Hacker News .hnnew | past | comments | ask | show | jobs | submitlogin
Google Code is going readonly in about a day – update URLs in project pages now (code.google.com)
18 points by i336_ on Aug 24, 2015 | hide | past | favorite | 6 comments


DEADLINE: The banner above every GC page says August 25, the Wiki page above says August 24. So let's just say "ACT IMMEDIATELY."

Do you know anyone (who knows anyone, ...) with anything up on GC? Go and tell them IMMEDIATELY to edit their project(s') homepages to add new homepage URLs. Remember that the URLs will need to be stable for a couple of years at least, since it will soon be nontrivial to change them.

To summarize the link,

- Google Code will go readonly in 1-2 days

- After Jan 2016, `git`, `hg`, `svn` - and possibly the `[Export to GitHub]` button - will all break, and the data behind these endpoints will be made available in `.zip` and `JSON` formats - provided it is PUBLIC: private data (`Restrict-View-` et. al.) will go away

- Archived data (`.zip` and `JSON`), to quote, "will remain online for a long time."

If you still need to export your data, [I heard](https://www.reddit.com/r/programming/comments/3i58k5/google_...) that [fast-export](https://github.com/frej/fast-export).py will cleanly turn a `hg` repo into a Git one without getting confused like GitHub's exporter sometimes can.

Apparently you'll be able to email google-code-shutdown@google.com after the deadline and have project pages edited manually (you can apparently even have "redirects" set up, not sure if this is a `Location:` redirect or something else) - but this will not be instantaneous, especially not to begin with (RIP this address's inbox in a couple days).

(PS. You might spot this message elsewhere online - not trying to spam, just trying to help everyone.)


I work at Google on Google Code, so I am happy to clarify a few things.

# What will happen between August 2015 and January 2016

The vast majority of projects will be "read-only", but everything will pretty much work the same. The frontend will look the same. You can brows issues, downloads, wikis the same. You can even sync code using a DVCS client like svn, git, or hg.

i.e. all the data that was there right now will stay there until January.

Projects will still be able to be exported to GitHub via code.google.com/export-to-github. In addition, project administrators can still use Google Takeout to get a JSON dump of their projects issues.

There are a few projects that will stay read-write for a few months, for example /p/chromium and /p/android. These projects will keep doing their issue tracking on Google Code until a replace issue tracker is ready.

If a project needs some administrative action, such as deleting a project or setting up a "project moved" URL you can contact google-code-shutdown@google.com and somebody (most likely me) will twiddle the bits on your project.

Important note: All links to code.google.com will continue to work as normal.

# Coming soon: The Google Code Archive!

Obviously deleting ~a decades worth of open-source project data and breaking millions of links would be a very bad thing. Nobody on the Google Code team wants that to happen, so we are working to ensure that doesn't happen.

Tomorrow we will hopefully launch "The Google Code Archive" which is a slimmed down frontend for Google Code that renders an archive of project data, served from App Engine and Google Cloud Storage.

This Archive site will continue to host public (and only public) Google Code project data years into the future.

While it is in beta, the Google Code Archive will use a different URL scheme than current Google Code projects. Later this year, we will have the Google Code Archive frontend replace URLs to the old project hosting frontend. So urls to old Google Code projects will continue to work, but be hosted from a new website. (But you should be able to switch to the old one if need be via URL parameter.)

# What will happen after January 2016

After January 2016, we will turn down the old frontend and only serve data from the Google Code Archive.

The data that won't be preserved is private data. Things like issues labeled with Restrict-View-* (which only project admins, committers, etc. can view). Or projects that are marked as "Members Only". I haven't analyzed the exact numbers here, but this doesn't make an appreciable amount of the data on Google Code.

The big difference in January however is that we won't have the DVCS frontends running any more. So you will not be able to sync project source code using svn, git, or hg. Another outcome of this is that we won't be serving raw repo contents.

Hopefully that clears things up. Happy to answer questions.


Forgot to add that in January, 2016 we will no longer the Google Code-to-GitHub exporter service. (Since the DVCS frontends won't be running.)

However, you will be able to download a tarball of repository contents (including the .git or .hg folder) from the archive.


I was considering asking if anyone at Google could shed some more light on this - thanks for taking the initiative and making accounts both here and on Reddit just to reply!

You pretty much covered everything I could think of and more. :)

The only thing I thought of would be to filter all of Google Code for projects with private data, then mass-email a form letter to everyone relevant telling them their non-public data soon won't be available. I don't have anything up on Google Code so I have no idea if you're not already doing something like this.


Just curious: Will Code staff eventually transition to 1 or N other departments, or are most already maintaining other projects?


Most are already working on other projects. For example, the people who are experts in the inner workings of Git and Mercurial are part of the Cloud Source Repositories project (https://cloud.google.com/tools/cloud-repositories/docs/).




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: