Differences between revisions 2 and 44 (spanning 42 versions)
Revision 2 as of 2005-08-26 01:23:08
Size: 633
Editor: waste
Comment:
Revision 44 as of 2013-08-31 08:52:17
Size: 2656
Editor: rcl
Comment:
Deletions are marked like this. Additions are marked like this.
Line 1: Line 1:
== Mercurial working practices == = Mercurial Working Practices =
Line 3: Line 3:
This page documents some ways to use Mercurial. Because the software is flexible, there's no "right way", but some methods are more scalable than others. This page documents some ways to use Mercurial. Because the software is flexible, there's no "right way", but some methods are more scalable than others. These are a few tips for creating a scaleable workflow.
Line 5: Line 5:
   * * How to write good ChangeSetComments
   * * RepositoryNaming is important, because you'll probably have lots of them
First, [[Merge|merge]] often! This makes merging easier for everyone and you
find out about conflicts (which are often rooted in incompatible
design decisions) earlier.
Line 8: Line 9:
=== Ways to collaborate === Second, don't hesitate to use multiple trees locally. Mercurial makes
this fast and light-weight. Typical usage is to have an incoming tree,
an outgoing tree, and a separate tree for each area being worked on.
Line 10: Line 13:
| '''Name''' | '''Scalability''' | '''Overhead''' | '''Description''' |
| CvsLikePractice | poor | low | keep things simple, use a few central repositories |
| KernelPractice | good | medium | distributed, semi-hierarchical development |
{{{#!dot
digraph G {
  upstream;
  subgraph {
    incoming; outgoing;
    rank=same
  }
  working;
  upstream -> incoming [label="pull"];
  incoming -> working [label="pull"];
  working -> working [label="update/merge"];
  working -> outgoing [label="push"];
  outgoing -> upstream [label="push or publish"];
}
}}}

The incoming tree is best maintained as a pristine copy of the
upstream [[Repository|repository]]. This works as a cache so that you don't have to
[[Pull|pull]] multiple copies over the network. No need to check files out here
as you won't be changing them.

The outgoing tree contains all the changes you intend for merge into
upstream. Publish this tree with `hg serve` or [[PublishingRepositories|hgweb]] or use
`hg push` to [[Push|push]] it to another publicly available repository.

Then, for each feature you work on, create a new tree. Commit early
and commit often, merge with incoming regularly, and once you're
satisfied with your feature, pull the changes into your outgoing tree.

== Other ways to collaborate ==

|| '''Name''' || '''Scalability''' || '''Overhead''' || '''Description''' ||
|| CvsLikePractice || poor || low || keep things simple, use a few central repositories ||
|| KernelPractice || good || medium || distributed, semi-hierarchical development ||
|| ControlledPractice || good || medium || hierarchical development ||

[[ContributingChanges|Mercurial development practices]] resemble those described in KernelPractice.

== Backup practices ==

To maintain a backup of commits, you can clone a repository and push your changes to it periodically.

== See also ==

 * How to write good [[ChangeSetComments|changeset comments]]
 * The [[RepositoryNaming|naming of repositories]] is important, because you'll probably have lots of them
 * Dealing with contributions from [[MultipleCommitters|multiple committers]]
 * More [[Workflows]] with usecase and requirements

----
CategoryHowTo

Mercurial Working Practices

This page documents some ways to use Mercurial. Because the software is flexible, there's no "right way", but some methods are more scalable than others. These are a few tips for creating a scaleable workflow.

First, merge often! This makes merging easier for everyone and you find out about conflicts (which are often rooted in incompatible design decisions) earlier.

Second, don't hesitate to use multiple trees locally. Mercurial makes this fast and light-weight. Typical usage is to have an incoming tree, an outgoing tree, and a separate tree for each area being worked on.

The incoming tree is best maintained as a pristine copy of the upstream repository. This works as a cache so that you don't have to pull multiple copies over the network. No need to check files out here as you won't be changing them.

The outgoing tree contains all the changes you intend for merge into upstream. Publish this tree with hg serve or hgweb or use hg push to push it to another publicly available repository.

Then, for each feature you work on, create a new tree. Commit early and commit often, merge with incoming regularly, and once you're satisfied with your feature, pull the changes into your outgoing tree.

Other ways to collaborate

Name

Scalability

Overhead

Description

CvsLikePractice

poor

low

keep things simple, use a few central repositories

KernelPractice

good

medium

distributed, semi-hierarchical development

ControlledPractice

good

medium

hierarchical development

Mercurial development practices resemble those described in KernelPractice.

Backup practices

To maintain a backup of commits, you can clone a repository and push your changes to it periodically.

See also


CategoryHowTo

WorkingPractices (last edited 2013-08-31 09:38:15 by rcl)