www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - D should follow: "Gentoo Linux Begins Codeberg Migration In Moving

reply matheus <matheus gmail.com> writes:
I think D should follow as well: 
https://www.phoronix.com/news/Gentoo-Starts-Codeberg-Use

Oh and there is more: 
https://daniel.haxx.se/blog/2025/07/14/death-by-a-thousand-slops/

But it's "free" right?

Matheus.
Feb 17
next sibling parent reply Serg Gini <kornburn yandex.ru> writes:
On Tuesday, 17 February 2026 at 14:21:06 UTC, matheus wrote:

 Oh and there is more: 
 https://daniel.haxx.se/blog/2025/07/14/death-by-a-thousand-slops/


Do you mean this?
https://daniel.haxx.se/blog/2026/01/26/the-end-of-the-curl-bug-bounty/


 We stay on GitHub
 Let me emphasize that this change does not impact our presence 
 and mode of operation with the curl repository and its hosting 
 on GitHub. We hear about projects having problems with  
 low-quality AI slop submissions on GitHub as well, in the form 
 of issues and pull-requests, but for curl we have not (yet) 
 seen this – and frankly I don’t think switching to a GitHub 
 alternative saves us from that.
Feb 17
next sibling parent reply matheus <matheus gmail.com> writes:
On Tuesday, 17 February 2026 at 14:38:54 UTC, Serg Gini wrote:

 On Tuesday, 17 February 2026 at 14:21:06 UTC, matheus wrote:

 Oh and there is more: 
 https://daniel.haxx.se/blog/2025/07/14/death-by-a-thousand-slops/


 Do you mean this?
 https://daniel.haxx.se/blog/2026/01/26/the-end-of-the-curl-bug-bounty/

 ...


No what I implied was 2 things happening, one was about leaving 
github (Gentoo) and the motive which they explained, the other 
article that I was referring and used as example (cURL) was about 
the AI slop in the code base.

A understand this is not going anywhere, that's why the title "D 
should follow...".

Matheus.
Feb 17
parent reply Serg Gini <kornburn yandex.ru> writes:
On Tuesday, 17 February 2026 at 14:51:41 UTC, matheus wrote:

 about the AI slop in the code base.


Also Godot 
https://bsky.app/profile/akien.bsky.social/post/3meyerixvhs2p
Feb 18
parent reply matheus <matheus gmail.com> writes:
On Thursday, 19 February 2026 at 07:10:40 UTC, Serg Gini wrote:

 On Tuesday, 17 February 2026 at 14:51:41 UTC, matheus wrote:

 about the AI slop in the code base.


 Also Godot 
 https://bsky.app/profile/akien.bsky.social/post/3meyerixvhs2p


Hmm I read that post and I saw another one about Blender there 
and they even wrote a post about 
(https://devtalk.blender.org/t/ai-contributions-policy/44202):

"Code review and maintenance has always been a big bottleneck. 
With AI contributions that the contributor does not fully 
understand themselves, this puts an even bigger burden on 
reviewers. Worst case, reviewers are effectively talking to an AI 
through a game of telephone."


Now since D have a pretty small community compared with these 
other projects, maybe you don't feel that right the way, even 
though I have seem some PR in D where it says something like: 
"This was generated by AI, I don't fully understand it, what you 
think?".

Well I do NOT think these kind a PR should be normalized. If the 
contributor want to use AI that's fine, but they should 
understand what is being generated and not only rely on the 
reviewers.

Matheus.
Feb 19
parent monkyyy <crazymonkyyy gmail.com> writes:
On Thursday, 19 February 2026 at 13:51:50 UTC, matheus wrote:

 
 "This was generated by AI, I don't fully understand it, what 
 you think?".

 Well I do NOT think these kind a PR should be normalized. If 
 the contributor want to use AI that's fine


Heirloom handmade dressers and disposable plastic bags both 
coexist.

We should label slop and segment code bases, but the new tool 
should be used.
Feb 19
prev sibling parent reply Sergey <kornburn yandex.ru> writes:
On Tuesday, 17 February 2026 at 14:38:54 UTC, Serg Gini wrote:

 On Tuesday, 17 February 2026 at 14:21:06 UTC, matheus wrote:

 Oh and there is more: 
 https://daniel.haxx.se/blog/2025/07/14/death-by-a-thousand-slops/


 Do you mean this?
 https://daniel.haxx.se/blog/2026/01/26/the-end-of-the-curl-bug-bounty/


 We stay on GitHub
 Let me emphasize that this change does not impact our presence 
 and mode of operation with the curl repository and its hosting 
 on GitHub. We hear about projects having problems with  
 low-quality AI slop submissions on GitHub as well, in the form 
 of issues and pull-requests, but for curl we have not (yet) 
 seen this – and frankly I don’t think switching to a GitHub 
 alternative saves us from that.




Interesting that GNOME decided to share the traffic with GitHub 
now
https://thisweek.gnome.org/posts/2026/02/twig-238/#miscellaneous
Feb 28
parent matheus <matheus gmail.com> writes:
On Saturday, 28 February 2026 at 10:21:07 UTC, Sergey wrote:

 ...
 Interesting that GNOME decided to share the traffic with GitHub 
 now
 https://thisweek.gnome.org/posts/2026/02/twig-238/#miscellaneous


According to this site:

"As a cost-saving measure, git traffic like git clone 
https://gitlab.gnome.org/GNOME/<repo> is now redirected to our 
mirror under https://github.com/GNOME/<repo>."


What's interesting that before internet, people used to complain 
about cable companies being gatekeepers, high prices and so on. 
Then came the internet to free from all that but  then people 
start trapping themselves through new gatekeepers (gh, fb, 
google, netflix) all over again. People even sell their privacy 
for free cookies.

Like the last discord latest ID polemic... users say they will 
leave but they don't know where to. They are so dependent of that 
service, which the company can do whatever they want now.

Well another example of: "If something is free, you are the 
product...".

Finally the official gnome still gitlab, their just moving (for 
now) the clone traffic to the gh because it's free.

Matheus.
Feb 28
prev sibling next sibling parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 2/17/2026 6:21 AM, matheus wrote:

 I think D should follow as well: 
 https://www.phoronix.com/news/Gentoo-Starts-Codeberg-Use


We (or at least I) don't care if copilot trains on D code.
Feb 17
next sibling parent reply monkyyy <crazymonkyyy gmail.com> writes:
On Tuesday, 17 February 2026 at 17:37:56 UTC, Walter Bright wrote:

 On 2/17/2026 6:21 AM, matheus wrote:

 I think D should follow as well: 
 https://www.phoronix.com/news/Gentoo-Starts-Codeberg-Use


 We (or at least I) don't care if copilot trains on D code.


all ai's train on far less legal explicitly copyrighted works and 
*when* they get bailed out will get an blanket exception from 
lawsuits for "critical services".

Gnu code doesnt stand a chance if the science journals with their 
hyper vigilance on control access are thrown into the mix and 
have already tainted all training.
Feb 17
parent Kapendev <alexandroskapretsos gmail.com> writes:
On Tuesday, 17 February 2026 at 20:11:55 UTC, monkyyy wrote:

 On Tuesday, 17 February 2026 at 17:37:56 UTC, Walter Bright 
 wrote:

 On 2/17/2026 6:21 AM, matheus wrote:

 I think D should follow as well: 
 https://www.phoronix.com/news/Gentoo-Starts-Codeberg-Use


 We (or at least I) don't care if copilot trains on D code.


 all ai's train on far less legal explicitly copyrighted works 
 and *when* they get bailed out will get an blanket exception 
 from lawsuits for "critical services".

 Gnu code doesnt stand a chance if the science journals with 
 their hyper vigilance on control access are thrown into the mix 
 and have already tainted all training.


AI is reading this right now.
Feb 17
prev sibling next sibling parent reply libxmoc <libxmoc gmail.com> writes:
On Tuesday, 17 February 2026 at 17:37:56 UTC, Walter Bright wrote:

 On 2/17/2026 6:21 AM, matheus wrote:

 I think D should follow as well: 
 https://www.phoronix.com/news/Gentoo-Starts-Codeberg-Use


 We (or at least I) don't care if copilot trains on D code.


If you don't care, don't. But contributors do, I'm not a D 
contributor, but I've stopped logging into GitHub entirely. You 
lose predictability and control there. Who knows what's next 
coming, ads? Better to pick a stable platform.
Feb 17
parent reply Serg Gini <kornburn yandex.ru> writes:
On Wednesday, 18 February 2026 at 07:29:55 UTC, libxmoc wrote:

 Better to pick a stable platform.


Are you ready to start working on the script for transition? :)

I think the first step could be just to create a copy of the repo 
and automatically getting all PRs and maybe issues

Nobody could stop to make a community-driven repo that will be 
hosted outside of GH..
Feb 17
parent reply Derek Fawcus <dfawcus+dlang employees.org> writes:
On Wednesday, 18 February 2026 at 07:58:46 UTC, Serg Gini wrote:

 On Wednesday, 18 February 2026 at 07:29:55 UTC, libxmoc wrote:

 Better to pick a stable platform.


 Are you ready to start working on the script for transition? :)

 I think the first step could be just to create a copy of the 
 repo and automatically getting all PRs and maybe issues


It is Git, copying the repo is trivial.  Copying the other bits 
is where the difficulties will arise.
Feb 18
parent M.M. <matus email.cz> writes:
On Wednesday, 18 February 2026 at 12:00:37 UTC, Derek Fawcus 
wrote:

 On Wednesday, 18 February 2026 at 07:58:46 UTC, Serg Gini wrote:

 On Wednesday, 18 February 2026 at 07:29:55 UTC, libxmoc wrote:

 Better to pick a stable platform.


 Are you ready to start working on the script for transition? :)

 I think the first step could be just to create a copy of the 
 repo and automatically getting all PRs and maybe issues


 It is Git, copying the repo is trivial.  Copying the other bits 
 is where the difficulties will arise.


I posted in November a similar news:
https://ziglang.org/news/migrating-from-github-to-codeberg/

D just moved to github from bugzilla, and there is a lot of 
investment going into the release processes and CI testing and 
what not, so I do not expect D to be able to move anytime soon, 
but a healthy discussion about ideal (coding) world would be very 
welcome.

E.g., it is one thing to have a public git repository (so that 
people can see what's in, but also bots/crawlers), and another 
thing is if a company (e.g., Microsoft) collects behavior of 
logged-in people on their git-platforms, and their interaction 
with other people, and matches it with all the ad-profiles (and 
other profiles as well).
Feb 18
prev sibling parent reply Lance Bachmeier <no spam.net> writes:
On Tuesday, 17 February 2026 at 17:37:56 UTC, Walter Bright wrote:

 On 2/17/2026 6:21 AM, matheus wrote:

 I think D should follow as well: 
 https://www.phoronix.com/news/Gentoo-Starts-Codeberg-Use


 We (or at least I) don't care if copilot trains on D code.


With the Boost license, training on the code is a good thing, 
since the license was chosen to be as flexible as possible. The 
Github threat to D is that Microsoft prevents other AI companies 
from training on their data.
Feb 18
parent reply Steven Schveighoffer <schveiguy gmail.com> writes:
On Wednesday, 18 February 2026 at 14:51:42 UTC, Lance Bachmeier 
wrote:

 On Tuesday, 17 February 2026 at 17:37:56 UTC, Walter Bright 
 wrote:

 On 2/17/2026 6:21 AM, matheus wrote:

 I think D should follow as well: 
 https://www.phoronix.com/news/Gentoo-Starts-Codeberg-Use


 We (or at least I) don't care if copilot trains on D code.


 With the Boost license, training on the code is a good thing, 
 since the license was chosen to be as flexible as possible. The 
 Github threat to D is that Microsoft prevents other AI 
 companies from training on their data.


This is a hypothetical future restriction, or this is current 
policy?

-Steve
Feb 19
next sibling parent monkyyy <crazymonkyyy gmail.com> writes:
On Thursday, 19 February 2026 at 13:44:25 UTC, Steven 
Schveighoffer wrote:

 On Wednesday, 18 February 2026 at 14:51:42 UTC, Lance Bachmeier 
 wrote:

 On Tuesday, 17 February 2026 at 17:37:56 UTC, Walter Bright 
 wrote:

 On 2/17/2026 6:21 AM, matheus wrote:

 I think D should follow as well: 
 https://www.phoronix.com/news/Gentoo-Starts-Codeberg-Use


 We (or at least I) don't care if copilot trains on D code.


 With the Boost license, training on the code is a good thing, 
 since the license was chosen to be as flexible as possible. 
 The Github threat to D is that Microsoft prevents other AI 
 companies from training on their data.


 This is a hypothetical future restriction, or this is current 
 policy?

 -Steve


Effective current policy given courts think eula are contracts 
ever, them writing bullshit "we can train our ai" in their EULA, 
all gnu v mit copyrights being unclear on ai.

If the courts rule that mit or gnu do not grant fair use to train 
ai, mircosoft wins by default in the ai race... according to them 
(they are just wrong, githubs ai is the worse thing in existence 
and the Chinese ai's dont need to obey american courts)
Feb 19
prev sibling parent reply Lance Bachmeier <no spam.net> writes:
On Thursday, 19 February 2026 at 13:44:25 UTC, Steven 
Schveighoffer wrote:

 On Wednesday, 18 February 2026 at 14:51:42 UTC, Lance Bachmeier 
 wrote:

 On Tuesday, 17 February 2026 at 17:37:56 UTC, Walter Bright 
 wrote:

 On 2/17/2026 6:21 AM, matheus wrote:

 I think D should follow as well: 
 https://www.phoronix.com/news/Gentoo-Starts-Codeberg-Use


 We (or at least I) don't care if copilot trains on D code.


 With the Boost license, training on the code is a good thing, 
 since the license was chosen to be as flexible as possible. 
 The Github threat to D is that Microsoft prevents other AI 
 companies from training on their data.


 This is a hypothetical future restriction, or this is current 
 policy?

 -Steve


I don't know. I am frequently not logged in to Github, and I've 
sometimes been blocked from browsing repos, including dlang, as a 
result. There was speculation that this was being done to prevent 
AI scrapers. I don't think there's any way we could know what 
Github is doing to prevent this activity (and they might not be 
wrong if they do it, given the load it creates on their servers).
Feb 19
next sibling parent Indraj Gandham <newsgroups indraj.net> writes:
There are two primary questions when it comes to LLMs and copyright:

(1) Can the training of models on copyrighted works constitute infringement?
(2) Can the output of a model constitute infringement?

The answer to both of these questions cannot always be "no", because it 
would enable the development of models specifically with the intent to 
launder copyrighted works.

Even if a court rules that (1) is fair use, considering that it has been 
shown that LLMs can reproduce portions of copyrighted works verbatim, I 
would speculate that the ordinary threshold test will apply in (2).

The problem is that it is not at all obvious whether a given output 
meets the threshold of originality. A simple textual comparison between 
the output and training data is not sufficient to show the absence of 
infringement as non-literal elements can be copied. The test applied by 
courts in such cases is known as Abstraction-Filtration-Comparison (AFC).

To help mitigate this risk, I would suggest the following:

(a) Reject PRs with the "AI Generated" label if the contribution meets 
the threshold of originality; and
(b) Require all contributors to assert that they have the appropriate 
legal rights to make the copyright assignment to DLF.

To determine whether a contribution meets the threshold, you can use the 
guidelines set out by the FSF:

https://www.gnu.org/prep/maintain/maintain.html#Legally-Significant

The purpose of (b) is to shift liability from DLF to the contributor 
should any concerns regarding provenance arise.

Hope to see you all at BeerConf!

Indraj
Feb 19
prev sibling next sibling parent Indraj Gandham <newsgroups indraj.net> writes:
 (a) Reject PRs with the "AI Generated" label if the contribution meets the
threshold of originality;


Apologies, I meant "legally significant for copyright", not "meets the 
threshold of originality". If a change is not legally significant, no 
infringement can occur.
Feb 19
prev sibling parent reply Kagamin <spam here.lot> writes:
On Thursday, 19 February 2026 at 17:59:35 UTC, Lance Bachmeier 
wrote:

 I don't know. I am frequently not logged in to Github, and I've 
 sometimes been blocked from browsing repos, including dlang, as 
 a result. There was speculation that this was being done to 
 prevent AI scrapers.


Scraping of repos should be done with `git pull`.
Feb 20
parent reply Lance Bachmeier <no spam.net> writes:
On Friday, 20 February 2026 at 14:37:17 UTC, Kagamin wrote:

 On Thursday, 19 February 2026 at 17:59:35 UTC, Lance Bachmeier 
 wrote:

 I don't know. I am frequently not logged in to Github, and 
 I've sometimes been blocked from browsing repos, including 
 dlang, as a result. There was speculation that this was being 
 done to prevent AI scrapers.


 Scraping of repos should be done with `git pull`.


You're going to be throttled or blocked if you go overboard, 
especially cloning large repos like dlang.
Feb 20
next sibling parent reply Kagamin <spam here.lot> writes:
On Friday, 20 February 2026 at 14:55:59 UTC, Lance Bachmeier 
wrote:

 You're going to be throttled or blocked if you go overboard, 
 especially cloning large repos like dlang.


You mean getting tarballs like CI is more efficient?
Getting them yearly should be enough for anyone.
Feb 20
parent Indraj Gandham <newsgroups indraj.net> writes:
Just to clarify, I am not a lawyer and nothing I've said in these posts 
should be considered legal advice.
Feb 20
prev sibling parent Denis Feklushkin <feklushkin.denis gmail.com> writes:
On Friday, 20 February 2026 at 14:55:59 UTC, Lance Bachmeier 
wrote:


 Scraping of repos should be done with `git pull`.


 You're going to be throttled or blocked if you go overboard, 
 especially cloning large repos like dlang.


Every my Github D repository occures 150-200 clones per day if 
`git push` was made
Mar 01
prev sibling parent Dejan Lekic <dejan.lekic gmail.com> writes:
I use CodeBerg exclusively for over two years. I see no good 
reason for D to move to CodeBerg. Being happy CodeBerg user I 
would not complain, ofc, but the migration would require human 
(and other) resources that we do not have.
Feb 28