HIVE

Distributed Bug Tracker
 
Home |  Documentation |  Code |  Issues

What Is It?

hive is a distributed issue tracker in the spirit of Bugs Everywhere. Issues in hive are included as "merge-tolerant" plain-text files that can be version-controlled alongside source code. The package includes both a command-line interface and a web interface for interacting with issues, and, if you want, issues can also be edited via a plain-text editor.

Is It Any Good?

Yes.

Can I Use It?

Of course! hive is free/libre software, available under the terms of the GNU General Public License version 3+ .

You can install hive via pip:

$ sudo pip3 install hive_bt

or by cloning the repository and installing manually:

$ hg clone https://hz.mit.edu/hg/hive_bt
$ cd hive_bt
$ sudo pip3 install .

The Documentation page has more information about using hive.

Do I Want To Use It?

That's a more complicated question. For one thing, hive is very much in a "pre-alpha" stage. In short, this means that, until a stable version is released, some things will only be partially implemented, some things will be broken, and some things may change dramatically without much notice. If you're OK with that, read on!

Beyond the fact that hive is very new, there are pros and cons associated with distributed version control in general, and there are pros and cons associated with this particular implementation. For example:

Pros:

• With hive, bugs and code are tracked together. Among other things, this means that merge requests can include not only changes to the code, but also changes to the relevant issues. It also means that it is possible for each branch to contain its own information about bugs that are relevant to that branch, which may be helpful when maintaining different branches for, for example, development and production versions of a piece of software. Merging branches can also merge changes to the issues.

• Since the issues are stored as part of the repository, browsing and modifying issues does not require an Internet connection.

• Issues are stored in a format that makes it impossible for merge conflicts to occur for most operations on issues.

• hive is decentralized and platform-agnostic. It currently supports both Git and Mercurial. It can be used on local repositories with no public access, and for public-facing repositories, it does not require using any particular platform to host your repository.

Cons:

• hive stores data in the repository, and its particular format can use a large number of files. This can mean that the repository contains many more files than it otherwise would, and that the history may contain commits related only to modifying the state of issues.

• hive can be slow. I'm not sure you'll notice this until you reach thousands of issues, but the particular choice of internal representation means that certain operations can be slow. Internal caching mechanisms help with this a little bit.

Other Distributed Bug Trackers

If you want a distributed bug tracker but are worried about this particular implementation, you might want to consider:

• Bugs Everywhere
• git-bug
• Artemis

Contact

If you want to get in touch, report a bug, send a patch, etc, feel free to send me an e-mail at hz@mit.edu.

Thanks!

hive is written in Python and relies on a few great pieces of software:

• the command-line interface is implemented using click, and it also makes use of the termcolor package for generating certain kinds of colored output. the progress bar that shows when running building the cache uses the tqdm package.
• the web interface is implemented in the flask web framework and uses Python-Markdown for interpreting Markdown in the bodies of issues and comments.