docs/instability.rst
author Pierre-Yves David <pierre-yves.david@logilab.fr>
Fri, 24 Aug 2012 11:17:18 +0200
changeset 492 7ecd41520dae
parent 365 9f6a77ec66cb
child 519 9825c7da5b54
permissions -rw-r--r--
rename `stabilize` to `evolve` stabilize keep being valide but as Matt still like "Changeset Evolution" as a name evolve make more sense. Stabilize is a bit too tight to unstable too.


-----------------------------------
The instability Principle
-----------------------------------



An intrinsic contradiction
-----------------------------------

XXX starts by talking about getting ride of changeset.

DVCSes bring two new major concepts to the Version Control Scene:

    * History is organized as a robust DAG,
    * History can be rewritten.

However, the two concepts are in contradiction:

To achieve a robust history, three key elements are gathered in *changesets*:

    * Full snapshot of the versioned content,
    * Reference to the previous full snapshot used to build the new one,
    * A description of the change who lead from the old content to the new old.

All three elements are to compute a *unique* hash that identify the changeset
(with various other metadata). This identification is a key part of DVCS design.

This is a very useful property because Changing B parent means
changing B content too. This requires the creation of **another**
changeset, which is semantically good.

::

  Schema base,  A, B and B'

To avoid duplication, the older changeset is usually discarded from accessible
history. I'm calling them *obsolete* changesets.


But rewriting a changeset with children does not change these
children's parent! And because children of the rewritten changeset
still **depend** on the older "dead" version of the changeset with
can not get rid of this dead version.

::

  Schema base,  A and A' and B.

I'm calling these children **unstable** because they are based on a
dead changeset and prevent people to get rid of it.

This instability is an **unavoidable consequence** of the strict dependency of
changeset.  History Rewriting history alway  need to take it in account and
provide a way to rewrite the descendant on the new changeset to avoid
coexistence of the old and new version of a rewritten changeset.


Everybody is working around the issue
------------------------------------------------

I'm not claiming that rewriting history is impossible. People are successfully
doing for years. However they all need to work around *instability*. Several
work around strategy exists.


Rewriting all at once
``````````````````````````

The simplest way to avoid instability is to ensure rewriting
operations always end in a stable situation. This is achieved by
rewriting all affected changesets at the same time.

Rewriting all descendants at the same time when rewriting a changeset.

::

  Schema!

Several Mercurial commands apply it: rebase, collapse, histedit.
Mercurial also refuses to amend changeset with descendant. The git
branch design enforces such approach in git too.


However, DVCS are **Distributed**. This means that you do not control what
happen outside your repository. Once a changeset have been exchanged *outside*,
there is no way to be sure it does not have descendants somewhere else.
Therefore **if you rewrite changeset that exists elsewhere, you can't eradicate
the risk of instability.**

Do not rewrite exchanged changeset
```````````````````````````````````

To work around the issue above, Mercurial introduced phases, which
prevent you from rewriting shared changesets and ensure others can't
pull certain changesets from you. But this is a very frustrating
limitation that prevents you to efficiently sharing, reviewing and
collaborating on mutable changesets.

In the Git world, they use another approach to prevent instability. By
convention only a single developper works on a changeset contained in
a named branch. But once again this is a huge blocker for
collaborating. Moreover clueless people **will** mess up social
convention soon or later.


Loose the DAG robustness
````````````````````````````

The other approach in Mercurial is to keep the mutable part of the
history outside the DVCS constraint. This is the MQ approach of
sticking a quilt queue over Mercurial.

This allow much more flexible workflow but two major feature are lost in the
process:

:Graceful merge: MQ use plain-patch to store changeset content and patch have
                 trouble to apply in changing context. Applying your queue
                 becomes very painful when context changes.

:easy branching: A quilt queue is by definition a linear queue. Increasing risk
                 of conflict

It is possible to collaborate over versioned mq! But you are going to
have a lot of troubles.

Ignore conflicts
```````````````````````````````````
 
Another ignored issue is a conflicting rewrite of the same changeset.
If a changeset is rewritten two times we have two newer versions,
and duplicated history is complicated to merge.

Mercurial work around by

The "One set of mutable changset == One developer" mantra is also a way to work
around conflicting rewriting of changeset. If two different people are able to

The git branch model allow to overwrite changeset version by another
one, but it does not care about divergent version. It is the equivalent
of "common ftp" source management for changesets.

Facing The Danger Once And For All
------------------------------------------------

Above we saw that, the more effort you put to avoid instability, the more option
you deny. And even most restrictive work flow can't guarantee that instability
will never show up!

Obsolete marker can handle the job
```````````````````````````````````

It is time to provide a full featured solution to deal with
instability and to stop working around the issue! This is why I
developing a new feature for mercurial called "Obsolete markers".
Obsolete markers have two key properties:


* Any "old" changeset we want to get ride of is **explicitly** marked
  as "obsolete" by history rewriting operation.

  By explicitly marking the obsolete part of the history, we will be able to
  easily detect instability situation.

* Relations between old and new version of changesets are tracked by obsolete
  markers.

  By Storing a meta-history of changeset evolution we are able to easily resolve
  instability and edition conflict [#]_ .

.. [#] edition conflict is another major obstable to collaboration. See the
       section dedicated to obsolete marker for details.

Improves robustness == improves simplicity
````````````````````````````````````````````````

This proposal should **first** be seen as a safety measure.

It allow to detect instability as soon as possible

::

    $ hg pull
    added 3 changeset
    +2 unstable changeset
    (do you want "hg evolve" ?)
    working directory parent is obsolete!
    $ hg push
    outgoing unstable changesets
    (use "hg evolve" or force the push)

And should not not encourage people to create instability

::

    $ hg up 42
    $ hg commit --amend
    changeset have descendant.
    $ hg commit --amend -f
    +5 unstable changeset

    $ hg rebase -D --rev 40::44
    rebasing already obsolete changeset 42:AAA will conflict with newer version 48:BBB

While allowing powerful feature
````````````````````````````````````````````````


* Help to automatically solve instability.

* "prune" changeset remotely.

* track resulting changeset when submitting patch//pull request.

* Focus on what you do:

  I do not like the "all at once" model of history rewriting. I'm comfortable
  with instability and obsolete marker offer all the tool to safely create and
  handle instability locally.