[shagie]

The first part sets up the second part that unlike the GPL, the MIT license doesn't require that future contributions to the project be any particular license.

Presumably the OP can find the open source project MIT licensed without the company's contributions to it.

The only thing that the MIT license requires is:

> The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

It doesn't even require attribution in the compiled application or any other notices.

And so, it is possible (and I would dare say likely) that the contributions that the OP made while working on the repo at the company unless specific permission was given otherwise would be considered as work for hire or as part of the work product as condition for employment and completely owned by the company (and not MIT licensed).

If the OP thinks that they should have access to the code because of the MIT license, that is something to take up with a lawyer. My IANAL senses suggest that that would be rather fruitless.

I don't find GitHub's model particularly surprising but rather the most reasonable one that opens GitHub up to the least liability for accidental disclosures of content. Error on the side of least privilege and if there's something to work out beyond that, that's something for the contributors to work out themselves - GitHub isn't the arbitrator for that.

[eyelidlessness]

> And so, it is possible (and I would dare say likely) that the contributions that the OP made while working on the repo at the company unless specific permission was given otherwise would be considered as work for hire or as part of the work product as condition for employment and completely owned by the company (and not MIT licensed).

This is a radical interpretation of the text if I’ve ever seen one. To the extent any of their contributions were merged upstream, they’re inherently MIT licensed by virtue of being in the same codebase which offers that license. To the extent they have unmerged changes, they may well be works for hire but it isn’t GitHub’s role to decide that between a second and third party.

Again nor do they want to. GitHub is extremely hands off about forks and the licensing implications thereof.

This isn’t a GH posture towards licensing disputes, it’s their posture towards their own authorization model. And that’s fine, but we shouldn’t conflate the two when they’re quite distinct.

[shagie]

Who's upstream?

Open source project and publicly accessible upstream? Yep. They're likely MIT licensed in accordance with the CLA.

Internal company copy of an MIT project? Unlikely unless legal says they are.

If the organization that I worked for made an internal copy of Influxdb ( https://github.com/influxdata/influxdb ), my contributions to the private and internally hosted copy are not inherently MIT licensed too and furthermore don't need to be redistributed.

If those changes were submitted upstream to the influxdata/influxdb repo, and I signed the CLA with them - then yes they would be.

What's more, if I left the organization, then I don't necessarily have a right to the contributions that I made to the private and internally hosted copy's codebase.

[eyelidlessness]

It doesn’t sound like we have any substantive disagreement other than potentially what GitHub’s role in it is or should be.