel, #390563)
178
* ``GroupCompress`` repositories now take advantage of the pack hints
179
parameter to permit cross-format fetching to incrementally pack the
180
converted data. (Robert Collins)
182
* ``Repository.commit_write_group`` now returns opaque data about what
183
was committed, for passing to the ``Repository.pack``. Repositories
184
without atomic commits will still return None. (Robert Collins)
186
* ``Repository.pack`` now takes an optional ``hint`` parameter
187
which will support doing partial packs for repositories that can do
188
that. (Robert Collins)
190
* RepositoryFormat has a new attribute 'pack_compresses' which is True
191
when doing a pack operation changes the compression of content in the
192
repository. (Robert Collins)
194
* ``StreamSink`` and ``InterDifferingSerialiser`` will call
195
``Repository.pack`` with the hint returned by
196
``Repository.commit_write_group`` if the formats were different and the
197
repository can increase compression by doing a pack operation.
198
(Robert Collins, #376748)
180
201
bzr 1.16 "It's yesterday in California" 2009-06-18
181
202
##################################################
182
203
:Codename: yesterday-in-california