mirrors/jj
1
0
Fork 1
mirror of https://github.com/martinvonz/jj.git synced 2024-12-01 03:45:55 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 2
jj/CHANGELOG.md

212 lines
8.3 KiB
Markdown
Raw Normal View History

release: add a changelog
2022-03-12 18:48:51 +00:00
# Changelog
All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [Unreleased]
cli: make operation to undo or restore to a positional argument I originally made the operation argument a named argument (`--operation`) to allow for a change ID to be passed as a positional argument, matching e.g. `hg revert -r <rev> <path>`. However, even if we add support for undoing changes only to certain change IDs, it's going to be done much less frequently than full undo/restore. We can therefore make that a named argument if we ever add it.
2022-04-24 19:31:54 +00:00
### Breaking changes
* `jj [op] undo` and `jj op restore` used to take the operation to undo or
restore to as an argument to `-o/--operation`. It is now a positional
argument instead (i.e. `jj undo -o abc123` is now written `jj undo abc123`).
cli: make invalid alias definition an error
2022-05-07 13:57:13 +00:00
* An alias that is not configured as a string list (e.g. `my-status = "status"`
instead of `my-status = ["status"]`) is now an error instead of a warning.
cli: make `log` show only local commits by default The default log output of showing all commits is not very useful when contributing to an existing repo. Let's have it default to showing commits not on any remote branch instead. I think that's the best we can do since we don't have a configurable main branch yet, and we don't even have per-repo configuration.. Closes #250.
2022-05-14 15:55:46 +00:00
* (#250) `jj log` now defaults to showing only commits that are not on any
remote branches (plus their closest commit on the remote branch for context).
Use `jj log -r 'all()'` for the old behavior. Read more about revsets
[here](https://github.com/martinvonz/jj/blob/main/docs/revsets.md).
cli: make `jj new` always check out the new commit Before this change, `jj new` would check out the new commit only if it was created on top of the current commit. I never liked that special-casing, and after thinking more about how the open/closed should work (see discussion #321), I think we want `jj new` to behave similar to how `git/hg checkout` works, so it can effectively replace the current `jj checkout` command for the use case of starting new work on top of an existing commit.
2022-05-21 13:38:56 +00:00
* `jj new` now always checks out the new commit (used to be only if the parent
was `@`).
cli: make `jj branch` take subcommands, not flags As per https://github.com/martinvonz/jj/issues/330.
2022-06-06 03:23:01 +00:00
* (#330) `jj branch` now uses subcommands like `jj branch create` and
`jj branch forget` instead of options like `jj branch --forget`.
cli: add a command for printing the contents of a file in a revision I'm adding this mostly because it's useful for testing. That's also the reason it supports displaying conflicts. I didn't call it `cat` like `hg cat` because I haven't found `hg cat` on multiple files useful.
2022-04-06 18:00:27 +00:00
### New features
cli: add mode for rebasing branch onto destination (#168)
2022-04-13 17:53:50 +00:00
* `jj rebase` now accepts a `--branch/-b <revision>` argument, which can be used
instead of `-r` or `-s` to specify which commits to rebase. It will rebase the
cli: make `jj rebase` default to `-b @` (#168) Closes #168.
2022-04-14 06:53:35 +00:00
whole branch, relative to the destination. The default mode has changed from
`-r @` to `-b @`.
cli: add mode for rebasing branch onto destination (#168)
2022-04-13 17:53:50 +00:00
cli: add a command for printing the contents of a file in a revision I'm adding this mostly because it's useful for testing. That's also the reason it supports displaying conflicts. I didn't call it `cat` like `hg cat` because I haven't found `hg cat` on multiple files useful.
2022-04-06 18:00:27 +00:00
* The new `jj print` command prints the contents of a file in a revision.
cli: add a command for listing git remotes and their URLs As requested by @talpr. I added this is a separate new command `jj git remote list`. One could also imagine showing the listing when there is no sub-command specified to `jj git remote`, but we don't have other commands that behave that way yet. Closes #243
2022-04-25 20:34:13 +00:00
* (#243) The new `jj git remotes list` command lists the configured remotes and
their URLs.
cli: teach `jj squash` to move only changes to specified paths
2022-04-09 17:27:17 +00:00
* `jj move` and `jj squash` now lets you limit the set of changes to move by
specifying paths on the command line (in addition to the `--interactive`
mode). For example, use `jj move --to @-- foo` to move the changes to file
(or directory) `foo` in the working copy to the grandparent commit.
cli: teach `jj move` to move only changes to specified paths I quite often want to move the changes to a particular file from one commit to another. We already support that using `jj move -i`, but that can be annoying to run because we don't have a TUI for it (#48). Let's make it possible to do `jj move --from X --to Y <path>`.
2022-04-09 16:58:33 +00:00
cli: teach `jj split` to include all specified paths in first commit
2022-04-09 18:35:28 +00:00
* `jj split` now lets you specify on the CLI which paths to include in the first
commit. The interactive diff-editing is not started when you do that.
cli: add commands for working with sparse patterns (#52) This adds a `jj sparse` command with options to list and manage the set of paths to include in the working copy. It only supports includes (postive matches) for now. I'm not sure "sparse" is the best name for the feature. Perhaps it would make sense as a subcommand under `jj workspace` - maybe `jj workspace track`? However, there's also `jj untrack` for removing a file from the working copy and leaving it in the working copy. I'm happy to hear suggestions, or we can get back to the naming later.
2022-02-06 23:10:50 +00:00
* Sparse checkouts are now supported. In fact, all working copies are now
"sparse", only to different degrees. Use the `jj sparse` command to manage
cli: `jj branch` can accept any number of branches
2022-05-02 19:58:32 +00:00
the paths included in the sparse checkout.
cli: add commands for working with sparse patterns (#52) This adds a `jj sparse` command with options to list and manage the set of paths to include in the working copy. It only supports includes (postive matches) for now. I'm not sure "sparse" is the best name for the feature. Perhaps it would make sense as a subcommand under `jj workspace` - maybe `jj workspace track`? However, there's also `jj untrack` for removing a file from the working copy and leaving it in the working copy. I'm happy to hear suggestions, or we can get back to the naming later.
2022-02-06 23:10:50 +00:00
cli: allow config in `~/.jjconfig.toml` As I said in 095fb9fef4b1, removing support for `~/.jjconfig` was an experiment. I've heard from a few people (including in #233) that they would prefer to have configs in the home directory. This patch therefore restores that functionality, except I added a `.toml` extension to the file to clarify the expected format to users and editors. After this patch, we still allow configs in `$XDG_CONFIG_HOME` (and the other paths used by `dirs::config_dir()`), but we error out there are config files in both that location and `~/.jjconfig.toml`.
2022-05-08 06:54:51 +00:00
* Configuration is now also read from `~/.jjconfig.toml`.
cli: if `$JJ_CONFIG` points to a directory, read all files in it It's annoying especially for tests to not be able to append to a config file without knowing the contents (as you have to do with TOML). Let's read all files in a directory if `$JJ_CONFIG` points to a directory. Mercurial does that for its `$HGRCPATH` variable.
2022-04-09 23:14:42 +00:00
* The `$JJ_CONFIG` environment variable can now point to a directory. If it
does, all files in the directory will be read, in alphabetical order.
cli: respect `$VISUAL`, overriding `$EDITOR` With this patch, the order is this: `$JJ_EDITOR` environement variable `ui.editor` config `$VISUAL` environement variable `$EDITOR` environement variable `pico` That matches git, except that git falls back to an editor determined at compile time (usually `vi`) instead of using `pico`.
2022-05-10 17:04:10 +00:00
* The `$VISUAL` environment is now respected and overrides `$EDITOR`. The new
`ui.editor` config has higher priority than both of them. There is also a new
`$JJ_EDITOR` environment variable, which has even higher priority than the
config.
cli: allow overriding `$EDITOR` variable by `ui.editor` config
2022-04-10 05:42:27 +00:00
revsets: allow single internal `-` and `+` characters in symbols (#46)
2022-04-27 16:42:18 +00:00
* You can now use `-` and `+` in revset symbols. You used to have to quote
branch names like `my-feature` in nested quotes (outer layer for your shell)
like `jj co '"my-feature"'`. The quoting is no longer needed.
revset: add a `connected()` function This introduces a `connected(x)` function, which is simply the same as `x:x`. It's occasionally useful if `x` is a long expression. It's also useful as a building block for `root(x)` (coming soon).
2022-04-14 04:26:21 +00:00
* The new revset function `connected(x)` is the same as `x:x`.
revset: add a `roots()` function
2022-04-13 20:55:47 +00:00
* The new revset function `roots(x)` finds commits in the set that are not
descendants of other commits in the set.
Remove out-of-date parenthetical in tutorial.md
2022-04-23 04:30:33 +00:00
* ssh-agent is now detected even if `$SSH_AGENT_PID` is not set (as long as
`$SSH_AUTH_SOCK` is set). This should help at least macOS users where
ssh-agent is launched by default and only `$SSH_AUTH_SOCK` is set.
git: when importing refs, abandon commits that were abandoned in git Now that I'm using GitHub PRs instead of pushing directly to the main branch, it's quite annoying to have to abandon the old commits after GitHub rebases them. This patch makes it so we compare the remote's previous heads to the new heads and abandons any commits that were removed on the remote. As usual, that means that descendants get rebased onto the closest remaining commit. This is half of #241. The other half is to detect rewritten branches and rebase on top.
2022-03-27 05:33:08 +00:00
* When importing from a git, any commits that are no longer referenced on the
git side will now be abandoned on the jj side as well. That means that
`jj git fetch` will now abandon unreferenced commits and rebase any local
changes you had on top.
cli: add a mode of pushing with branch taken from change ID This adds `jj git push --change <revision>` which creates a branch with a name based on the revision's change ID, and then pushes that like with `--branch`. That can be useful so you don't have to manually add the branch (and come up with a name for it). The created branch behaves like any other branch, so it's possible to make it point to a commit with a different change ID.
2022-04-29 03:58:41 +00:00
* `jj git push` gained a `--change <revision>` argument. When that's used, it
will create a branch named after the revision's change ID, so you don't have
Add configuration option to override branch name prefix.
2022-05-24 09:52:45 +00:00
to create a branch yourself. By default, the branch name will start with
`push-`, but this can be overridden by the `push.branch-prefix` config
setting.
cli: add a mode of pushing with branch taken from change ID This adds `jj git push --change <revision>` which creates a branch with a name based on the revision's change ID, and then pushes that like with `--branch`. That can be useful so you don't have to manually add the branch (and come up with a name for it). The created branch behaves like any other branch, so it's possible to make it point to a commit with a different change ID.
2022-04-29 03:58:41 +00:00
diff_edit: load command arguments from merge-tools.<name> config Apparently, I need to pass `--merge` option to use kdiff3 as a diff editor. We could add `diff-editor-args` or extend `diff-editor` to a list of command arguments, but we'll eventually add stock merge tools and the configuration would look like: [merge-tools.<name>] program = ... diff-args = [...] edit-args = [...] merge-args = [...]
2022-05-01 01:45:22 +00:00
* Diff editor command arguments can now be specified by config file.
Example:
[merge-tools.kdiff3]
program = "kdiff3"
edit-args = ["--merge", "--cs", "CreateBakFiles=0"]
cli: `jj branch` can accept any number of branches
2022-05-02 19:58:32 +00:00
* `jj branch` can accept any number of branches to update, rather than just one.
cli: allow alias after global args, and recursive aliases Our support for aliases is very naively implemented; it assumes the alias is the first argument in argv. It therefore fails to resolve aliases after global arguments such as `--at-op`. This patch fixes that by modifying the command defintion to have an "external subcommand" in the list of available commands. That makes `clap` give us the remainder of the arguments when it runs into an unknown command. The first in the list will then be an alias or simply an unknown command. Thanks to @epage for the suggestion on in clap-rs/clap#3672. With the new structure, it was easy to handle recursive alias definitions, so I added support for that too. Closes #292.
2022-05-01 15:54:35 +00:00
* Aliases can now call other aliases.
cli: teach `log` flag to show commits in reverse order This adds a `--reversed` flag to `jj log` to show commits with later commits further down. It works both with and without the graph. Since the graph-drawing code is already independent of the relationship between commits, it doesn't need any updating.
2022-05-09 17:38:23 +00:00
* `jj log` now accepts a `--reversed` option, which will show older commits
first.
cli: print errors to stderr
2022-04-07 06:25:01 +00:00
### Fixed bugs
merge: fix modify/delete conflict to not resolve
2022-04-06 19:18:24 +00:00
* When rebasing a conflict where one side modified a file and the other side
deleted it, we no longer automatically resolve it in favor of the modified
content (this was a regression from commit c0ae4b16e8c4).
cli: print errors to stderr
2022-04-07 06:25:01 +00:00
* Errors are now printed to stderr (they used to be printed to stdout).
working_copy: fix crash when updating and only executable bit changed
2022-04-15 06:03:08 +00:00
* Updating the working copy to a commit where a file's executable bit changed
but the contents was the same used to lead to a crash. That has now been
cli: `jj branch` can accept any number of branches
2022-05-02 19:58:32 +00:00
fixed.
working_copy: fix crash when updating and only executable bit changed
2022-04-15 06:03:08 +00:00
trees: when merging trees and one is missing, treat it as empty When a directory is missing in one merge input (base or one side), we would consider that a merge conflict. This patch changes that so we instead merge trees by treating the missing tree as empty.
2022-04-18 03:51:24 +00:00
* If one side of a merge modified a directory and the other side deleted it, it
used to be considered a conflict. The same was true if both sides added a
directory with different files in. They are now merged as if the missing
directory had been empty.
rebase: update bookkeeping for branches as we rebase descendants The `DescendantRebaser` keeps a map of branches from the source commit, so it gets efficient lookup of branches to update when a commit has been rebased. This map was not kept up to date as we rebased. That could lead to branches getting left on hidden intermediate commits. Specifically, if a commit with a branch was rewritten by some command, and an ancestor of it was also rewritten, then we'd only update the branch only the first step and not update it again when rebasing onto the rewritten ancestor.
2022-04-23 05:58:16 +00:00
* When using `jj move` to move part of a commit into an ancestor, any branches
pointing to the source commit used to be left on a hidden intermediate commit.
They are now correctly updated.
cli: require at least one path for `jj untrack` It seems very unlikely that the user would want to untrack all paths (that's still possible with `jj untrack .`, if they really want to, and have added all their current paths to the `.gitignore`).
2022-04-09 16:20:00 +00:00
* `jj untrack` now requires at least one path (allowing no arguments was a UX
bug).
cli: require at least one destination for rebase Rebasing onto no commits (i.e. onto the root commit) is almost definitely not what you meant to do if you run `jj rebase` (without any arguments).
2022-04-11 17:40:56 +00:00
* `jj rebase` now requires at least one destination (allowing no arguments was a
UX bug).
cli: make `jj restore --to <rev>` restore from the working copy
2022-04-24 19:13:43 +00:00
* `jj restore --to <rev>` now restores from the working copy (it used to restore
from the working copy's parent).
cli: don't crash if $EDITOR fails
2022-04-09 22:53:32 +00:00
* You now get a proper error message instead of a crash when `$EDITOR` doesn't
exist or exits with an error.
cli: allow alias after global args, and recursive aliases Our support for aliases is very naively implemented; it assumes the alias is the first argument in argv. It therefore fails to resolve aliases after global arguments such as `--at-op`. This patch fixes that by modifying the command defintion to have an "external subcommand" in the list of available commands. That makes `clap` give us the remainder of the arguments when it runs into an unknown command. The first in the list will then be an alias or simply an unknown command. Thanks to @epage for the suggestion on in clap-rs/clap#3672. With the new structure, it was easy to handle recursive alias definitions, so I added support for that too. Closes #292.
2022-05-01 15:54:35 +00:00
* Global arguments, such as `--at-op=<operation>`, can now be passed before
an alias.
cli: use `.` in output for current directory instead of empty string
2022-04-20 05:15:38 +00:00
* Fixed relative path to the current directory in output to be `.` instead of
empty string.
cli: some fixes and touch-ups for `jj workspace`, plus tests
2022-04-22 04:49:57 +00:00
* When adding a new workspace, the parent of the current workspace's current
checkout will be checked out. That was always the intent, but the root commit
was accidentally checked out instead.
checkout: don't abandon old commit if it has non-empty description If the user entered a description, we shouldn't abandon it even if it has no changes to the content.
2022-05-21 17:23:46 +00:00
* When checking out a commit, the previous commit is no longer abandoned if it
has a non-empty description.
cli: make `move/squash/unsquash` keep empty working-copy commit If the source commit becomes empty as a result of `move/squash/unsquash`, we abandon it. However, perhaps we shouldn't do that if the source commit is a working-copy commit because working-copy commits are often work-in-progress commits. The background for this change is that @arxanas had just started a new change and had set a description on it, and then decided to make some changes in the working copy that should be in the parent commit. Running `jj squash` then abandoned the working-copy commit, resuling in the description getting lost.
2022-05-30 00:18:56 +00:00
* When using `jj move/squash/unsquash` results in the source commit becoming
empty, it is no longer abandoned if it is checked out (in any workspace).
release: release version 0.4.0 This release is mostly about the fix for #177, which looks pretty bad even though I think it is actually harmless. It also has `jj log -p` contributed by @yuja!
2022-04-03 05:56:59 +00:00
## [0.4.0] - 2022-04-02
config: drop support for `~/.jjconfig` I'm a little hesitant to do this because most tools I'm familiar with have the config file directly in `~/`. It's also easier to describe where to put the file if it doesn't vary across platforms. But we're still early in the project, so let's try it and see if we get any complaints.
2022-03-19 17:17:30 +00:00
### Breaking changes
* Dropped support for config in `~/.jjconfig`. Your configuration is now read
from `<config dir>/jj/config.toml`, where `<config dir>` is
`${XDG_CONFIG_HOME}` or `~/.config/` on Linux,
`~/Library/Application Support/` on macOS, and `~\AppData\Roaming\` on
Windows.
config: add support for a `$JJ_CONFIG` environment variable It's useful for tests, scripts, and debugging to be able to use specific config instead of the user's config. That's especially true for our automated tests because they didn't have a place to read config from on Windows before this patch (they read their config from `{FOLDERID_RoamingAppData}`, which I don't think we can override in tests).
2022-03-19 17:37:13 +00:00
### New features
* You can now set an environment variable called `$JJ_CONFIG` to a path to a
config file. That will then be read instead of your regular config file. This
is mostly intended for testing and scripts.
cli: respect `$NO_COLOR` environment variable
2022-03-19 17:00:13 +00:00
* The [standard `$NO_COLOR` environment variable](https://no-color.org/) is now
respected.
cli: teach `jj new` to set initial description with `-m` I rarely use `jj new`, so this feature is mostly for use in tests.
2022-03-26 21:02:04 +00:00
* `jj new` now lets you specify a description with `--message/-m`.
repo: don't abandon empty commit if it has descendants It's unusual for the current commit to have descendants, but it can happen. In particular, it can easily happen when you run `jj new`. You probably don't want to abandon it in those cases.
2022-03-26 21:53:51 +00:00
* When you check out a commit, the old commit no longer automatically gets
abandoned if it's empty and has descendants, it only gets abandoned if it's
empty and does not have descendants.
repo: when merging in removed head, rebase descendants (#111)
2022-03-17 03:58:04 +00:00
* (#111) When undoing an earlier operation, any new commits on top of commits
from the undone operation will be rebased away. For example, let's say you
rebase commit A so it becomes a new commit A', and then you create commit B
on top of A'. If you now undo the rebase operation, commit B will be rebased
to be on top of A instead. The same logic is used if the repo was modified
by concurrent operations (so if one operation added B on top of A, and one
operation rebased A as A', then B would be automatically rebased on top of
A'). See #111 for more examples.
log: add -p/--patch option to show diff along with commit meta data "log -p | less" is the option I often use with hg/git to find interesting bits from the changelog, and I think it's also valid with jj. Unlike "hg log -p --stat", "jj log -p --summary" does not show both diff summary and patch to reflect the internal structure. This behavoir is arguable and may be changed later. The logic of show_patch() is extracted from cmd_show().
2022-03-28 09:12:08 +00:00
* `jj log` now accepts `-p`/`--patch` option.
cli: fix git-colocated working copy on Windows This patch adds a very simple e2e test of having a working copy shared with Git. The test initially failed on Windows. The symptom was that the "master" branch did not get updated when we create a commit using `jj`. That suggested that we didn't correctly detect that the working copy was shared. After a lot of troubleshooting, I think I mostly understand what we going on here (thanks to @arxanas for suggesting https://github.com/mxschmitt/action-tmate). The path we get from `git2::Repository::workdir()` seems to not be canonicalized in the same way as `std::fs::canonicalize()` canonicalizes. Specifically, it does not have the "\\?\" prefix we get from that function. I suppose that's because libgit2 is a C library and canonicalizes the path using some other system call.
2022-03-29 06:03:37 +00:00
### Fixed bugs
cli: fix crash when initializing workspace colocated with Git repo When initializing a workspace that shares its working copy with a Git repo (i.e. `jj init --git-repo=.`), we import refs and HEAD when creating the `WorkspaceCommandHelper` (as we do for all commands when the working copy is shared). That makes the explicit import we do in `cmd_init()` unnecessary. It also makes the checkout of HEAD I added for the fix of #102 unnecessary. More importantly, as @yuja reported in #177, it makes the command crash (at least if the repo is small enough that the two checkouts happen within a second). I think the problem is that the second checkout tries to create the same commit except that the Change ID is different (the problem is not the predecessors as I speculated in the issue tracker). The fix is to simply avoid doing the redundant work. We still need a proper fix for #27 eventually. Closes #177.
2022-03-28 13:46:21 +00:00
* Fixed crash on `jj init --git-repo=.` (it almost always crashed).
cli: fix git-colocated working copy on Windows This patch adds a very simple e2e test of having a working copy shared with Git. The test initially failed on Windows. The symptom was that the "master" branch did not get updated when we create a commit using `jj`. That suggested that we didn't correctly detect that the working copy was shared. After a lot of troubleshooting, I think I mostly understand what we going on here (thanks to @arxanas for suggesting https://github.com/mxschmitt/action-tmate). The path we get from `git2::Repository::workdir()` seems to not be canonicalized in the same way as `std::fs::canonicalize()` canonicalizes. Specifically, it does not have the "\\?\" prefix we get from that function. I suppose that's because libgit2 is a C library and canonicalizes the path using some other system call.
2022-03-29 06:03:37 +00:00
* When sharing the working copy with a Git repo, the automatic importing and
exporting (sometimes?) didn't happen on Windows.
release: include the two recent no-op releases in the changelog I keep forgetting to update things while doing a release...
2022-03-17 06:01:42 +00:00
## [0.3.3] - 2022-03-16
No changes, only trying to get the automated build to work.
## [0.3.2] - 2022-03-16
No changes, only trying to get the automated build to work.
release: update release notes for 0.3.1 I forgot to do this when I bumped the version number.
2022-03-14 04:51:21 +00:00
## [0.3.1] - 2022-03-13
cli: expand `~` in `core.excludesFile`, don't crash when missing I thought that `std::fs::canonicalize()` expanded "~", but it doesn't seem to do that, which caused #131. Git seems to do the expansion itself, so we probably also should. More importantly `std::fs::canonicalize()` crashes when the file doesn't exist. The manual expansion we do now does not. Closes #131.
2022-03-13 15:14:01 +00:00
### Fixed bugs
- (#131) Fixed crash when `core.excludesFile` pointed to non-existent file, and
made leading `~/` in that config expand to `$HOME/`
release: add a changelog
2022-03-12 18:48:51 +00:00
## [0.3.0] - 2022-03-12
Last release before this changelog started.
Reference in a new issue Copy permalink