Updating Rigbox

This guide contains infomation on how to update Rigbox and how to switch back to older versions. Updates can either be done automatically via Git or manually.

Contents

Before you update

It's often useful to read the release notes or CHANGELOG. The former provides an overview of all changes, including a 'major changes' section which details the changes that may affect your analysis code or how experiments are run. For instance changes to function names and signature are considered major changes.

Note that you will always be able to go back to previous releases (more below).

Finding the current version

The current version can be found by running git.repoVersion in MATLAB if you installed Rigbox via Git, otherwise the CHANGELOG file will tell you what version of Rigbox you're currently using.

Automatic updating

Update checks can be made when running either mc or srv.expServer. The 'updateSchedule' field in your dat.paths file determines when new code may be pulled from the master (most stable) branch. Options include weekdays (e.g. 'Monday'), 'Never' (to never update), and 'Everyday' (to update every day). String or cell arrays can be used to update on multiple days:

% Check for updates on Mondays and Fridays
p.updateSchedule = ["Monday", "Friday"];

When updateSchedule is set for a particular day, the code is fetched on the day given, or if the code hasn't been fetched in over a week. If the code has already been fetched for that day it won't be fetched again until the next scheduled day. When updateSchedule is set to 'everyday', the code is fetched at most once an hour.

NB: This requires Rigbox to be installed via Git (instructions can be found here). The location of the Git Bash executable should be set in the 'gitExe' field in the dat.paths file. The default path can be found in the paths_template:

% Location of git for automatic updates
p.gitExe = 'C:\Program Files\Git\cmd\git.exe';

Updating manually via Git

You can check for updates (and download them) by running git.update in the MATLAB command window.

To do this directly in Git:

  1. Open Git Bash
  2. Change directory by typing cd and the path to your main Rigbox folder. You can use a tilda as shorthand for C:/Users/<username>, e.g. cd ~/Documents/Github/Rigbox
  3. Type git pull --recurse-submodules. The recurse flag ensures that signals, alyx-matlab, etc. are also updated.
  4. Check the end of the message. Ensure you don't see this messgae:
CONFLICT (content): Merge conflict in path/to/example_file.m
Automatic merge failed; fix conflicts and then commit the result.

This message means that there was a local change to the code which can't be merged with the recent remote changes. If you're happy to discard your local changes you can run this command:

git reset --hard origin/master
git pull --recurse-submodules

This resets the code to the current remote master branch state. The second line ensures that the other modules are updates too.

Switching between releases

On the all releases are tagged with a version number. The most recent version, including patches, is on the master (stable) branch and can be updated by following the instructions in the previous section.

To go back to a previous release version first list the versions available:

git.listVersions;

To switch to one of the available versions run the following in the MATLAB command window:

git.switchVersion('2.4.0'); % Switch to version 2.4.0
git.switchVersion('latest'); % Switch to latest version
git.switchVersion('previous'); % Switch to the previous version
git.switchVersion(2.3); % Switch to the most recently patched version 2.3

In Git Bash this can be done with the following command:

git checkout tags/version 2.1

NB: Once you've switched to a specific version, the code will no longer update.

If you believe that the most recent commit on the master branch has caused a problem, either switch to the previous version (see above) or revert the most recent change with the following command in Git Bash:

git reset --hard master@{"10 minutes ago"}

NB: By reverting the code in this way, the code will still update at the next scheduled time.

Updating without Git

You can download the source code for the most recent version manually:

  1. navigate to the Rigbox GitHub repository Webpage
  2. click the green button on the right-hand side that says 'Clone or download'
  3. in the pop-up balloon click on 'Download ZIP'
  4. replace the previous code with the code extracted from the zip file

To download a previous version:

  1. go to https://github.com/cortex-lab/Rigbox/releases
  2. under the release you wish to download, click the 'Assets' dropdown and download the Source code zip file
  3. replace the previous code with the code extracted from the zip file

Etc.

Author: Miles Wells

v1.1.0


Published with MATLAB® R2019b