Bitbucket deadline postponed to July 1st

By Georges Racinet

Last week, Bitbucket updated their "sunsetting" announcement with the new July 1st, 2020 deadline. Here we explore some implications of this piece of good news, for Heptapod and the Mercurial ecosystem.

To quote the announcement: "Due to the impact of COVID-19, we are extending the deprecation date by 30 days to give you more time to transition".

Many people have found the original 9 months to be short. It takes time to explore alternatives, especially in a moving landscape: Since the original Bitbucket announcement, some new Mercurial hosting services have been launched, including of course our own These are getting stronger by the day.

We think this additional month could give Mercurial Bitbucket users the opportunity to reevaluate their plans, especially for those that were considering migration to another version control system: given the progress made by Mercurial solutions, is it really worth the added disruption?

There are a lot of inactive repositories on Bitbucket. Many of them are at risk to be lost forever. Fortunately, there is an ongoing archival effort of public repositories by Software Heritage in partnership with Octobus, the creators of Heptapod. More on this and how the Mercurial expertise of Octobus will be leveraged by Software Heritage can be read here. One month more time can only be good news for this rescue mission as well.

Let's turn ourselves to what Heptapod specifically has to offer right now and where it should be by the end of June.

Current state of Heptapod as Bitbucket replacement

As we've already stated, it was a primary concern at Octobus that someday Bitbucket could simply drop Mercurial support. Still, it happened faster and more brutally than we feared.

That means we had to shift priorities. That let us to spend lots of time with the rescue of Bitbucket repositories in mind:

These are achievements we can be proud of, and they should already be enough for many projects to migrate to Heptapod right now, be it on, or a self-hosted instance. Let's see now what one more month of development may bring us.

Heptapod in June (towards version 1.0)

We've always been very cautious about roadmaps, because we prefer to deliver rather than to make promises. The Bitbucket situation alone only underlines that all plans are liable to change due to external events. That being said, Heptapod has made enough progress these past few months that we can be more confident. Here are our current projections:

  • Early May: Heptapod 0.13 will bring back the support for wikis, of course with Bitbucket import. Development is well under way, things are looking good.
  • Early June: in Heptapod 0.14, Mercurial repository web browsing starts going native, as it's always been with a Mercurial client. In the meanwhile, GitLab will reach version 13.0.0, and we'll try to land on the closest 12.x as possible.

In short, there was a risk that the previous Bitbucket deadline would have found us still working on polishing Heptapod 0.14. Things will certainly be much more comfortable this way.

What next in June, then? Heptapod 1.0, of course! We've always had two major items to leave the realm of 0.x versions: catching up with upstream GitLab and full native Mercurial support.

At this point, it is a mere possibility. Whether we reach 1.0 before Bibucket's deadline or shortly after doesn't matter so much. Heptapod 0.14 will be a strong version anyway, we don't need to rush.