{"id":7339,"date":"2020-02-29T11:08:34","date_gmt":"2020-02-29T11:08:34","guid":{"rendered":"https:\/\/gtechbooster.com\/?p=7339"},"modified":"2023-04-01T01:36:46","modified_gmt":"2023-04-01T01:36:46","slug":"fsf-to-launch-free-hosting","status":"publish","type":"post","link":"https:\/\/gtechbooster.com\/fsf-to-launch-free-hosting\/","title":{"rendered":"FSF to launch free hosting"},"content":{"rendered":"\n

Free Software Foundation intends to launch a “fully free” public code hosting and collaboration platform this year. Rather than creating a new forge, it is going to adapt an existing one with Fedora Pagure currently the front runner. <\/p>\n\n\n\n

\"\"<\/a><\/div><\/div>

The idea of establishing another code hosting platform to complement \nthe FSF’s existing Savannah software forge was included in an end-of-year blog post<\/a> by Ruben Rodriguez who explained:<\/p>\n\n\n\n

We know that many software projects \nand individual hackers are looking for more options on code hosting \nservices that focus on freedom and privacy, so we are working to set up a\n public site that anybody can use to publish, collaborate, or document \ntheir progress on free software projects. We will follow strict criteria\n to ensure that this code repository hosts only fully free software, and\n that it follows the very best practices towards freedom and privacy.<\/em><\/p>\n\n\n\n

YesterdayAndrew Engelbrecht reported progress towards the new forge and his post<\/a> also outlined why the FSF feels it is necessary:<\/p>\n\n\n\n

Infrastructure is very important for \nfree software, and it’s unfortunate that so much free software \ndevelopment currently relies on sites that don’t publish their source \ncode, and require or encourage the use of proprietary software.<\/em><\/p>\n\n\n\n

Engelbrecht went on to refer to the GNU ethical re<\/a>pository criteria<\/a> as\n setting the standard for free software code hosting and how in the \nsearch for an existing forge for FSF to work with the aim is to find one\n that would score a B rating (Good enough to recommend) and then work \ntowards an A (Excellent) rating after launch. He states:<\/p>\n\n\n\n

Reaching a B will require LibreJS<\/a> support,\n no third party tracking, proper license information, and more. We also \ncame up with a list of practical criteria, which includes two-factor \nauthentication (2FA), high performance, being well supported upstream, \nand other common forge features.<\/em><\/p>\n\n\n\n

According to the Forge Software Evaluation<\/a> report\n on LibrePlanet 15 different forge programs have been evaluated and \neliminated from the search. The only one for which details are provided \nis GitLab which currently has an ethical criteria rating C (Acceptable \nhosting for a GNU package). However the evaluation states:<\/p>

\"\"<\/a><\/div><\/div>\n\n\n\n

 listed as C, but has been operating \nat an F and will be reclassified soon because it sometimes requires \nusers to run non-free Google ReCAPTCHA code they have been very slowly \nworking on moving away from for almost 2 years now.<\/em><\/p>\n\n\n\n

There are three candidates still being evaluated. Gitea has as its first  Pro that “LibreJS is partially supported”<\/em>, but Cons start with “Poor usability with no js from site”<\/em>. Sourcehut on the other hand is said to have  “Perfect usability with no js from site”<\/em>, plus some other useful features. However among its Cons we find:<\/p>\n\n\n\n

No Web merge request UI. A Web merge \nrequest can be done by creating a ticket and including the forked git \nrepo url and branch name, but this missing feature seems like a key \nthing we are looking for.<\/em><\/p>\n\n\n\n

The third candidate is Pagure which has the annotation “(seems most likely, its our current focus)” <\/em>and the following lists of good and bad points:<\/p>\n\n\n\n