README_REPOSITORY
author wenzelm
Thu Sep 13 16:10:20 2012 +0200 (2012-09-13)
changeset 49348 01d2d01bf9d1
parent 48986 037d32448e29
child 49443 75633efcc70d
permissions -rw-r--r--
instructions for quick start in 20min;
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@49348
     7
1. Install Mercurial http://www.selenic.com/mercurial
wenzelm@49348
     8
wenzelm@49348
     9
2. Create file $HOME/.isabelle/etc/settings and insert the following
wenzelm@49348
    10
   line near its beginning:
wenzelm@49348
    11
wenzelm@49348
    12
    init_components "$HOME/.isabelle/contrib" "$ISABELLE_HOME/Admin/components/main"
wenzelm@49348
    13
wenzelm@49348
    14
3. Execute shell commands as follows:
wenzelm@49348
    15
wenzelm@49348
    16
    hg clone http://isabelle.in.tum.de/repos/isabelle
wenzelm@49348
    17
wenzelm@49348
    18
    ./isabelle/bin/isabelle components -a
wenzelm@49348
    19
wenzelm@49348
    20
    ./isabelle/bin/isabelle build -b HOL
wenzelm@49348
    21
wenzelm@49348
    22
    ./isabelle/bin/isabelle jedit
wenzelm@49348
    23
wenzelm@49348
    24
wenzelm@40601
    25
Introduction
wenzelm@40601
    26
------------
wenzelm@28907
    27
wenzelm@40601
    28
Mercurial http://www.selenic.com/mercurial belongs to the current
wenzelm@40601
    29
generation of source code management systems that follow the so-called
wenzelm@40601
    30
paradigm of "distributed version control".  This is a terrific name
wenzelm@40601
    31
for plain revision control without the legacy of CVS or SVN.  See also
wenzelm@40601
    32
http://hginit.com/ for an introduction to the main ideas.  The
wenzelm@40601
    33
Mercurial book http://hgbook.red-bean.com/ explains many more details.
wenzelm@40601
    34
wenzelm@40601
    35
Mercurial offers great flexibility in organizing the flow of changes,
wenzelm@40601
    36
both between individual developers and designated pull/push areas that
wenzelm@40601
    37
are shared with others.  This additional power demands some additional
wenzelm@40601
    38
responsibility to maintain a certain development process that fits to
wenzelm@40601
    39
a particular project.
wenzelm@28907
    40
wenzelm@40601
    41
Regular Mercurial operations are strictly monotonic, where changeset
wenzelm@40601
    42
transactions are only added, but never deleted.  There are special
wenzelm@40601
    43
tools to manipulate repositories via non-monotonic actions (such as
wenzelm@40601
    44
"rollback" or "strip"), but recovering from gross mistakes that have
wenzelm@40601
    45
escaped into the public can be hard and embarrassing.  It is much
wenzelm@40601
    46
easier to inspect and amend changesets routinely before pushing.
wenzelm@28907
    47
wenzelm@40601
    48
The effect of the critical "pull" / "push" operations can be tested in
wenzelm@40601
    49
a dry run via "incoming" / "outgoing".  The "fetch" extension includes
wenzelm@40601
    50
useful sanity checks beyond raw "pull" / "update" / "merge".  Most
wenzelm@40601
    51
other operations are local to the user's repository clone.  This gives
wenzelm@40601
    52
some freedom for experimentation without affecting anybody else.
wenzelm@40601
    53
wenzelm@40601
    54
Mercurial provides nice web presentation of incoming changes with a
wenzelm@40601
    55
digest of log entries; this also includes RSS/Atom news feeds.  There
wenzelm@47449
    56
are add-on history browsers such as "hg view" and TortoiseHg.  Unlike
wenzelm@47449
    57
the default web view, some of these tools help to inspect the semantic
wenzelm@47449
    58
content of non-trivial merge nodes.
wenzelm@28907
    59
wenzelm@28907
    60
wenzelm@28907
    61
Initial configuration
wenzelm@29481
    62
---------------------
wenzelm@28907
    63
wenzelm@28913
    64
The official Isabelle repository can be cloned like this:
wenzelm@28907
    65
wenzelm@28907
    66
  hg clone http://isabelle.in.tum.de/repos/isabelle
wenzelm@28907
    67
wenzelm@28907
    68
This will create a local directory "isabelle", unless an alternative
wenzelm@28907
    69
name is specified.  The full repository meta-data and history of
wenzelm@28907
    70
changes is in isabelle/.hg; local configuration for this clone can be
wenzelm@28907
    71
added to isabelle/.hg/hgrc, but note that hgrc files are never copied
wenzelm@40601
    72
by another clone operation.
wenzelm@28907
    73
wenzelm@28917
    74
wenzelm@28913
    75
There is also $HOME/.hgrc for per-user Mercurial configuration.  The
wenzelm@40601
    76
initial configuration requires at least an entry to identify yourself
wenzelm@40601
    77
as follows:
wenzelm@28907
    78
wenzelm@28907
    79
  [ui]
wenzelm@40601
    80
  username = XXX
wenzelm@28907
    81
wenzelm@40601
    82
Isabelle contributors are free to choose either a short "login name"
wenzelm@40601
    83
(for accounts at TU Munich) or a "full name" -- with or without mail
wenzelm@40601
    84
address.  It is important to stick to this choice once and for all.
wenzelm@40601
    85
The machine default that Mercurial produces for unspecified
wenzelm@40601
    86
[ui]username is not appropriate.
wenzelm@28907
    87
wenzelm@40601
    88
Such configuration can be given in $HOME/.hgrc (for each home
wenzelm@40601
    89
directory on each machine) or in .hg/hgrc (for each repository clone).
wenzelm@28907
    90
wenzelm@28907
    91
wenzelm@40601
    92
Here are some further examples for hgrc.  This is how to provide
wenzelm@40601
    93
default options for common commands:
wenzelm@28907
    94
wenzelm@28907
    95
  [defaults]
wenzelm@28907
    96
  log = -l 10
wenzelm@28907
    97
wenzelm@40601
    98
This is how to configure some extension, which is called "graphlog"
wenzelm@40601
    99
and provides the "glog" command:
wenzelm@28907
   100
wenzelm@28907
   101
  [extensions]
wenzelm@28907
   102
  hgext.graphlog =
wenzelm@28907
   103
wenzelm@28907
   104
wenzelm@28907
   105
Shared pull/push access
wenzelm@29481
   106
-----------------------
wenzelm@28907
   107
wenzelm@28907
   108
The entry point http://isabelle.in.tum.de/repos/isabelle is world
wenzelm@28907
   109
readable, both via plain web browsing and the hg client as described
wenzelm@40601
   110
above.  Anybody can produce a clone, change it locally, and then use
wenzelm@40601
   111
regular mechanisms of Mercurial to report changes upstream, say via
wenzelm@40601
   112
mail to someone with write access to that file space.  It is also
wenzelm@40601
   113
quite easy to publish changed clones again on the web, using the
wenzelm@40601
   114
ad-hoc command "hg serve -v", or the hgweb.cgi or hgwebdir.cgi scripts
wenzelm@40601
   115
that are included in the Mercurial distribution, and send a "pull
wenzelm@40601
   116
request" to someone else.  There are also public hosting services for
wenzelm@40601
   117
Mercurial repositories.
wenzelm@28907
   118
wenzelm@28913
   119
The downstream/upstream mode of operation is quite common in the
wenzelm@28907
   120
distributed version control community, and works well for occasional
wenzelm@28907
   121
changes produced by anybody out there.  Of course, upstream
wenzelm@28907
   122
maintainers need to review and moderate changes being proposed, before
wenzelm@40601
   123
pushing anything onto the official Isabelle repository at TUM.  Direct
wenzelm@40601
   124
pushes are also reviewed routinely in a post-hoc fashion; everybody
wenzelm@40601
   125
hooked on the main repository is called to keep an eye on incoming
wenzelm@40601
   126
changes.  In any case, changesets need to be understandable, at the
wenzelm@40601
   127
time of writing and many years later.
wenzelm@28907
   128
wenzelm@40601
   129
Push access to the Isabelle repository requires an account at TUM,
wenzelm@47449
   130
with properly configured ssh to the local machines (e.g. macbroy20 ..
wenzelm@47449
   131
macbroy29).  You also need to be a member of the "isabelle" Unix
wenzelm@40601
   132
group.
wenzelm@28907
   133
wenzelm@28913
   134
Sharing a locally modified clone then works as follows, using your
wenzelm@28913
   135
user name instead of "wenzelm":
wenzelm@28913
   136
wenzelm@47449
   137
  hg out ssh://wenzelm@macbroy20//home/isabelle-repository/repos/isabelle
wenzelm@28907
   138
wenzelm@28913
   139
In fact, the "out" or "outgoing" command performs only a dry run: it
wenzelm@28913
   140
displays the changesets that would get published.  An actual "push",
wenzelm@28913
   141
with a lasting effect on the Isabelle repository, works like this:
wenzelm@28907
   142
wenzelm@47449
   143
  hg push ssh://wenzelm@macbroy20//home/isabelle-repository/repos/isabelle
wenzelm@28913
   144
wenzelm@28907
   145
wenzelm@40601
   146
Default paths for push and pull can be configured in
wenzelm@40601
   147
isabelle/.hg/hgrc, for example:
wenzelm@28907
   148
wenzelm@28907
   149
  [paths]
wenzelm@47449
   150
  default = ssh://wenzelm@macbroy20//home/isabelle-repository/repos/isabelle
wenzelm@28907
   151
wenzelm@28913
   152
Now "hg pull" or "hg push" will use that shared file space, unless a
wenzelm@28913
   153
different URL is specified explicitly.
wenzelm@28913
   154
wenzelm@28913
   155
When cloning a repository, the default path is set to the initial
wenzelm@28913
   156
source URL.  So we could have cloned via that ssh URL in the first
wenzelm@28913
   157
place, to get exactly to the same point:
wenzelm@28913
   158
wenzelm@47449
   159
  hg clone ssh://wenzelm@macbroy20//home/isabelle-repository/repos/isabelle
wenzelm@28907
   160
wenzelm@28907
   161
wenzelm@40601
   162
Simple merges
wenzelm@40601
   163
-------------
wenzelm@30182
   164
wenzelm@30182
   165
The main idea of Mercurial is to let individual users produce
wenzelm@30182
   166
independent branches of development first, but merge with others
wenzelm@30182
   167
frequently.  The basic hg merge operation is more general than
wenzelm@30182
   168
required for the mode of operation with a shared pull/push area.  The
wenzelm@40601
   169
"fetch" extension accommodates this case nicely, automating trivial
wenzelm@30182
   170
merges and requiring manual intervention for actual conflicts only.
wenzelm@30182
   171
wenzelm@40601
   172
The fetch extension can be configured via $HOME/.hgrc like this:
wenzelm@30182
   173
wenzelm@30182
   174
  [extensions]
wenzelm@30182
   175
  hgext.fetch =
wenzelm@30182
   176
wenzelm@30182
   177
  [defaults]
wenzelm@30182
   178
  fetch = -m "merged"
wenzelm@30182
   179
wenzelm@40601
   180
To keep merges semantically trivial and prevent genuine merge
wenzelm@40601
   181
conflicts or lost updates, it is essential to observe to following
wenzelm@40601
   182
general discipline wrt. the public Isabelle push area:
wenzelm@40601
   183
wenzelm@40601
   184
  * Before editing, pull or fetch the latest version.
wenzelm@40601
   185
wenzelm@40601
   186
  * Accumulate private commits for a maximum of 1-3 days.
wenzelm@40601
   187
wenzelm@40601
   188
  * Start publishing again by pull or fetch, which normally produces
wenzelm@40601
   189
    local merges.
wenzelm@40601
   190
wenzelm@40601
   191
  * Test the merged result as usual and push back in real time.
wenzelm@40601
   192
wenzelm@40601
   193
Piling private changes and public merges longer than 0.5-2 hours is
wenzelm@40601
   194
apt to produce some mess when pushing eventually!
wenzelm@30182
   195
wenzelm@30182
   196
wenzelm@28907
   197
Content discipline
wenzelm@29481
   198
------------------
wenzelm@28907
   199
wenzelm@40601
   200
The following principles should be kept in mind when producing
wenzelm@40601
   201
changesets that are meant to be published at some point.
wenzelm@28907
   202
wenzelm@40601
   203
  * The author of changes needs to be properly identified, using
wenzelm@40601
   204
    [ui]username configuration as described above.
wenzelm@28907
   205
wenzelm@28907
   206
    While Mercurial also provides means for signed changesets, we want
wenzelm@28907
   207
    to keep things simple and trust that users specify their identity
wenzelm@40601
   208
    correctly (and uniquely).
wenzelm@28907
   209
wenzelm@28907
   210
  * The history of sources is an integral part of the sources
wenzelm@28907
   211
    themselves.  This means that private experiments and branches
wenzelm@40601
   212
    should not be published by accident.  Named branches are not
wenzelm@40601
   213
    allowed on the public version.  Note that old SVN-style branching
wenzelm@40601
   214
    is replaced by regular repository clones in Mercurial.
wenzelm@28907
   215
wenzelm@40601
   216
    Exchanging local experiments with some other users can be done
wenzelm@40601
   217
    directly on peer-to-peer basis, without affecting the central
wenzelm@40601
   218
    pull/push area.
wenzelm@28907
   219
wenzelm@28907
   220
  * Log messages are an integral part of the history of sources.
wenzelm@40601
   221
    Other people will have to inspect them routinely, to understand
wenzelm@40601
   222
    why things have been done in a certain way at some point.
wenzelm@40601
   223
wenzelm@40601
   224
    Authors of log entries should be aware that published changes are
wenzelm@40601
   225
    actually read.  The main point is not to announce novelties, but
wenzelm@40601
   226
    to describe faithfully what has been done, and provide some clues
wenzelm@40601
   227
    about the motivation behind it.  The main recipient is someone who
wenzelm@40601
   228
    needs to understand the change in the distant future to isolate
wenzelm@40601
   229
    problems.  Sometimes it is helpful to reference past changes via
wenzelm@40601
   230
    changeset ids in the log entry.
wenzelm@28907
   231
wenzelm@40601
   232
  * The standard changelog entry format of the Isabelle repository
wenzelm@40601
   233
    admits several (vaguely related) items to be rolled into one
wenzelm@40601
   234
    changeset.  Each item is then described on a separate line that
wenzelm@40601
   235
    may become longer as screen line and is terminated by punctuation.
wenzelm@40601
   236
    These lines are roughly ordered by importance.
wenzelm@28907
   237
wenzelm@40601
   238
    This format conforms to established Isabelle tradition.  In
wenzelm@40601
   239
    contrast, the default of Mercurial prefers a single headline
wenzelm@40601
   240
    followed by a long body text.  The reason for that is a somewhat
wenzelm@40601
   241
    different development model of typical "distributed" projects,
wenzelm@40601
   242
    where external changes pass through a hierarchy of reviewers and
wenzelm@40601
   243
    maintainers, until they end up in some authoritative repository.
wenzelm@40601
   244
    Isabelle changesets can be more spontaneous, growing from the
wenzelm@40601
   245
    bottom-up.
wenzelm@40601
   246
wenzelm@40601
   247
    The web style of http://isabelle.in.tum.de/repos/isabelle/
wenzelm@40601
   248
    accommodates the Isabelle changelog format.  Note that multiple
wenzelm@40601
   249
    lines will sometimes display as a single paragraph in HTML, so
wenzelm@40601
   250
    some terminating punctuation is required.  Do not squeeze multiple
wenzelm@40601
   251
    items on the same line in the original text!
wenzelm@28907
   252
wenzelm@28907
   253
wenzelm@32361
   254
Building a repository version of Isabelle
wenzelm@32361
   255
-----------------------------------------
wenzelm@28908
   256
wenzelm@48986
   257
This first requires to resolve add-on components first, including the
wenzelm@48986
   258
ML system.  Some extra configuration is required to approximate some
wenzelm@48986
   259
of the system integration of official Isabelle releases from a
wenzelm@48986
   260
bare-bones repository snapshot.  The special directory Admin/ -- which
wenzelm@48986
   261
is absent in official releases -- might provide some further clues.
wenzelm@48497
   262
wenzelm@48844
   263
Here is a reasonably easy way to include important Isabelle components
wenzelm@48844
   264
on the spot:
wenzelm@48844
   265
wenzelm@48854
   266
  (1) The bash script ISABELLE_HOME_USER/etc/settings is augmented by
wenzelm@48844
   267
  some shell function invocations like this:
wenzelm@28908
   268
wenzelm@48844
   269
      init_components "$HOME/.isabelle/contrib" "$ISABELLE_HOME/Admin/components/main"
wenzelm@48844
   270
      init_components "$HOME/.isabelle/contrib" "$ISABELLE_HOME/Admin/components/optional"
wenzelm@48844
   271
wenzelm@48844
   272
  This uses some central place "$HOME/.isabelle/contrib" to keep
wenzelm@48844
   273
  component directories that are shared by all Isabelle versions.
wenzelm@48844
   274
wenzelm@48844
   275
  (2) Missing components are resolved on the command line like this:
wenzelm@48497
   276
wenzelm@48844
   277
      isabelle components -a
wenzelm@48844
   278
wenzelm@48844
   279
  This will saturate the "$HOME/.isabelle/contrib" directory structure
wenzelm@48855
   280
  according to $ISABELLE_COMPONENT_REPOSITORY.
wenzelm@48497
   281
wenzelm@48844
   282
Since the given component catalogs in $ISABELLE_HOME/Admin/components
wenzelm@48844
   283
are subject to the Mercurial history, it is possible to bisect over a
wenzelm@48844
   284
range of Isabelle versions while references to the contributing
wenzelm@48844
   285
components change accordingly.
wenzelm@48986
   286
wenzelm@48986
   287
The Isabelle build process is managed as follows:
wenzelm@48986
   288
wenzelm@48986
   289
  * regular "isabelle build" to build session images, e.g. HOL;
wenzelm@48986
   290
wenzelm@48986
   291
  * administrative "isabelle build_doc" to populate the doc/
wenzelm@48986
   292
    directory, such that "isabelle doc" will find the results.