README_REPOSITORY
author wenzelm
Wed Sep 19 13:19:45 2012 +0200 (2012-09-19)
changeset 49443 75633efcc70d
parent 49348 01d2d01bf9d1
child 50281 cbba16084784
permissions -rw-r--r--
tuned;
wenzelm@28907
     1
Important notes on Mercurial repository access for Isabelle
wenzelm@28907
     2
===========================================================
wenzelm@28907
     3
wenzelm@49348
     4
Quick start in 20min
wenzelm@49348
     5
--------------------
wenzelm@49348
     6
wenzelm@49443
     7
1. Ensure that "hg" (Mercurial) is installed; see also
wenzelm@49443
     8
   http://www.selenic.com/mercurial
wenzelm@49348
     9
wenzelm@49348
    10
2. Create file $HOME/.isabelle/etc/settings and insert the following
wenzelm@49348
    11
   line near its beginning:
wenzelm@49348
    12
wenzelm@49348
    13
    init_components "$HOME/.isabelle/contrib" "$ISABELLE_HOME/Admin/components/main"
wenzelm@49348
    14
wenzelm@49348
    15
3. Execute shell commands as follows:
wenzelm@49348
    16
wenzelm@49348
    17
    hg clone http://isabelle.in.tum.de/repos/isabelle
wenzelm@49348
    18
wenzelm@49348
    19
    ./isabelle/bin/isabelle components -a
wenzelm@49348
    20
wenzelm@49348
    21
    ./isabelle/bin/isabelle build -b HOL
wenzelm@49348
    22
wenzelm@49348
    23
    ./isabelle/bin/isabelle jedit
wenzelm@49348
    24
wenzelm@49348
    25
wenzelm@40601
    26
Introduction
wenzelm@40601
    27
------------
wenzelm@28907
    28
wenzelm@40601
    29
Mercurial http://www.selenic.com/mercurial belongs to the current
wenzelm@40601
    30
generation of source code management systems that follow the so-called
wenzelm@40601
    31
paradigm of "distributed version control".  This is a terrific name
wenzelm@40601
    32
for plain revision control without the legacy of CVS or SVN.  See also
wenzelm@40601
    33
http://hginit.com/ for an introduction to the main ideas.  The
wenzelm@40601
    34
Mercurial book http://hgbook.red-bean.com/ explains many more details.
wenzelm@40601
    35
wenzelm@40601
    36
Mercurial offers great flexibility in organizing the flow of changes,
wenzelm@40601
    37
both between individual developers and designated pull/push areas that
wenzelm@40601
    38
are shared with others.  This additional power demands some additional
wenzelm@40601
    39
responsibility to maintain a certain development process that fits to
wenzelm@40601
    40
a particular project.
wenzelm@28907
    41
wenzelm@40601
    42
Regular Mercurial operations are strictly monotonic, where changeset
wenzelm@40601
    43
transactions are only added, but never deleted.  There are special
wenzelm@40601
    44
tools to manipulate repositories via non-monotonic actions (such as
wenzelm@40601
    45
"rollback" or "strip"), but recovering from gross mistakes that have
wenzelm@40601
    46
escaped into the public can be hard and embarrassing.  It is much
wenzelm@40601
    47
easier to inspect and amend changesets routinely before pushing.
wenzelm@28907
    48
wenzelm@40601
    49
The effect of the critical "pull" / "push" operations can be tested in
wenzelm@40601
    50
a dry run via "incoming" / "outgoing".  The "fetch" extension includes
wenzelm@40601
    51
useful sanity checks beyond raw "pull" / "update" / "merge".  Most
wenzelm@40601
    52
other operations are local to the user's repository clone.  This gives
wenzelm@40601
    53
some freedom for experimentation without affecting anybody else.
wenzelm@40601
    54
wenzelm@40601
    55
Mercurial provides nice web presentation of incoming changes with a
wenzelm@40601
    56
digest of log entries; this also includes RSS/Atom news feeds.  There
wenzelm@47449
    57
are add-on history browsers such as "hg view" and TortoiseHg.  Unlike
wenzelm@47449
    58
the default web view, some of these tools help to inspect the semantic
wenzelm@47449
    59
content of non-trivial merge nodes.
wenzelm@28907
    60
wenzelm@28907
    61
wenzelm@28907
    62
Initial configuration
wenzelm@29481
    63
---------------------
wenzelm@28907
    64
wenzelm@28913
    65
The official Isabelle repository can be cloned like this:
wenzelm@28907
    66
wenzelm@28907
    67
  hg clone http://isabelle.in.tum.de/repos/isabelle
wenzelm@28907
    68
wenzelm@28907
    69
This will create a local directory "isabelle", unless an alternative
wenzelm@28907
    70
name is specified.  The full repository meta-data and history of
wenzelm@28907
    71
changes is in isabelle/.hg; local configuration for this clone can be
wenzelm@28907
    72
added to isabelle/.hg/hgrc, but note that hgrc files are never copied
wenzelm@40601
    73
by another clone operation.
wenzelm@28907
    74
wenzelm@28917
    75
wenzelm@28913
    76
There is also $HOME/.hgrc for per-user Mercurial configuration.  The
wenzelm@40601
    77
initial configuration requires at least an entry to identify yourself
wenzelm@40601
    78
as follows:
wenzelm@28907
    79
wenzelm@28907
    80
  [ui]
wenzelm@40601
    81
  username = XXX
wenzelm@28907
    82
wenzelm@40601
    83
Isabelle contributors are free to choose either a short "login name"
wenzelm@40601
    84
(for accounts at TU Munich) or a "full name" -- with or without mail
wenzelm@40601
    85
address.  It is important to stick to this choice once and for all.
wenzelm@40601
    86
The machine default that Mercurial produces for unspecified
wenzelm@40601
    87
[ui]username is not appropriate.
wenzelm@28907
    88
wenzelm@40601
    89
Such configuration can be given in $HOME/.hgrc (for each home
wenzelm@40601
    90
directory on each machine) or in .hg/hgrc (for each repository clone).
wenzelm@28907
    91
wenzelm@28907
    92
wenzelm@40601
    93
Here are some further examples for hgrc.  This is how to provide
wenzelm@40601
    94
default options for common commands:
wenzelm@28907
    95
wenzelm@28907
    96
  [defaults]
wenzelm@28907
    97
  log = -l 10
wenzelm@28907
    98
wenzelm@40601
    99
This is how to configure some extension, which is called "graphlog"
wenzelm@40601
   100
and provides the "glog" command:
wenzelm@28907
   101
wenzelm@28907
   102
  [extensions]
wenzelm@28907
   103
  hgext.graphlog =
wenzelm@28907
   104
wenzelm@28907
   105
wenzelm@28907
   106
Shared pull/push access
wenzelm@29481
   107
-----------------------
wenzelm@28907
   108
wenzelm@28907
   109
The entry point http://isabelle.in.tum.de/repos/isabelle is world
wenzelm@28907
   110
readable, both via plain web browsing and the hg client as described
wenzelm@40601
   111
above.  Anybody can produce a clone, change it locally, and then use
wenzelm@40601
   112
regular mechanisms of Mercurial to report changes upstream, say via
wenzelm@40601
   113
mail to someone with write access to that file space.  It is also
wenzelm@40601
   114
quite easy to publish changed clones again on the web, using the
wenzelm@40601
   115
ad-hoc command "hg serve -v", or the hgweb.cgi or hgwebdir.cgi scripts
wenzelm@40601
   116
that are included in the Mercurial distribution, and send a "pull
wenzelm@40601
   117
request" to someone else.  There are also public hosting services for
wenzelm@40601
   118
Mercurial repositories.
wenzelm@28907
   119
wenzelm@28913
   120
The downstream/upstream mode of operation is quite common in the
wenzelm@28907
   121
distributed version control community, and works well for occasional
wenzelm@28907
   122
changes produced by anybody out there.  Of course, upstream
wenzelm@28907
   123
maintainers need to review and moderate changes being proposed, before
wenzelm@40601
   124
pushing anything onto the official Isabelle repository at TUM.  Direct
wenzelm@40601
   125
pushes are also reviewed routinely in a post-hoc fashion; everybody
wenzelm@40601
   126
hooked on the main repository is called to keep an eye on incoming
wenzelm@40601
   127
changes.  In any case, changesets need to be understandable, at the
wenzelm@40601
   128
time of writing and many years later.
wenzelm@28907
   129
wenzelm@40601
   130
Push access to the Isabelle repository requires an account at TUM,
wenzelm@47449
   131
with properly configured ssh to the local machines (e.g. macbroy20 ..
wenzelm@47449
   132
macbroy29).  You also need to be a member of the "isabelle" Unix
wenzelm@40601
   133
group.
wenzelm@28907
   134
wenzelm@28913
   135
Sharing a locally modified clone then works as follows, using your
wenzelm@28913
   136
user name instead of "wenzelm":
wenzelm@28913
   137
wenzelm@47449
   138
  hg out ssh://wenzelm@macbroy20//home/isabelle-repository/repos/isabelle
wenzelm@28907
   139
wenzelm@28913
   140
In fact, the "out" or "outgoing" command performs only a dry run: it
wenzelm@28913
   141
displays the changesets that would get published.  An actual "push",
wenzelm@28913
   142
with a lasting effect on the Isabelle repository, works like this:
wenzelm@28907
   143
wenzelm@47449
   144
  hg push ssh://wenzelm@macbroy20//home/isabelle-repository/repos/isabelle
wenzelm@28913
   145
wenzelm@28907
   146
wenzelm@40601
   147
Default paths for push and pull can be configured in
wenzelm@40601
   148
isabelle/.hg/hgrc, for example:
wenzelm@28907
   149
wenzelm@28907
   150
  [paths]
wenzelm@47449
   151
  default = ssh://wenzelm@macbroy20//home/isabelle-repository/repos/isabelle
wenzelm@28907
   152
wenzelm@28913
   153
Now "hg pull" or "hg push" will use that shared file space, unless a
wenzelm@28913
   154
different URL is specified explicitly.
wenzelm@28913
   155
wenzelm@28913
   156
When cloning a repository, the default path is set to the initial
wenzelm@28913
   157
source URL.  So we could have cloned via that ssh URL in the first
wenzelm@28913
   158
place, to get exactly to the same point:
wenzelm@28913
   159
wenzelm@47449
   160
  hg clone ssh://wenzelm@macbroy20//home/isabelle-repository/repos/isabelle
wenzelm@28907
   161
wenzelm@28907
   162
wenzelm@40601
   163
Simple merges
wenzelm@40601
   164
-------------
wenzelm@30182
   165
wenzelm@30182
   166
The main idea of Mercurial is to let individual users produce
wenzelm@30182
   167
independent branches of development first, but merge with others
wenzelm@30182
   168
frequently.  The basic hg merge operation is more general than
wenzelm@30182
   169
required for the mode of operation with a shared pull/push area.  The
wenzelm@40601
   170
"fetch" extension accommodates this case nicely, automating trivial
wenzelm@30182
   171
merges and requiring manual intervention for actual conflicts only.
wenzelm@30182
   172
wenzelm@40601
   173
The fetch extension can be configured via $HOME/.hgrc like this:
wenzelm@30182
   174
wenzelm@30182
   175
  [extensions]
wenzelm@30182
   176
  hgext.fetch =
wenzelm@30182
   177
wenzelm@30182
   178
  [defaults]
wenzelm@30182
   179
  fetch = -m "merged"
wenzelm@30182
   180
wenzelm@40601
   181
To keep merges semantically trivial and prevent genuine merge
wenzelm@40601
   182
conflicts or lost updates, it is essential to observe to following
wenzelm@40601
   183
general discipline wrt. the public Isabelle push area:
wenzelm@40601
   184
wenzelm@40601
   185
  * Before editing, pull or fetch the latest version.
wenzelm@40601
   186
wenzelm@40601
   187
  * Accumulate private commits for a maximum of 1-3 days.
wenzelm@40601
   188
wenzelm@40601
   189
  * Start publishing again by pull or fetch, which normally produces
wenzelm@40601
   190
    local merges.
wenzelm@40601
   191
wenzelm@40601
   192
  * Test the merged result as usual and push back in real time.
wenzelm@40601
   193
wenzelm@40601
   194
Piling private changes and public merges longer than 0.5-2 hours is
wenzelm@40601
   195
apt to produce some mess when pushing eventually!
wenzelm@30182
   196
wenzelm@30182
   197
wenzelm@28907
   198
Content discipline
wenzelm@29481
   199
------------------
wenzelm@28907
   200
wenzelm@40601
   201
The following principles should be kept in mind when producing
wenzelm@40601
   202
changesets that are meant to be published at some point.
wenzelm@28907
   203
wenzelm@40601
   204
  * The author of changes needs to be properly identified, using
wenzelm@40601
   205
    [ui]username configuration as described above.
wenzelm@28907
   206
wenzelm@28907
   207
    While Mercurial also provides means for signed changesets, we want
wenzelm@28907
   208
    to keep things simple and trust that users specify their identity
wenzelm@40601
   209
    correctly (and uniquely).
wenzelm@28907
   210
wenzelm@28907
   211
  * The history of sources is an integral part of the sources
wenzelm@28907
   212
    themselves.  This means that private experiments and branches
wenzelm@40601
   213
    should not be published by accident.  Named branches are not
wenzelm@40601
   214
    allowed on the public version.  Note that old SVN-style branching
wenzelm@40601
   215
    is replaced by regular repository clones in Mercurial.
wenzelm@28907
   216
wenzelm@40601
   217
    Exchanging local experiments with some other users can be done
wenzelm@40601
   218
    directly on peer-to-peer basis, without affecting the central
wenzelm@40601
   219
    pull/push area.
wenzelm@28907
   220
wenzelm@28907
   221
  * Log messages are an integral part of the history of sources.
wenzelm@40601
   222
    Other people will have to inspect them routinely, to understand
wenzelm@40601
   223
    why things have been done in a certain way at some point.
wenzelm@40601
   224
wenzelm@40601
   225
    Authors of log entries should be aware that published changes are
wenzelm@40601
   226
    actually read.  The main point is not to announce novelties, but
wenzelm@40601
   227
    to describe faithfully what has been done, and provide some clues
wenzelm@40601
   228
    about the motivation behind it.  The main recipient is someone who
wenzelm@40601
   229
    needs to understand the change in the distant future to isolate
wenzelm@40601
   230
    problems.  Sometimes it is helpful to reference past changes via
wenzelm@40601
   231
    changeset ids in the log entry.
wenzelm@28907
   232
wenzelm@40601
   233
  * The standard changelog entry format of the Isabelle repository
wenzelm@40601
   234
    admits several (vaguely related) items to be rolled into one
wenzelm@40601
   235
    changeset.  Each item is then described on a separate line that
wenzelm@40601
   236
    may become longer as screen line and is terminated by punctuation.
wenzelm@40601
   237
    These lines are roughly ordered by importance.
wenzelm@28907
   238
wenzelm@40601
   239
    This format conforms to established Isabelle tradition.  In
wenzelm@40601
   240
    contrast, the default of Mercurial prefers a single headline
wenzelm@40601
   241
    followed by a long body text.  The reason for that is a somewhat
wenzelm@40601
   242
    different development model of typical "distributed" projects,
wenzelm@40601
   243
    where external changes pass through a hierarchy of reviewers and
wenzelm@40601
   244
    maintainers, until they end up in some authoritative repository.
wenzelm@40601
   245
    Isabelle changesets can be more spontaneous, growing from the
wenzelm@40601
   246
    bottom-up.
wenzelm@40601
   247
wenzelm@40601
   248
    The web style of http://isabelle.in.tum.de/repos/isabelle/
wenzelm@40601
   249
    accommodates the Isabelle changelog format.  Note that multiple
wenzelm@40601
   250
    lines will sometimes display as a single paragraph in HTML, so
wenzelm@40601
   251
    some terminating punctuation is required.  Do not squeeze multiple
wenzelm@40601
   252
    items on the same line in the original text!
wenzelm@28907
   253
wenzelm@28907
   254
wenzelm@32361
   255
Building a repository version of Isabelle
wenzelm@32361
   256
-----------------------------------------
wenzelm@28908
   257
wenzelm@48986
   258
This first requires to resolve add-on components first, including the
wenzelm@48986
   259
ML system.  Some extra configuration is required to approximate some
wenzelm@48986
   260
of the system integration of official Isabelle releases from a
wenzelm@48986
   261
bare-bones repository snapshot.  The special directory Admin/ -- which
wenzelm@48986
   262
is absent in official releases -- might provide some further clues.
wenzelm@48497
   263
wenzelm@48844
   264
Here is a reasonably easy way to include important Isabelle components
wenzelm@48844
   265
on the spot:
wenzelm@48844
   266
wenzelm@48854
   267
  (1) The bash script ISABELLE_HOME_USER/etc/settings is augmented by
wenzelm@48844
   268
  some shell function invocations like this:
wenzelm@28908
   269
wenzelm@48844
   270
      init_components "$HOME/.isabelle/contrib" "$ISABELLE_HOME/Admin/components/main"
wenzelm@48844
   271
      init_components "$HOME/.isabelle/contrib" "$ISABELLE_HOME/Admin/components/optional"
wenzelm@48844
   272
wenzelm@48844
   273
  This uses some central place "$HOME/.isabelle/contrib" to keep
wenzelm@48844
   274
  component directories that are shared by all Isabelle versions.
wenzelm@48844
   275
wenzelm@48844
   276
  (2) Missing components are resolved on the command line like this:
wenzelm@48497
   277
wenzelm@48844
   278
      isabelle components -a
wenzelm@48844
   279
wenzelm@48844
   280
  This will saturate the "$HOME/.isabelle/contrib" directory structure
wenzelm@48855
   281
  according to $ISABELLE_COMPONENT_REPOSITORY.
wenzelm@48497
   282
wenzelm@48844
   283
Since the given component catalogs in $ISABELLE_HOME/Admin/components
wenzelm@48844
   284
are subject to the Mercurial history, it is possible to bisect over a
wenzelm@48844
   285
range of Isabelle versions while references to the contributing
wenzelm@48844
   286
components change accordingly.
wenzelm@48986
   287
wenzelm@48986
   288
The Isabelle build process is managed as follows:
wenzelm@48986
   289
wenzelm@48986
   290
  * regular "isabelle build" to build session images, e.g. HOL;
wenzelm@48986
   291
wenzelm@48986
   292
  * administrative "isabelle build_doc" to populate the doc/
wenzelm@48986
   293
    directory, such that "isabelle doc" will find the results.