It's 2020 And GCC Has Finally Converted From SVN To Git

[lamikr]

I think it will make sense for GCC to host their own servers for code due to privacy reasons that has always been important for gnu projects. Github offers nice tools and good infra but when you are using it you are always also giving some information about yourself that can be potentially useful also for example for the linkedin.

[starshipeleven]

Enough with the stupid conspiracy theories about Github and the Microsoft buyout being somehow tainted because Microsoft can monetize what's already public information anyway.

Usage statistics aren't public, yo

[CommunityMember]

This would be true if Github had been purchased by ...

Or even if not purchased. They were a US company (I have walked past the GitHub HQ a number of times). It is probably true that their previous legal team did not have the depth of the entire Microsoft legal team (money does requiring hiring even more lawyers), but the underlying regulations and responsibility to comply are the same. The details are often under various restrictions, but both GitHub and Microsoft have provided what are called "transparency reports" that provide the quantity of the impacts.

[duby229]

Enough with the stupid conspiracy theories about Github and the Microsoft buyout being somehow tainted because Microsoft can monetize what's already public information anyway. No one that's not wearing an aluminum foil hat cares. Nothing would stop Microsoft from mining the GCC servers already if they so cared. For that matter nothing stops any other "big data" company from doing the same thing.

The only legitimate reason not to use Github (outside of simply because they can) at this time from the standpoint of a public core infrastructure project point of view is that Microsoft, like any other corporation and individual IN THE WORLD is subject to the political and legal realities of its home nation. They would need to make sure all the core contributors are not under embargo from the West. This would be true if Github had been purchased by Google, Siemens in Germany, or some deep pocketed altruistic individual in Japan. They would all be subject to the export and international cooperation restrictions of their host country and they change nearly daily. Violate those restrictions in your native country at your own peril.

Regardless of whether MS has access to publicly available information, now they -own- your publicly available information. Call it paranoid or whatever, but MS is what MS is and history is already done and happened. While it may be true that quantum particles can erase their history, MS can't...

[stormcrow]

I'm not saying it has to make sense for you, I'm saying they will do their best to monetize it. Big Data is still a buzzword.

They have better control over the data because they own the infrastructure, extracting info is cheaper.

They also have access and power over your account, know your email and phone number (if you have two-factor auth) which is not something Gitlab has over self-hosted instances.


I'll tell you one thing, Github was never profitable, Microsoft didn't buy them for 7.5B as an act of selfishness towards the masses.

As usual with most Big Data the issue is more like "what bad decisions they can take by looking at the data, and how that will be bad for you".

For example after MS aquisition they started blocking contributors from nations facing US sanctions, because fuck them I guess. https://www.zdnet.com/article/github...ade-sanctions/

In other news, there are horribly broken algorithms in other Big Data analysis systems that resulted in a lot of kids ending in no-fly-lists as "alleged terrorists", when even a fucking noob would have figured out to put some kind of age filter on that to avoid this issue. https://noflylistkids.ca/en/home/

Enough with the stupid conspiracy theories about Github and the Microsoft buyout being somehow tainted because Microsoft can monetize what's already public information anyway. No one that's not wearing an aluminum foil hat cares. Nothing would stop Microsoft from mining the GCC servers already if they so cared. For that matter nothing stops any other "big data" company from doing the same thing.

The only legitimate reason not to use Github (outside of simply because they can) at this time from the standpoint of a public core infrastructure project point of view is that Microsoft, like any other corporation and individual IN THE WORLD is subject to the political and legal realities of its home nation. They would need to make sure all the core contributors are not under embargo from the West. This would be true if Github had been purchased by Google, Siemens in Germany, or some deep pocketed altruistic individual in Japan. They would all be subject to the export and international cooperation restrictions of their host country and they change nearly daily. Violate those restrictions in your native country at your own peril.

[starshipeleven]

Wow, thats insane, you can store information that's available for the the public in a database?

I'm not saying it has to make sense for you, I'm saying they will do their best to monetize it. Big Data is still a buzzword.

How is that different if you scrap information from Gitlab/Bugzilla?

They have better control over the data because they own the infrastructure, extracting info is cheaper.

They also have access and power over your account, know your email and phone number (if you have two-factor auth) which is not something Gitlab has over self-hosted instances.

In the context of the thread: What effect has MS as owner on this?

I'll tell you one thing, Github was never profitable, Microsoft didn't buy them for 7.5B as an act of selfishness towards the masses.

What mind-boggling (no-trivial) information can be concluded from the data dump?

As usual with most Big Data the issue is more like "what bad decisions they can take by looking at the data, and how that will be bad for you".

For example after MS aquisition they started blocking contributors from nations facing US sanctions, because fuck them I guess. https://www.zdnet.com/article/github...ade-sanctions/

In other news, there are horribly broken algorithms in other Big Data analysis systems that resulted in a lot of kids ending in no-fly-lists as "alleged terrorists", when even a fucking noob would have figured out to put some kind of age filter on that to avoid this issue. https://noflylistkids.ca/en/home/

[discordian]

https://blog.novoda.com/github-data-mining-101/

Wow, thats insane, you can store information that's available for the the public in a database?
So, enlighten me:

- How is that different if you scrap information from Gitlab/Bugzilla?
- In the context of the thread: What effect has MS as owner on this?
- What mind-boggling (no-trivial) information can be concluded from the data dump?

[Veerappan]

I use GitLab in house hosted on our own server so it will only go down if I make updates or mistakes.
It will be around as long as people care about it - it's free software (MIT License for the community edition).

Same here. My company (work at, not owned by me) is using self-hosted gitlab for our source control, and we'll probably continue to do that for the foreseeable future.

Our primary product is closed-source software, and giving up control of the servers that software is hosted on is not going to happen.

For myself, I use github for all of the open-source stuff I work on, but I wouldn't mind spinning up my own personal gitlab instance at home and tying it into a jenkins/travis/whatever instance for mini-projects I work on.

[slalomsk8er]

True but with a project this large, and fundamental honestly being independent does have some advantages.
Because now the work-flow can be change to fit the project itself and give some kind of ownership and agency to GNU members. Because GitHub and GitLab simply will not be around forever plus they do go down.

I use GitLab in house hosted on our own server so it will only go down if I make updates or mistakes.
It will be around as long as people care about it - it's free software (MIT License for the community edition).

[starshipeleven]

What is this effing bigotry? What can you mine off an open source project whose entire workflow is public?

https://blog.novoda.com/github-data-mining-101/