27 ms·

> Wikipedia is useful for looking up technical info that there is no conflict over. Even this isn't true. As a former contributor to the Linux desktop communit

> Wikipedia is useful for looking up technical info that there is no conflict over. Even this isn't true. As a former contributor to the Linux desktop community, false information about projects and codebases I had runs rampant on Wikipedia. One Wikipedia editor with more time on his hands than any of us [0] devotes his time to maintaining those articles, despite self-admittedly having little topical knowledge [1]. Trying to correct these things quickly caused me to spiral down into edit wars I wasn't that interested in winning, and tomes of naming policy after policy. To get something incorrect on Wikipedia, you just have to be persistent and cast your WP:NPOV, WP:NOR, WP:DEMOCRACY, etc. spells at the right time. Anybody interested in correcting the record will eventually tire out and have to go back to doing the work... [0] https://commons.wikimedia.org/wiki/User:ScotXW https://commons.wikimedia.org/wiki/User:ScotXW [1] https://lists.freedesktop.org/archives/systemd-devel/2013-November/013970.html https://lists.freedesktop.org/archives/systemd-devel/2013-No...

tptacek 8y ago
If it's "rampant" on Wikipedia, you should have no trouble pointing out examples. So, respectfully, I'd ask that to support your argument with evidence, you do that. I had trouble editing technical topics I was familiar with, too, and found it incredibly aggravating. I'd write demonstrably true, helpful, factual content about something only to have it struck for lack of sourcing. It took awhile to get through my head the model of the project, which included as a basic premise that I, as an editor, am not a valid source. I came to the project knowing that, of course, but didn't realize viscerally how much of the technical writing I do implicitly relied on my own experience and understanding as a sort of "source". On articles that WP'ians pay attention to, you won't get away with doing that. You'll have to tie claims to a secondary source (you usually can't just point to source code, because then you're doing interpretation, which is a form of original research) both for the claim itself and, sometimes, for the noteworthiness of the claim. The best way to metabolize this is to realize that Wikipedia isn't the Hitchhiker's Guide; it's an encyclopedia, which means the project has explicitly opted-in to some limitations as a tradeoff to accomplish other things. (Some articles also just have possessive editors; there is dysfunction everywhere, and WP is no exception).
- Jasper_ 8y ago
  I don't remember the exact claims I tried to fight, but let's pick a random diagram from the page I linked: https://upload.wikimedia.org/wikipedia/commons/4/43/Linux_API.svg https://upload.wikimedia.org/wikipedia/commons/4/43/Linux_AP... See if you can spot any inaccuracies. Here are a few: 1. SDL doesn't use libevdev. It uses the kernel ioctls directly: http://hg.libsdl.org/SDL/file/3a50eb90e4b2/src/core/linux/SDL_evdev.c http://hg.libsdl.org/SDL/file/3a50eb90e4b2/src/core/linux/SD... 2. Sockets, netfilter, network protocols are all listed under "everything is a file", but notably, sockets are one of the few places of the UNIX API that isn't a file namespace! Sockets can be files in the case of UNIX domain sockets, but configuring Netfilter goes through NETLINK: http://man7.org/linux/man-pages/man7/netlink.7.html http://man7.org/linux/man-pages/man7/netlink.7.html 3. Similarly, the I/O scheduler isn't a file. Admittedly, I don't have a source for this other than the kernel source code itself. These are nitpicks, yes, but they add up, especially considering the near-100 diagrams this guy has made and cloned all over Wikipedia. When we saw people joining our community and asking questions, this information made them think things were different than they were, and that was frustrating. And yes, I used sources like these when trying to correct this misinformation.
  - tptacek 8y ago
    You haven't linked to an article, so I don't have the context here. Where is this diagram included? But I suspect you're running into exactly what I ran into: you're trying to correct a page using yourself as a source. I'm sure you're right (although I think you're nitpicking about the sockets-are-files thing --- not that they are, but rather than I don't think the diagram is making that claim --- it is a _really_ bad diagram though). It can be easier on WP to strike content from a page (because it is inaccurate) than it is to replace that content.
    - MattHeard 8y ago
      https://commons.wikimedia.org/wiki/File:Linux_API.svg https://commons.wikimedia.org/wiki/File:Linux_API.svg
AndyNemmity 8y ago
The same was true for technical articles on databases. Just factually incorrect information, and lots of marketing. Tried to fix it, but there were a few people more devoted than I was, and it became too much of an effort to play the revert/lawyer game just to get consistent articles with similar information across all databases.
JdeBP 8y ago
Another example of problematic diagrams in the computing field is the one for systemd. It took me citing one of the systemd developers to point out that it was not in fact an accurate diagram. * https://en.wikipedia.org/w/index.php?title=Systemd&diff=631112520&oldid=631031336 https://en.wikipedia.org/w/index.php?title=Systemd&diff=6311... I observe that this was again that account that you mention, in 2013 taking someone else's work off another WWW site without attributing the creator (or indeed any indication on that other WWW site, that has no free copyright licence that I can find, that the original diagram was freely licensed by that creator and usable on Wikipedia in the first place), taking it out of its original context, and falsely marking as "own work". That account maintains a personal version of the article still calling it "The entire systemd software bundle and its components". * https://en.wikipedia.org/wiki/User:ScotXW/systemd#Components_of_the_systemd_software_bundle https://en.wikipedia.org/wiki/User:ScotXW/systemd#Components...

tptacek 8y ago

If it's "rampant" on Wikipedia, you should have no trouble pointing out examples. So, respectfully, I'd ask that to support your argument with evidence, you do that. I had trouble editing technical topics I was familiar with, too, and found it incredibly aggravating. I'd write demonstrably true, helpful, factual content about something only to have it struck for lack of sourcing. It took awhile to get through my head the model of the project, which included as a basic premise that I, as an editor, am not a valid source. I came to the project knowing that, of course, but didn't realize viscerally how much of the technical writing I do implicitly relied on my own experience and understanding as a sort of "source". On articles that WP'ians pay attention to, you won't get away with doing that. You'll have to tie claims to a secondary source (you usually can't just point to source code, because then you're doing interpretation, which is a form of original research) both for the claim itself and, sometimes, for the noteworthiness of the claim. The best way to metabolize this is to realize that Wikipedia isn't the Hitchhiker's Guide; it's an encyclopedia, which means the project has explicitly opted-in to some limitations as a tradeoff to accomplish other things. (Some articles also just have possessive editors; there is dysfunction everywhere, and WP is no exception).

Jasper_ 8y ago
I don't remember the exact claims I tried to fight, but let's pick a random diagram from the page I linked: https://upload.wikimedia.org/wikipedia/commons/4/43/Linux_API.svg https://upload.wikimedia.org/wikipedia/commons/4/43/Linux_AP... See if you can spot any inaccuracies. Here are a few: 1. SDL doesn't use libevdev. It uses the kernel ioctls directly: http://hg.libsdl.org/SDL/file/3a50eb90e4b2/src/core/linux/SDL_evdev.c http://hg.libsdl.org/SDL/file/3a50eb90e4b2/src/core/linux/SD... 2. Sockets, netfilter, network protocols are all listed under "everything is a file", but notably, sockets are one of the few places of the UNIX API that isn't a file namespace! Sockets can be files in the case of UNIX domain sockets, but configuring Netfilter goes through NETLINK: http://man7.org/linux/man-pages/man7/netlink.7.html http://man7.org/linux/man-pages/man7/netlink.7.html 3. Similarly, the I/O scheduler isn't a file. Admittedly, I don't have a source for this other than the kernel source code itself. These are nitpicks, yes, but they add up, especially considering the near-100 diagrams this guy has made and cloned all over Wikipedia. When we saw people joining our community and asking questions, this information made them think things were different than they were, and that was frustrating. And yes, I used sources like these when trying to correct this misinformation.
- tptacek 8y ago
  You haven't linked to an article, so I don't have the context here. Where is this diagram included? But I suspect you're running into exactly what I ran into: you're trying to correct a page using yourself as a source. I'm sure you're right (although I think you're nitpicking about the sockets-are-files thing --- not that they are, but rather than I don't think the diagram is making that claim --- it is a _really_ bad diagram though). It can be easier on WP to strike content from a page (because it is inaccurate) than it is to replace that content.
  - MattHeard 8y ago
    https://commons.wikimedia.org/wiki/File:Linux_API.svg https://commons.wikimedia.org/wiki/File:Linux_API.svg

AndyNemmity 8y ago

The same was true for technical articles on databases. Just factually incorrect information, and lots of marketing. Tried to fix it, but there were a few people more devoted than I was, and it became too much of an effort to play the revert/lawyer game just to get consistent articles with similar information across all databases.

JdeBP 8y ago

Another example of problematic diagrams in the computing field is the one for systemd. It took me citing one of the systemd developers to point out that it was not in fact an accurate diagram. * https://en.wikipedia.org/w/index.php?title=Systemd&diff=631112520&oldid=631031336 https://en.wikipedia.org/w/index.php?title=Systemd&diff=6311... I observe that this was again that account that you mention, in 2013 taking someone else's work off another WWW site without attributing the creator (or indeed any indication on that other WWW site, that has no free copyright licence that I can find, that the original diagram was freely licensed by that creator and usable on Wikipedia in the first place), taking it out of its original context, and falsely marking as "own work". That account maintains a personal version of the article still calling it "The entire systemd software bundle and its components". * https://en.wikipedia.org/wiki/User:ScotXW/systemd#Components_of_the_systemd_software_bundle https://en.wikipedia.org/wiki/User:ScotXW/systemd#Components...