~theoryware.net - Stupid source "distribution" models

Stupid source "distribution" models

Posted: 2022-09-20 #rant #software
2 minute read.

Sometimes I laugh so hard at the audacity of developers who feel the need to “protect” the source code of any software project that they write. It always seems that there are people who just don’t get the overall benefit of what free software, but sill want a slice of the popularity pie.

As a preface: I am not a lawyer, nor do I consider my licensing opinons legal advice

Today’s case study is of irchiver, one of the most hilariously licensed software projects that I have seen to date. I have seen projects with absolutely terrible foss licences (BSD clause 1 moment), but this is on a whole different level of what in the hell.

First of all, the software itself (not the source) is distributed under a CC BY-NC-SA 4.0 International licence. The first problem with this is that CC licences are mainly targeted at creative works in media, not software, so there could be potential abuse of this. Secondly, there are two parts to this software, the search backend and the daemon that captures pictures of your browser (a security issue in my opinion). The search backend is “open source”, but without a licence, so therefore it is assumed to be copyright. Ironically enough, this precludes this software from being considered open source by the OSI definition of open source.

However, here is the crazy part, the capture software (doing something that could be considered very privacy violating), is not open source at all, its just “source-verifiable”. Pretty much, this guy just has a video of him compiling the software, then showing of a SHA-256 hash for us to verify that the binary is the same that he distributes. In my mind, the source-available model is more open than bullshit like this, even if under a copyright licence.

Additional nitpick: the overall presentation seems very startup-like. Thanks, I hate it.

Ultimately, the list of competitors that this software lists that is supposedly is “better” than, are probably a better bet if you are serious about web archival1. If you really want to get everything, you can just download a whole website too, not just individual pages like most of these tools do.2


  1. Archivebox is among this list and I can personally recommend it. ↩︎

  2. wget is a great tool for this, and it licensed under GPLv3 ↩︎

Have a comment or question? Shoot an email to ~theorytoe/public-inbox@lists.sr.ht, my public inbox.

Articles:  <= =>

Liked the article? Send it to a friend!