Free Software Directory talk:Antifeatures

From Free Software Directory
Jump to: navigation, search



  • Minified: "Minification (also minimisation or minimization), in computer programming languages and especially JavaScript, is the process of removing all unnecessary characters from source code without changing its functionality. These unnecessary characters usually include white space characters, new line characters, comments, and sometimes block delimiters, which are used to add readability to the code but are not required for it to execute." However, minification is often used by people who want to "protect" their work from being modified.
    • Minification is equivalent to compilation, and just another form of non-source (aka binary) code. It is not an anti-feature, it's simply not the source code. Iank (talk) 20:52, 12 November 2017 (EST)
  • Service as a Software Substitute (SaaSS). Other terms which may refer to the same idea: Tethered, Nonfree Network Services
    • Tethered - Program depends on communication with a particular server, and is not very useful without that.
      • Stallman says: But maybe such programs should not be in the directory at all. Is there any reason to include them? Could you describe the examples?
      • If the front end is free back the back end isn't, and isn't otherwise useful, then we probably shouldn't include it - Donaldr3
  • Support channels/places that suggest people to use non-free functional data (generally: software, but not always).
    • Example (fictional): A user asking for support on VLC due to some bug or impossibility to do something, and receiving a reply suggesting him to use some non-free software.
    • Exception would be: for testing or development of the free software, or some free/libre software replacement for such non-(free/libre) ones, as described at Is It Ever a Good Thing to Use a Nonfree Program?.
  • Software known to have grave flaws – Software that fails to fully do the job that it says it does.
    • This was discussed on the mailing list. RMS argued against it, most notably by saying: "We don't judge whether a program is truly useful or not. Anyway, different users might judge a program differently in regard to its utility."
    • I'm not a fan either. It is also too broad. --IanK (talk)
  • Outdated upstream fork: Software that is not based on the latest version of the software it is derived from. Upstream forks is software that is not a straight fork; instead, they are parallel efforts that works closely with and re-bases in synchronization on the latest base software as the upstream supplier.
    • Outdated software is a very common phenomenon and not an easy one to judge, I don't see why this subset deserves a tag. --IanK (talk)
  • Discrimination against tor users, other other ways that are hostile to user privacy in documentation, forum, website, such as requesting real name registration before reading. users of shared IP addresses (often Tor) have chosen not to be tracked and thus not to share their unique and personal IP address. When servers hosting documentation give these users degraded treatment that is unequal to others, this is an attack on user freedom. By extension it's also an attack on the freedom of all those who would otherwise benefit from cover traffic generated by the repressed user.
    • This was added to the main page, but did not really fit well into the scope of the tag it was on, so I removed it and added it here for discussion. It at least needs a good name before being re-added. --IanK (talk)
      • I don't see how there is a scope or naming problem. Can you be more specific? When a freedom is reduced, limited, or attacked in some way, "nonfree" is the typical FSF adjective. In this case, the freedom at issue is with respect to access to documentation. Someone changed my naming from "closed documentation" to "nonfree documentation". I think that was a good change. Ian, what specifically wrong ("unfit") with "nonfree documentation" as nomenclature for the issues that were enumerated?
        • The gnu philosophy tries to as much as possible to first, make free and nonfree software/documentation a binary distinction. As concerning software,, and for documentation is basically the subset of that which applies, or as the gfdl puts in it's intro "to make a manual, textbook, or other functional and useful document "free" in the sense of freedom: to assure everyone the effective freedom to copy and redistribute it, with or without modifying it, either commercially or noncommercially". So, if your freedom is restricted, for example because the authors don't publish on the internet and require you to buy a physical cd in the mail, or they track your ip address, it is still free. it is not nonfree. That is the crux of why I removed what you wrote from the category "nonfree documentation". Obstructed freedom, restricted freedom, or some term like that would apply, but generally after still qualifying it as still being free; eg. free, but obstructed. --IanK (talk) 17:35, 14 January 2017 (EST)
          • The failure here is your attempt to define documentation freedom as software freedom. Documentation is different. It doesn't make sense to (for example) have "freedom to run the documentation as you wish", because documentation is not executable. The flaw in your argument is saying that we must necessarily be limited to a *subset* of software freedoms, as opposed to selecting freedoms that affect documentation users. Cypherpunks (talk) 08:59, 15 January 2017 (EST)
            • Note, I said "basically" and cited another definition because I'm not claiming it's a perfect subset. The core reason I deleted it still stands. --IanK (talk) 07:32, 19 January 2017 (EST)
              • It doesn't "stand" because you've not established that the GNU Free Documentation License should be disregarded. Your deletions undermine what the GFDL and the opposition of SaaSS try to accomplish.
        • To add to that, some restrictions on freedom would push the work into the category of nonfree, and some wouldn't. For example, if it's license says it can't be distributed in a specific country, that would make it nonfree. --IanK (talk) 17:57, 14 January 2017 (EST)
      • The new version is vague. What does "freely licensed" mean? Wild speculation leads one to think that somehow there is an attempt at separating between 'contributors' and 'consumers' of the documentation. But OTOH, the removed text impacts the freedom of contributors just the same. If "freely licensed" means "GPL", should that be mentioned or linked to a definition of sorts? In any case, it seems like a good idea to include licensing as an additional criteria to what was there.
        • Freely licensed means free, as in it's available under a free software or free documentation license, where the free is like the freedom in my above bullet point. --IanK (talk) 17:35, 14 January 2017 (EST)
      • The division seems to be a premature evolution. That is, if an antifeature catagory were too broad, the negative effect that would manifest is that users see rampant tagging and get conditioned to accept it. But that's not the case. There are only 4 known projects to have this flag. Consider usenet, which has a sensible approach to grouping. That is, if for example rec.bicycles becomes swamped with e-bike chatter, only then does it make sense to introduce more division, and thus create a rec.bicycles.e-bike group. But introducing division in a very low volume group, and the division is a bad idea. Hence, such attempts are voted down by the usenet panel. IMO it's more sensible to have a "nonfree documentation" category for all related matters. If many documentation projects are in a widespread unacceptable state, and at the same time the need to split the category presents, then that is the time to do so. But I don't envision that.
        • I'm not opposed to a category that includes all the things related to documentation, just the current name would not fit well for that. So, if I used too broad as a justification, that was probably me using the wrong word. Note, if you are wondering if the original "closed documentation" would be good for this, I think "closed" in this context fits with, so I do not think it's a good name. --IanK (talk) 17:35, 14 January 2017 (EST)
          • Whether "closed documentation" is a good name or not is not a point of contention. David improved the name, and I agreed. David demonstrated how to make a civil change to a wiki. Despite David and I agreeing with the name, you've decided otherwise and also that text should be removed. Of course all the items that were enumerated were there because they each singly attacked user freedom.Cypherpunks (talk) 08:59, 15 January 2017 (EST)
      • If nomenclature division is needed, then inventing a new name on the fly is less intrusive than deleting content. In order to not disrupt progress, Ian could have simply renamed the section to "Foo" or "TBD", and let the wiki make forward progress. Why delete the content? It's rather hasty, haphazard, and off-the-cuff. Users are demoralized and contributions are discouraged when someone cavalier is fast with the /delete/ button.
        • Sure. That is one way, it would have been ok. Generally, I like to leave unfinished work off the current edit and put it elsewhere. I think that does the best service to readers, and it seems that wikis which leave unfinished work on the main edit tend to leave it there for a long time, and it's not worth it to have the page quality impacted by it. --IanK (talk)
          • A wiki is *always* unfinished. It is a work in progress at all times. This is not YOUR work. This is a community effort. When you delete someones text, you block them and others from making improvements. It's a step backwards. Cypherpunks (talk) 08:59, 15 January 2017 (EST)
      • I've only just started contributing, and this is the 2nd time I've seen someone too eager to delete my work. I'm a volunteer, contributing w/out being on a payroll. Earlier I asked the question in here: "who is in control?", and my question was deleted before I could see an answer. Feels like I'm working with a loose cannon here. Whoever is moderating shouldn't have a reason to conceal that they are the moderator.
        • I'm not sure this addresses your concern completely. I'm also a volunteer. I'm the same as you. In general, improvement via "editing" of a document to improve it includes deleting parts. It's not meant to be hostile, it's not moderation. Often it's the start of a conversation, one person deletes something, gives a reason in the edit field, the other person can undelete it, responding to why the first person deleted it. If it's likely that further discussion than a single sentence or so is warranted, discussion is invited elsewhere, usually the talk page. I'm not an expert at mediawiki norms like some wikipedia editors, so this is just my impression. Deletion (edit: or more generally, reversion of an edit) is generally not seen as hostile unless it's repeated more than once on the same content without a reason, or any new reasoning. Also, for longer discussions like this, I'm always up for talking on irc, or even over voice chat, as that can often be more efficient and friendlier. Also, on talk pages, its helpful to sign things you say with a dash or two, then 3 tildes to put your username, or 4 to include the date. I'm IanK on freenode. --IanK (talk) 17:35, 14 January 2017 (EST).
          • Deleting content is an improvement if the content was spam, or if it was redundant. In this case, carefully worded user freedoms were discarded, and not replaced. It's very hostile to delete someones work to "start a discussion", like punching someone in the face to get their attention. Who are you to decide for everyone what freedoms can be discarded, if you don't have any authority at FSF? Cypherpunks (talk) 08:59, 15 January 2017 (EST)
            • Deleting content is justified for many more reasons which I would generalize as article quality. An edit can have a good quality, but it does not mean it's bad qualities outweigh the good. Say there is a wikipedia article "dogs", and someone adds to it, "I walk my dog. I do it at 3 pm, after I get home from school .... you should walk your dog after you get home from school", keeping an edit like that is not justified by something like "but there is no mention of walking dogs, there should be a mention of that, and my edit mentions it, so it should not be removed unless it's replaced by a better mention of walking dogs." It doesn't work that way. --IanK (talk) 07:32, 19 January 2017 (EST)
              • This is a fallacy of analogy. Dog walking is not central to the definition of a dog. Non-conformities with the GNU Free Documentation License are very central to the meaning of "nonfree documention". Discarding the text is both malicious and reduces the quality. It also disservices users who need an appraisal of the freedom they will have with a application (and documentation thereof).


Indirectly promoting the use of non-free software or non-ethical platforms while free/libre solutions exist. Example: Discord, facebook, twitter, github etc...

0 listed:

Declined suggestions

  • Nonfree software support - The software support nonfree software but free software can be used as well. For example, Django supports both PostgreSQL and the proprietary Oracle Database.
  • Nonfree software dependencies
    • Decline: I don't think the directory page should mention this AT ALL. - RMS
  • Nonfree software promotion
    • Its meaning is rather vague. - RMS

F-Droid excluded categories

  • Most categories in are not included in our antifeatures list
    • Nonfree Network Services
      • network services arent free or nonfree, so we would not use this tag. Consideration of service as a software substitute, saass is above. --IanK (talk)
      • "The term is not clearly defined. I think we most of the time were actually talking about the code aspect of it: Does an app rely on a networking service that is not (by our knowledge) run or easy to reproduce with free software. Sigh.. don't think that makes it any better. Maybe some examples help: Telegram requires interacting with a server which source code has not been released, so its NonFreeNet. On the other hand, serving html files from an MS IIS http server is not affected, because you can easily replace it with your favored one. -- Krt (talk) 08:56, 24 November 2016 (UTC) " -
    • Non free Dependencies
      • This kind of software would not work on a free gnu/linux distribution, so it would not be included in the directory. --IanK (talk)
    • Upstream Non-free
      • This refers to f-droid's own repository version. The FSD does not have a repository or host downloads, so it does not apply.
  • F-droid categories without a consensus, still need discussion:
    • Non-Free Assets
      • Free_Software_Directory:Requirements says "The software program itself should not package any program-data, art assets loaded by the program, or software which is under a nonfree license. If art or data is available for the game under a nonfree license but not packaged directly with it, that is a different matter and one we should be more flexible about." So I think we might create this tag for software which promotes using nonfree art or program data but does not package it. --IanK (talk)

Bait and surrender, MySQL examples

Non-free software profit models

--Adfeno (talk) 16:32, 15 May 2016 (EDT)

I think this is an important criteria to consider because when these "premium"/"commercial" editions are better than the "open source"/"community" editions, or when they aren't free/libre, they still allow the proprietors to exercise unjust/unfair power over society.
We must also investigate ProcessMaker, because this page's HTML source evidences the existence of different editions.
There might be other entries to be investigated.
Besides, if you don't mind, I would like to apologize for making the Odoo entry. I created it because someone from Brazil blindly told me that it was free/libre software, and so I had to create the entry to see if that was true, then I saw Odoo being approved, and since the approval, I have recommended it at least two times (always referencing to the approved entry).
--Adfeno (talk) 10:27, 16 May 2016 (EDT)
Just for the record, for those wanting to change the requirements according to what is taken as consideration with this proposal: I spoke with contributors of the directory and some organizers of the meetings (Jgay participated on the debate), and we consider that these crippleware are somehow useful free/libre software (specially because a free software activist can make a parent project out of it), and so we came up with two alternatives that can be used instead of removing the entries. At least one of the following is recommended:
  • Insert a message telling the viewer that the entry is about the free/libre edition of the software, and not about the non-free one, and that the free software movement doesn't recommend the non-free one. I, personally, suggest that this message is displayed with emphasis (like red boxes, with warning symbol and so on), but it's OK to have a simple text message (similar but longer than what the entry about Krita has).
  • Instead of linking to the projects official page, resources, documentation, source code and so on: Link to a page from a free system distribution that provides such software. This avoids attracting people to the non-free functional data.
--Adfeno (talk) 09:23, 18 June 2016 (EDT)

Nonfree documentation

Rationale for specific project inclusion

Since these antifeatures are apparently simple flags to associate with the various projects, people can (and perhaps should) question why a project has been flagged. Is there no place organized to detail the flagging rationale? In the meantime, I will include the rationale for flagging some projects as having nonfree documentation here, if anyone is interested:

Anki Both and are CloudFlared without Tor whitelisting. The site is where the user manual appears, and is a place where users share databases that they have produced. The knowledgebase lives on a free website (, but so many answers there link back to that it's also unusable to Tor users anyway. Not that it matters, the user manual is a quite essential piece to have free. Anki maintainers are aware of the situation so the freedom repression is deliberate. [Discussion]

GNU Radio The website is cloudflared without a Tor whitelist. Note that the gnuradio maintainers are aware and in fact quite hostile toward Tor users, so the attack on user freedom is clearly deliberate. (BTW, it would be possible for other projects to naïvely repress user freedom by using CloudFlare without fully understanding the ramifications -- and hopefully the antifeature flag will remedy such cases)

Tcl There is an index of Tcl documentation on, which is itself free, but the most popular and official sources for documentation is the Tcl wiki (, which is cloudflared without Tor whitelisting.

Wireshark The website is cloudflared without a Tor whitelist.

CloudFlare analysis

The above-mentioned projects for which the documentation is behind a CloudFlare-freedom-paywall (all at the moment) can be further traced to FSF principles outlined in [GNU Free Documentation License]. Specifically, "a free program should come with manuals providing the same freedoms that the software does." Instead of including manuals with the software, the 4 projects above jail the documentation on a webpage within a corporate walled-garden that threatens the freedoms of visitors. That loss of freedom is enabled by SaaSS-like issues, and described in detail in [Threats on documentation freedom need a SaaSS and free s/w composite remedy]. Cypherpunks (talk) 13:17, 19 January 2017 (EST)

Flag rendering

The flags for the above-mentioned projects are not rendering, despite my attempt at such. Is my syntax wrong?

Cypherpunks (talk) 11:28, 12 January 2017 (EST)

I updated these entries by modifying "Closed documentation" to "Nonfree documentation" in their source. Just update Free_Software_Directory:Antifeatures to see all "Nonfree documentation" entries. --David Hedlund (talk) 11:08, 16 January 2017 (EST)

Mind the unintended consequences

Sometimes an anti-feature tag can pose as a bad thing to already-approved entries, or even have people understanding the tags as though they would mean intentions different from those initially thought about.

For example, the person placing the tag can have a supportive intention ("will warn people so that they can join us and contribute to fix it"), but the person seeing it can think of it as a bad thing (including the community behind the package/software).

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.3 or any later version published by the Free Software Foundation; with no Invariant Sections, no Front-Cover Texts, and no Back-Cover Texts. A copy of the license is included in the page “GNU Free Documentation License”.

The copyright and license notices on this page only apply to the text on this page. Any software or copyright-licenses or other similar notices described in this text has its own copyright notice and license, which can usually be found in the distribution or license text itself.