in fact the development of Pharo 7 has already started. But in a limited mode.
For Pharo 7 we have the #development branch on GitHub (pharo-project/pharo)
…and we are already merging pull requests.
We currently have (mostly technical) problems with the infrastructure so the monkey is not running. But we have a temporary CI jobs that can check specific pull request (that means that we bootstrap Pharo for them and run tests for all platforms).
The latest Pharo 7 image can be found here:
The image is currently not uploaded to files.pharo.org and it is not accessible by zero-conf scripts, sorry.
If you want to try to create own pull request directly from Pharo 7, you need to do this:
– create own fork of pharo-project/pharo repository on GitHub and keep it in sync
– download Pharo 7 image
– clone pharo-project/pharo repository into the image directory, the directory with the clone MUST be named “pharo-core”
– checkout the development branch or better checkout the commit from which the Pharo 7 image was bootstrapped (SystemVersion current commitHash). Beware, in that case you will be in detached head state (you need to create own issue branch later to be able to commit something). The checkout to a particular commit cannot be done currently from Iceberg, you need to use git
– add your fork repository as a remote and mark it as default push target
– create issue on FogBugz to get issue ID
– create a branch for this issue (can be done easily from the repository context menu in Iceberg [Pharo], this option is valid only if you cloned pharo-project/pharo repository)
– commit the fix to your fork repository
– create the pull request (can be done easily from the repository context menu in Iceberg [Pharo], this option is valid only if you cloned pharo-project/pharo repository)
– add link to the pull request into the FogBugz issue entry
Pull requests can be done from Pharo 6 too. Then you do not need to have clone in “pharo-core” directory, it can be everywhere, and you mast create your branches from the master branch (and create pull requests on #development)
Because we currently have limited possibilities to check pull requests, we prefer to merge PRs needed for start of the new integration process. We really do not need to break the system in this fragile state 😉