5 ms·

GitHub giveth; Wikipedia taketh away

35 points by fern12 9y ago

sshine 9y ago
Interesting points. Most popular alternatives to Wikipedia, commercial and non-commercial, work with ownership of articles, or a stricter editorial policy. [1] This includes Scholarpedia, Citizendium, Conservapedia. It comes at the cost of volume and recency. In a sense, Wikipedia deals with all the advantages and disadvantages of modern, free democracy. An aspect that the article seems to miss is the different demographics of Wikipedia and GitHub. (There are a lot fewer high-schoolers vandalizing GitHub repositories out of boredom.) The drawback of GitHub is not just that pull-requests must be accepted by authorized users, but also the bureaucracy of learning version control. Compared to Wikipedia, anyone with domain knowledge can hit "Edit" and add their bit. I would rather like to see a synthesis of StackOverflow's point system, so the quality of your contributions grants you privileges for certain sections of the site, applied to wiki articles. Another difference for comparison between GitHub/StackOverflow and Wikipedia is that you have ownership of your contributions. I occasionally go back and revise the most popular StackOverflow answers I've made for posterity. It's been years since I last made a change to one of my Wikipedia contributions. The last I remember was that someone had restructured a part of an article I'd written from scratch and added wrong information. [1]: http://oedb.org/ilibrarian/top-7-alternatives-to-wikipedia/ http://oedb.org/ilibrarian/top-7-alternatives-to-wikipedia/
- gonzo41 9y ago
  Why don't you just open an empty repo, and start accepting pull requests. See what happens.
- fredley 9y ago
  Is editing Wikipedia easy? Although I'll agree it's a lot easier than pushing to GitHub, it's still way beyond the skills of the average Wikipedia reader. Click Edit on any article. It's a mess of opaque syntax, presented in a textbox. Above the textbox are two different warning paragraphs, one because I'm not logged in and another warning against copyright. Any casual user who clicked Edit out of a genuine desire to improve the article would probably just click away. I would say the ease of editing needs to be much better than GitHub or current Wikipedia. A decent, WYSIWYG editor (with a Markdown mode, don't panic) with tools to make adding citations etc. possible for mortal users.
  - DanBC 9y ago
    > Click Edit on any article. It's a mess of opaque syntax, presented in a textbox. They've had visual editing for years: https://imgur.com/a/mnePe https://imgur.com/a/mnePe
    - fredley 9y ago
      But that's not what's presented by default. I just tried again and got a popup asking me to switch (not sure why that didn't show up first time around), but the default button for the popup ('Start Editing') takes you to the old interface.
      - sshine 9y ago
        Also, getting the visual editor back once you say no thanks doesn't seem trivial. I had to look up how. My guess is that regular contributors hardly use it, and its purpose is mostly symbolic rather than practical.
        fredley 9y ago
        How I would love to see the data on how many edits submitted from the WYSIWYG tool are/aren't reverted.
        tgr_ 9y ago
        Slightly less than with the text editor: https://meta.wikimedia.org/wiki/Research:VisualEditor%27s_effect_on_newly_registered_editors/November_2015_follow-up https://meta.wikimedia.org/wiki/Research:VisualEditor%27s_ef...
      - DanBC 9y ago
        I just tried in a private window, and got this: https://imgur.com/a/isTAw https://imgur.com/a/isTAw I guess this just proves OP's point: Wikipedia has been struggling to introduce visual editing because the current editing userbase rejected it.
  - 9y ago
- Aissen 9y ago
  Github has an Edit button, too. It will automatically fork the repo, make a commit, and then create a pull-request. Very handy.
- reificator 9y ago
  > The drawback of GitHub is not just that pull-requests must be accepted by authorized users, but also the bureaucracy of learning version control. Compared to Wikipedia, anyone with domain knowledge can hit "Edit" and add their bit. Editing from the site is incredibly simple. Browse to a file, click `edit` and make your changes. Then at the bottom of the page, give a commit message and click "Propose file change".
  - Singletoned 9y ago
    Indeed. I've had non-technical content writers feeling comfortable adding files directly to GitHub and writing Markdown in the them. Basically using it as a CMS. Pull Requests and Branching remained confusing issues for them though.
- cabalamat 9y ago
  > Most popular alternatives to Wikipedia, commercial and non-commercial, work with ownership of articles, or a stricter editorial policy. I'm currently doing an inclusionist fork of Wikipedia -- http://www.wikifork.org/mw/index.php/Main_Page http://www.wikifork.org/mw/index.php/Main_Page . Wikifork (when I get it up and running) will have much looser policies on notability than Wikipedia has.
mosselman 9y ago
"...annoyance at the legalistic (and typically white male) Wikipedian." How are skin color and gender relevant to the article or the point being made?
- purrcat259 9y ago
  It read like low hanging identity politics fruit. I stopped reading after that statement.
- nightcracker 9y ago
  We've entered a new age of racism and sexism, where as long as whatever group you're targeting is perceived to be 'privileged', it's okay to be hateful and divisive. IMO there's a simple rule of thumb. If you can't replace 'white male' with some other group without the statement becoming incredibly offensive to you, you probably shouldn't say it. E.g., I sincerely doubt the author would be okay with saying "(and typically female)" or "(and typically black)" in nearly any context, so I fail to see why he/she feels this is okay.
- dang 9y ago
  Would you please not pick the most inflammatory and off-topic bit of an article and drag it in here to start a flamewar with? Threads are highly sensitive to initial conditions, so you can do a lot of damage this way. This is covered by the following site guideline: Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
  - mosselman 9y ago
    Firstly, I am fine with you collapsing my comment. The following is just an addition to why I think it is somewhat relevant: The author claims that there are social constructs in place at wikipedia that lead to a bad working environment for contributors. There is not evidence provided to support this, but there is this offhand sexist and racist comment that gives insight into the way in which this author processes information around him/her. I find this a relevant bit of information for evaluating the merit of the article. What if you have an article with the title "SpaceX doomed to fail" "SpaceX could never succeed, their goals are unrealistic and they haven't done enough research.... The earth (which is flat by the way)..." Attacking the author on his flat earth believes would be a no go according to that policy, but it does provide information about the value of the author as a credible source.
flixic 9y ago
Wikipedia is like blockchain; sure you can fork it but the value is in the fact that there is one Wikipedia. That makes a lot of the problems worse.
- cabalamat 9y ago
  > but the value is in the fact that there is one Wikipedia Why?
  - flixic 9y ago
    Single source of truth. It would be too easy to split into "alternative facts" if people could successfully establish alternative Wikipedias. Of course, seeing the direction the world is going, this may happen. Also, while seeing the benefit of single Wikipedia, I also think it could benefit from larger number of pages about less important topics, not being that strict on what can and can't be a page topic. Extended Wikipedia of sorts.
    - avian 9y ago
      > Single source of truth https://en.wikipedia.org/wiki/Wikipedia:Verifiability,_not_truth https://en.wikipedia.org/wiki/Wikipedia:Verifiability,_not_t...
    - cabalamat 9y ago
      > It would be too easy to split into "alternative facts" if people could successfully establish alternative Wikipedias. Bear in mind that Wikipedia isn't the only website out there! But you make an interesting point about how authoritative it is perceived. > I also think it could benefit from larger number of pages about less important topics, not being that strict on what can and can't be a page topic. Extended Wikipedia of sorts. I agree, that's why I'm doing WikiFork.
z3t4 9y ago
Github is immutable, while Wikipedia is mutable with r/w conflicts and corruption.
avian 9y ago
I wonder how many times the author spent time making a nice GitHub pull request for a popular repo. You know, taking care that you match the coding style, document your changes and reasons for them, make sure the CI tests pass, etc. only to see it sit in the tracker for months without anyone merging it or commenting. Then after a year or two, when you already forgot about it, you get a mail that someone closed your ticket for being outdated or not merging cleanly (to the master that had a year worth of work done on it since you last looked) GitHub can be a source of bitter disappointment and annoyance just like Wikipedia.
- MaxBarraclough 9y ago
  It's disheartening enough to see your well-crafted bug-report go unanswered.
- gluejar 9y ago
  This point now referenced in the article.
hh3k0 9y ago
> […] the legalistic (and typically white male) Wikipedian. Yawn.
- cabalamat 9y ago
  Male is relevant, since men are more drawn to being Wikipedia editors than women. it has been argued that the deletionist policies of the people running the wikipedia project put women off to a greater degree than they put men off; I've no idea whether that is true. White isn't relevant -- it's mainly just because most native English speakers are white. I would imagine that for the Japanese Wikipedia, most editors are Japanese.
  - gluejar 9y ago
    Perhaps the distribution of fluent english speakers is more relevant.
- dang 9y ago
  Would you please not post unsubstantive comments, or flamebait? More here: https://news.ycombinator.com/item?id=16185156 https://news.ycombinator.com/item?id=16185156
pvdebbe 9y ago
> most trusted I don't personally know anyone who actually trusts Wikipedia. Everyone uses it for sheer laziness, as they admit.
_joe 9y ago
When you people will be able to scale git to accept the number of revisions Wikipedia has, and add several new ones per second with millisecond latencies, please let me know. Also, only a programmer could think git is a friendlier interface than a wiki. There are things that are wrong in the wiki comunity, none of which would be solved by using github instead. Not to point out that the forking model would lead to fragmentation and GitHub is actually a privately owned for-profit platform, of course.
bArray 9y ago
I agree with @Vorticalbox in the comments: >(and typically white male) that is an unnecessary comment. Sex or gender, >even if they are the majority, has very little to do with the way wiki works. > >Even the minority groups will act the same way. I would go further and say that it probably depends on what article you are editing as to what gender/ethnicity causes "disappointment and annoyance". I have no idea why it is suddenly popular (and not disputed) to make remarks like this in otherwise good articles. That said, I think some of the ideas in this blog are not correct. One of Wikipedia's benefits is that it is consolidated into one effort, rather than many forks and a major distributed effort. Also as a number of comments have suggested, PR's can sit for a while before being rejected depending on how active a repository is. Other than having to use wiki markup (which in my opinion is confusing and terrible), there's a reasonably low entry barrier for contributers to jump over. Expecting people to use a form of Git (even with a UI) could end up being a lot of hassle.
kawsper 9y ago
> But I imagine that new and old contributors get some satisfaction when their contribution get "merged into master", no matter how much that sounds like yielding to the hierarchy. I spent one day in my weekend upgrading a couple of Rubygems to be compatible with Ruby 2.5 after some things were deprecated. One of the authors just closed my PR, left no comments, and committed his own changes to master. His solution was a little bit different than my own, but he was able to commit to my branch. A "Thank you", or a "Hey! I did not know that it was deprecated in 2.5, thanks!" would have been appreciated, and I felt a bit miffed even though it is not a big thing.
gluejar 9y ago
Thanks for the comments, folks. Apparently this post is too controversial or off-topic (or maybe just deemed silly) for Hacker News - it's been removed from the ranking!

35 points by fern12 9y ago

Interesting points. Most popular alternatives to Wikipedia, commercial and non-commercial, work with ownership of articles, or a stricter editorial policy. [1] This includes Scholarpedia, Citizendium, Conservapedia. It comes at the cost of volume and recency. In a sense, Wikipedia deals with all the advantages and disadvantages of modern, free democracy. An aspect that the article seems to miss is the different demographics of Wikipedia and GitHub. (There are a lot fewer high-schoolers vandalizing GitHub repositories out of boredom.) The drawback of GitHub is not just that pull-requests must be accepted by authorized users, but also the bureaucracy of learning version control. Compared to Wikipedia, anyone with domain knowledge can hit "Edit" and add their bit. I would rather like to see a synthesis of StackOverflow's point system, so the quality of your contributions grants you privileges for certain sections of the site, applied to wiki articles. Another difference for comparison between GitHub/StackOverflow and Wikipedia is that you have ownership of your contributions. I occasionally go back and revise the most popular StackOverflow answers I've made for posterity. It's been years since I last made a change to one of my Wikipedia contributions. The last I remember was that someone had restructured a part of an article I'd written from scratch and added wrong information. [1]: http://oedb.org/ilibrarian/top-7-alternatives-to-wikipedia/ http://oedb.org/ilibrarian/top-7-alternatives-to-wikipedia/

gonzo41 9y ago
Why don't you just open an empty repo, and start accepting pull requests. See what happens.
fredley 9y ago
Is editing Wikipedia easy? Although I'll agree it's a lot easier than pushing to GitHub, it's still way beyond the skills of the average Wikipedia reader. Click Edit on any article. It's a mess of opaque syntax, presented in a textbox. Above the textbox are two different warning paragraphs, one because I'm not logged in and another warning against copyright. Any casual user who clicked Edit out of a genuine desire to improve the article would probably just click away. I would say the ease of editing needs to be much better than GitHub or current Wikipedia. A decent, WYSIWYG editor (with a Markdown mode, don't panic) with tools to make adding citations etc. possible for mortal users.
- DanBC 9y ago
  > Click Edit on any article. It's a mess of opaque syntax, presented in a textbox. They've had visual editing for years: https://imgur.com/a/mnePe https://imgur.com/a/mnePe
  - fredley 9y ago
    But that's not what's presented by default. I just tried again and got a popup asking me to switch (not sure why that didn't show up first time around), but the default button for the popup ('Start Editing') takes you to the old interface.
    - sshine 9y ago
      Also, getting the visual editor back once you say no thanks doesn't seem trivial. I had to look up how. My guess is that regular contributors hardly use it, and its purpose is mostly symbolic rather than practical.
      - fredley 9y ago
        How I would love to see the data on how many edits submitted from the WYSIWYG tool are/aren't reverted.
        tgr_ 9y ago
        Slightly less than with the text editor: https://meta.wikimedia.org/wiki/Research:VisualEditor%27s_effect_on_newly_registered_editors/November_2015_follow-up https://meta.wikimedia.org/wiki/Research:VisualEditor%27s_ef...
    - DanBC 9y ago
      I just tried in a private window, and got this: https://imgur.com/a/isTAw https://imgur.com/a/isTAw I guess this just proves OP's point: Wikipedia has been struggling to introduce visual editing because the current editing userbase rejected it.
- 9y ago
Aissen 9y ago
Github has an Edit button, too. It will automatically fork the repo, make a commit, and then create a pull-request. Very handy.
reificator 9y ago
> The drawback of GitHub is not just that pull-requests must be accepted by authorized users, but also the bureaucracy of learning version control. Compared to Wikipedia, anyone with domain knowledge can hit "Edit" and add their bit. Editing from the site is incredibly simple. Browse to a file, click `edit` and make your changes. Then at the bottom of the page, give a commit message and click "Propose file change".
- Singletoned 9y ago
  Indeed. I've had non-technical content writers feeling comfortable adding files directly to GitHub and writing Markdown in the them. Basically using it as a CMS. Pull Requests and Branching remained confusing issues for them though.
cabalamat 9y ago
> Most popular alternatives to Wikipedia, commercial and non-commercial, work with ownership of articles, or a stricter editorial policy. I'm currently doing an inclusionist fork of Wikipedia -- http://www.wikifork.org/mw/index.php/Main_Page http://www.wikifork.org/mw/index.php/Main_Page . Wikifork (when I get it up and running) will have much looser policies on notability than Wikipedia has.

mosselman 9y ago

"...annoyance at the legalistic (and typically white male) Wikipedian." How are skin color and gender relevant to the article or the point being made?

purrcat259 9y ago
It read like low hanging identity politics fruit. I stopped reading after that statement.
nightcracker 9y ago
We've entered a new age of racism and sexism, where as long as whatever group you're targeting is perceived to be 'privileged', it's okay to be hateful and divisive. IMO there's a simple rule of thumb. If you can't replace 'white male' with some other group without the statement becoming incredibly offensive to you, you probably shouldn't say it. E.g., I sincerely doubt the author would be okay with saying "(and typically female)" or "(and typically black)" in nearly any context, so I fail to see why he/she feels this is okay.
dang 9y ago
Would you please not pick the most inflammatory and off-topic bit of an article and drag it in here to start a flamewar with? Threads are highly sensitive to initial conditions, so you can do a lot of damage this way. This is covered by the following site guideline: Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html
- mosselman 9y ago
  Firstly, I am fine with you collapsing my comment. The following is just an addition to why I think it is somewhat relevant: The author claims that there are social constructs in place at wikipedia that lead to a bad working environment for contributors. There is not evidence provided to support this, but there is this offhand sexist and racist comment that gives insight into the way in which this author processes information around him/her. I find this a relevant bit of information for evaluating the merit of the article. What if you have an article with the title "SpaceX doomed to fail" "SpaceX could never succeed, their goals are unrealistic and they haven't done enough research.... The earth (which is flat by the way)..." Attacking the author on his flat earth believes would be a no go according to that policy, but it does provide information about the value of the author as a credible source.

flixic 9y ago

Wikipedia is like blockchain; sure you can fork it but the value is in the fact that there is one Wikipedia. That makes a lot of the problems worse.

cabalamat 9y ago
> but the value is in the fact that there is one Wikipedia Why?
- flixic 9y ago
  Single source of truth. It would be too easy to split into "alternative facts" if people could successfully establish alternative Wikipedias. Of course, seeing the direction the world is going, this may happen. Also, while seeing the benefit of single Wikipedia, I also think it could benefit from larger number of pages about less important topics, not being that strict on what can and can't be a page topic. Extended Wikipedia of sorts.
  - avian 9y ago
    > Single source of truth https://en.wikipedia.org/wiki/Wikipedia:Verifiability,_not_truth https://en.wikipedia.org/wiki/Wikipedia:Verifiability,_not_t...
  - cabalamat 9y ago
    > It would be too easy to split into "alternative facts" if people could successfully establish alternative Wikipedias. Bear in mind that Wikipedia isn't the only website out there! But you make an interesting point about how authoritative it is perceived. > I also think it could benefit from larger number of pages about less important topics, not being that strict on what can and can't be a page topic. Extended Wikipedia of sorts. I agree, that's why I'm doing WikiFork.

z3t4 9y ago

Github is immutable, while Wikipedia is mutable with r/w conflicts and corruption.

avian 9y ago

I wonder how many times the author spent time making a nice GitHub pull request for a popular repo. You know, taking care that you match the coding style, document your changes and reasons for them, make sure the CI tests pass, etc. only to see it sit in the tracker for months without anyone merging it or commenting. Then after a year or two, when you already forgot about it, you get a mail that someone closed your ticket for being outdated or not merging cleanly (to the master that had a year worth of work done on it since you last looked) GitHub can be a source of bitter disappointment and annoyance just like Wikipedia.

MaxBarraclough 9y ago
It's disheartening enough to see your well-crafted bug-report go unanswered.
gluejar 9y ago
This point now referenced in the article.

hh3k0 9y ago

> […] the legalistic (and typically white male) Wikipedian. Yawn.

cabalamat 9y ago
Male is relevant, since men are more drawn to being Wikipedia editors than women. it has been argued that the deletionist policies of the people running the wikipedia project put women off to a greater degree than they put men off; I've no idea whether that is true. White isn't relevant -- it's mainly just because most native English speakers are white. I would imagine that for the Japanese Wikipedia, most editors are Japanese.
- gluejar 9y ago
  Perhaps the distribution of fluent english speakers is more relevant.
dang 9y ago
Would you please not post unsubstantive comments, or flamebait? More here: https://news.ycombinator.com/item?id=16185156 https://news.ycombinator.com/item?id=16185156

pvdebbe 9y ago

> most trusted I don't personally know anyone who actually trusts Wikipedia. Everyone uses it for sheer laziness, as they admit.

_joe 9y ago

When you people will be able to scale git to accept the number of revisions Wikipedia has, and add several new ones per second with millisecond latencies, please let me know. Also, only a programmer could think git is a friendlier interface than a wiki. There are things that are wrong in the wiki comunity, none of which would be solved by using github instead. Not to point out that the forking model would lead to fragmentation and GitHub is actually a privately owned for-profit platform, of course.

bArray 9y ago

I agree with @Vorticalbox in the comments: >(and typically white male) that is an unnecessary comment. Sex or gender, >even if they are the majority, has very little to do with the way wiki works. > >Even the minority groups will act the same way. I would go further and say that it probably depends on what article you are editing as to what gender/ethnicity causes "disappointment and annoyance". I have no idea why it is suddenly popular (and not disputed) to make remarks like this in otherwise good articles. That said, I think some of the ideas in this blog are not correct. One of Wikipedia's benefits is that it is consolidated into one effort, rather than many forks and a major distributed effort. Also as a number of comments have suggested, PR's can sit for a while before being rejected depending on how active a repository is. Other than having to use wiki markup (which in my opinion is confusing and terrible), there's a reasonably low entry barrier for contributers to jump over. Expecting people to use a form of Git (even with a UI) could end up being a lot of hassle.

kawsper 9y ago

> But I imagine that new and old contributors get some satisfaction when their contribution get "merged into master", no matter how much that sounds like yielding to the hierarchy. I spent one day in my weekend upgrading a couple of Rubygems to be compatible with Ruby 2.5 after some things were deprecated. One of the authors just closed my PR, left no comments, and committed his own changes to master. His solution was a little bit different than my own, but he was able to commit to my branch. A "Thank you", or a "Hey! I did not know that it was deprecated in 2.5, thanks!" would have been appreciated, and I felt a bit miffed even though it is not a big thing.

gluejar 9y ago

Thanks for the comments, folks. Apparently this post is too controversial or off-topic (or maybe just deemed silly) for Hacker News - it's been removed from the ranking!