www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Voting for forum posts

reply Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> writes:
I wonder how difficult it would be to add a voting system for forum 
messages. Votes wouldn't be available in the NNTP interface but would be 
accessible in the Web interface and would allow sorting and filtering by 
votes.

cc Vladimir :o)
Mar 31 2021
next sibling parent reply Adam D. Ruppe <destructionator gmail.com> writes:
On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei Alexandrescu 
wrote:
 I wonder how difficult it would be to add a voting system for 
 forum messages. Votes wouldn't be available in the NNTP 
 interface but would be accessible in the Web interface and 
 would allow sorting and filtering by votes.
I'm sure this wouldn't be terribly difficult to implement, but I also don't think it is a good idea. Online voting is emotionally toxic and logically fallacious. We're better off without it.
Mar 31 2021
next sibling parent reply tsbockman <thomas.bockman gmail.com> writes:
On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe wrote:
 On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei 
 Alexandrescu wrote:
 I wonder how difficult it would be to add a voting system for 
 forum messages. Votes wouldn't be available in the NNTP 
 interface but would be accessible in the Web interface and 
 would allow sorting and filtering by votes.
I'm sure this wouldn't be terribly difficult to implement, but I also don't think it is a good idea. Online voting is emotionally toxic and logically fallacious. We're better off without it.
I agree Filtering or sorting by votes usually creates echo chambers. It is a pretty effective way to kill any possibility of fruitful debate: anyone who doesn't *sound* right to the majority of voters at the beginning quickly becomes invisible, largely eliminating the possibility of ever changing the majority's mind about anything. Even without filtering and sorting, voting turns discussion into a popularity contest, attracts politicians, and creates new ones. In general, it's only really good for things that aren't controversial. All that said, many websites make it worse than it needs to be with bad systems. (Bad for users, that is - maybe not bad for ruthlessly manipulating public opinion.) Up votes and down votes should always be tracked and displayed separately. Combined net counts are stupid and misleading: 0 upvotes and 100 downvotes means something very different from 10_000 upvotes and 10_100 downvotes.
Mar 31 2021
next sibling parent jmh530 <john.michael.hall gmail.com> writes:
On Wednesday, 31 March 2021 at 22:06:38 UTC, tsbockman wrote:
 [snip]

 I agree

 Filtering or sorting by votes usually creates echo chambers. It 
 is a pretty effective way to kill any possibility of fruitful 
 debate: anyone who doesn't *sound* right to the majority of 
 voters at the beginning quickly becomes invisible, largely 
 eliminating the possibility of ever changing the majority's 
 mind about anything.
 [snip]
Obligatory https://www.evanmiller.org/how-not-to-sort-by-average-rating.html
Mar 31 2021
prev sibling parent reply RazvanN <razvan.nitu1305 gmail.com> writes:
On Wednesday, 31 March 2021 at 22:06:38 UTC, tsbockman wrote:
 On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe 
 wrote:
 On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei 
 Alexandrescu wrote:
 I wonder how difficult it would be to add a voting system for 
 forum messages. Votes wouldn't be available in the NNTP 
 interface but would be accessible in the Web interface and 
 would allow sorting and filtering by votes.
I'm sure this wouldn't be terribly difficult to implement, but I also don't think it is a good idea. Online voting is emotionally toxic and logically fallacious. We're better off without it.
I agree Filtering or sorting by votes usually creates echo chambers. It is a pretty effective way to kill any possibility of fruitful debate: anyone who doesn't *sound* right to the majority of voters at the beginning quickly becomes invisible, largely eliminating the possibility of ever changing the majority's mind about anything.
I don't think that Andrei wanted this implemented in a way in which downvoted comments become invisible, but rather the option of upvoting and downvoting a comment so that people who are interested in the most popular/unpopular opinions would have a way to search for this. Ideally, by default, the forum stays the same, however, if you want to search for the most upvoted comments (which would most likely represent the main points of the discussion) you should have a way to do that. There are numerous posts with 200+ replies that are concentrated around 3-4 key comments. It is really hard to identify them with the current interface, you really have to go through all of the posts (some of them quite large). A voting system would basically select the most important comments that are debated.
 Even without filtering and sorting, voting turns discussion 
 into a popularity contest, attracts politicians, and creates 
 new ones. In general, it's only really good for things that 
 aren't controversial.
I think that this thread represents the best argument why a voting system is needed. You just posted a comment and the next 3-4 comments were basically "+1". If we had a voting system, your comment would have been voted and the verbosity of this thread would have diminished.
 All that said, many websites make it worse than it needs to be 
 with bad systems. (Bad for users, that is - maybe not bad for 
 ruthlessly manipulating public opinion.) Up votes and down 
 votes should always be tracked and displayed separately. 
 Combined net counts are stupid and misleading: 0 upvotes and 
 100 downvotes means something very different from 10_000 
 upvotes and 10_100 downvotes.
What I get from this is that voting can be also good or bad depending on how we use it. IMHO having an upvote/downvote scheme could be beneficial in some situations if it is implemented properly. Cheers, RazvanN
Apr 01 2021
next sibling parent reply Timon Gehr <timon.gehr gmx.ch> writes:
On 01.04.21 10:05, RazvanN wrote:
 ...
 I think that this thread represents the best argument why a voting system
 is needed. You just posted a comment and the next 3-4 comments were 
 basically "+1".
Most of them with additional text. If we had a voting system, your comment would have been voted The post that got so much support _literally_ said that a voting system would be a bad idea, and why.
 and the verbosity of this thread would have diminished.
 ...
If that's a problem for web interface users, improve the web interface.
 All that said, many websites make it worse than it needs to be with 
 bad systems. (Bad for users, that is - maybe not bad for ruthlessly 
 manipulating public opinion.) Up votes and down votes should always be 
 tracked and displayed separately. Combined net counts are stupid and 
 misleading: 0 upvotes and 100 downvotes means something very different 
 from 10_000 upvotes and 10_100 downvotes.
What I get from this is that voting can be also good or bad depending on how we use it. IMHO having an upvote/downvote scheme could be beneficial in some situations if it is implemented properly. ...
Downvotes are worse than useless. Just upvote the rebuttal instead; at least it will be clear what you voted for. E.g., if I just downvoted your post, that would not have helped you understand why.
Apr 01 2021
next sibling parent reply user1234 <user1234 12.de> writes:
On Thursday, 1 April 2021 at 12:18:33 UTC, Timon Gehr wrote:
 that would not have helped you understand why.
Honestly people who get upvoted dont understand why either but they're just happy.
Apr 01 2021
parent Timon Gehr <timon.gehr gmx.ch> writes:
On 01.04.21 14:36, user1234 wrote:
 On Thursday, 1 April 2021 at 12:18:33 UTC, Timon Gehr wrote:
 that would not have helped you understand why.
Honestly people who get upvoted dont understand why either but they're just happy.
Fair point, though I think it's a bit better there because an upvote usually expresses agreement with all the points in the post, while a downvote may be due to disagreement with any of possibly multiple points in the post. And then there's also the possibility that a vote is based on some misunderstanding, which affects downvotes more than upvotes.
Apr 01 2021
prev sibling parent RazvanN <razvan.nitu1305 gmail.com> writes:
On Thursday, 1 April 2021 at 12:18:33 UTC, Timon Gehr wrote:
 On 01.04.21 10:05, RazvanN wrote:
 Most of them with additional text.

 If we had a voting system, your comment would have been voted

 The post that got so much support _literally_ said that a 
 voting system would be a bad idea, and why.
Yes, people that wanted to understand the flow of the conversation without going through all of the messages would have seen the initial comment which would have been voted by a lot of people and maybe my comment which may have been supported by people who agree with me. This way you read 2 comments and you understand the general debate. If you want to zoom in to the details, you can just select the normal, currently used, algorithm where votes don't count. I see it as an optional feature which you can use if you want to.
 If that's a problem for web interface users, improve the web 
 interface.
I don't see how the interface would solve the verbosity problem. If you have 200+ messages in a thread, how can the interface help you without having some filters?
 Downvotes are worse than useless. Just upvote the rebuttal 
 instead; at least it will be clear what you voted for. E.g., if 
 I just downvoted your post, that would not have helped you 
 understand why.
I am not fixated on a specific proposal, I was just stating that a voting scheme, however we choose to implement it may bring some benefit. I agree that downvoting isn't very useful, but upvoting could be used as a mean to find the comments that were mostly appreciated by people. If you don't care about that, you can just use the old interface where posts are presented in a chronological order.
Apr 01 2021
prev sibling parent tsbockman <thomas.bockman gmail.com> writes:
On Thursday, 1 April 2021 at 08:05:29 UTC, RazvanN wrote:
 What I get from this is that voting can be also good or bad 
 depending on how we use it.
The problem is that it's extremely difficult to establish and enforce a healthy voting culture. Many websites have "rules" like "Don't downvote just because you disagree," "Upvote content that contributes to the discussion," "No brigading," and "No ballot stuffing." But, the breaking of these rules still ends up dominating the results. Individual misuse of the voting system is much harder to detect than abuse of the right to comment, and so bad actors tend to stick around. Eventually they take over entirely, as voting begins to influence meta discussions like this one.
 IMHO having an upvote/downvote scheme could be beneficial
 in some situations if it is implemented properly.
I think forum voting systems are beneficial in some circumstances, but not specifically for the big technical debates that people are suggesting we use them for.
Apr 01 2021
prev sibling next sibling parent deadalnix <deadalnix gmail.com> writes:
On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe wrote:
 On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei 
 Alexandrescu wrote:
 I wonder how difficult it would be to add a voting system for 
 forum messages. Votes wouldn't be available in the NNTP 
 interface but would be accessible in the Web interface and 
 would allow sorting and filtering by votes.
I'm sure this wouldn't be terribly difficult to implement, but I also don't think it is a good idea. Online voting is emotionally toxic and logically fallacious. We're better off without it.
+1
Mar 31 2021
prev sibling next sibling parent "H. S. Teoh" <hsteoh quickfur.ath.cx> writes:
On Wed, Mar 31, 2021 at 09:40:15PM +0000, Adam D. Ruppe via Digitalmars-d wrote:
 On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei Alexandrescu wrote:
 I wonder how difficult it would be to add a voting system for forum
 messages. Votes wouldn't be available in the NNTP interface but
 would be accessible in the Web interface and would allow sorting and
 filtering by votes.
I'm sure this wouldn't be terribly difficult to implement, but I also don't think it is a good idea. Online voting is emotionally toxic and logically fallacious. We're better off without it.
+1. It will also lead to totally pointless and fallacious reactions like "this comment was rated highest, how come Walter refuses to implement what it said?". (And yes, I realize the irony of this post. :-D) T -- Never step over a puddle, always step around it. Chances are that whatever made it is still dripping.
Mar 31 2021
prev sibling next sibling parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 3/31/2021 2:40 PM, Adam D. Ruppe wrote:
 I'm sure this wouldn't be terribly difficult to implement, but I also don't 
 think it is a good idea. Online voting is emotionally toxic and logically 
 fallacious. We're better off without it.
Votes on Reddit and Hackernews are heavily biased towards first posts. Posts the next day might as well have never happened. Another thing is on HN, downvoting based on politics rather than quality is proscribed but happens relentlessly anyway. I prefer the egalitarian approach of all voices here are equal. The only moderation we do is deleting posts that don't belong here, like spam.
Mar 31 2021
parent FeepingCreature <feepingcreature gmail.com> writes:
On Thursday, 1 April 2021 at 02:11:17 UTC, Walter Bright wrote:
 On 3/31/2021 2:40 PM, Adam D. Ruppe wrote:
 I'm sure this wouldn't be terribly difficult to implement, but 
 I also don't think it is a good idea. Online voting is 
 emotionally toxic and logically fallacious. We're better off 
 without it.
Votes on Reddit and Hackernews are heavily biased towards first posts. Posts the next day might as well have never happened. Another thing is on HN, downvoting based on politics rather than quality is proscribed but happens relentlessly anyway. I prefer the egalitarian approach of all voices here are equal. The only moderation we do is deleting posts that don't belong here, like spam.
I like the following vote logic: - topics sort by vote count mixed with newness - top-level replies sort by new (recency) - mid-level replies sort by vote. This is what Reddit does when you set a thread to "New by default." The downside of a purely recency based sorting is that it favors verbosity and controversy rather than consensus - threads stay on the frontpage by being *debated*, rather than by being agreed or disagreed on, resulting in posts where everyone agrees and nobody has anything to add quickly falling off the forum. I agree with all the downsides of votes you highlight, but as long as new toplevel replies can become visible, I think ratings are useful to keep valuable topics alive and promote interesting replies.
Apr 01 2021
prev sibling next sibling parent Guillaume Piolat <first.name spam.org> writes:
On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe wrote:
 I'm sure this wouldn't be terribly difficult to implement, but 
 I also don't think it is a good idea. Online voting is 
 emotionally toxic and logically fallacious. We're better off 
 without it.
+1 I think most posters here, even the slightly abrasive ones (and this is becoming rare), implement a sort of self-moderation, based upon what is valued on these forums.
Apr 01 2021
prev sibling next sibling parent reply Q. Schroll <qs.il.paperinik gmail.com> writes:
On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe wrote:
 Online voting is emotionally toxic and logically fallacious.
You're wrong, like really wrong. If that were true, [SO](https://stackoverflow.com) should be the pinnacle of emotionally toxic and logically fallacious posts when it comes to programming topics, but it's exactly the contrary: SO is one of the best if not *the* best resource for programming questions &mdash; and I'd claim it is because of their voting system and not in spite of it. But I do see where you're coming from. You're probably thinking of Facebook, Twitter, Reddit, you name it. Notice that those are for general audiences handling all sorts of topics. Voting turns bad stuff worse, when there's only one option (say, "Like", but no "Dislike"). YouTuber EmpLemon made a great video¹ that not only explains why dislikes are good but also why likes are good. One of the best features a voting system brings is enabling people to say "this is good" or "this is bad" without disrupting a discussion and without needing to qualify it. Posts containing *only* "+1" (or "Disagree") are a bad replacement of voting and, in the presence of a voting system, can be regarded as spam.
 We're better off without it.
Now, you're talking. *We,* the users of the D Language Forum, probably don't need voting. That's because there's a difference between a general forum like social networks, a specialized forum like SO, and very specialized forums like this one. The most value voting would generate is on DIP discussions and other suggestions. It may be stuff as simple as *Should the `=>` shorthand be extended to functions?* that have a subjective nuance. Maybe it could be a checkbox next to the *Enable Markdown* that enables voting on a new post. ¹ EmpLemon: [*Why You Should Like Dislikes*](https://www.youtube.com/watch?v=fdGbOT8NXnE) (~ 30 min.)
Apr 01 2021
next sibling parent reply matheus <matheus gmail.com> writes:
On Thursday, 1 April 2021 at 14:33:43 UTC, Q. Schroll wrote:
 On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe 
 wrote:
 Online voting is emotionally toxic and logically fallacious.
You're wrong, like really wrong. If that were true, [SO](https://stackoverflow.com) should be the pinnacle of emotionally toxic and logically fallacious posts when it comes to programming topics, but it's exactly the contrary: SO is one of the best if not *the* best resource for programming questions &mdash; and I'd claim it is because of their voting system and not in spite of it.
Well I don't agree with this, SO voting system isn't that good at all. It was already debatable on HN and in different medias too. For example, it's common to see that in different cases the most up voted answer is "not" the "best answer". Matheus.
Apr 01 2021
next sibling parent matheus <matheus gmail.com> writes:
On Thursday, 1 April 2021 at 14:59:44 UTC, matheus wrote:
 ...
By the way about SO there is another problem, they lock a topic/question after sometime, and now the most up voted answer remains as solution. But what is/was best in one point of time, may not be suitable in the future, but when looking for the problem, the old solution will show on top of the search engines. For me SO popularity may be connected with laziness too. Where I work I already saw people just copying and pasting source direct from there, I know that because the language (I'm foreigner) and the code had English text, so looking over I hit that site. Matheus.
Apr 01 2021
prev sibling next sibling parent Q. Schroll <qs.il.paperinik gmail.com> writes:
On Thursday, 1 April 2021 at 14:59:44 UTC, matheus wrote:
 On Thursday, 1 April 2021 at 14:33:43 UTC, Q. Schroll wrote:
 On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe 
 wrote:
 Online voting is emotionally toxic and logically fallacious.
You're wrong, like really wrong. If that were true, [SO](https://stackoverflow.com) should be the pinnacle of emotionally toxic and logically fallacious posts when it comes to programming topics, but it's exactly the contrary: SO is one of the best if not *the* best resource for programming questions &mdash; and I'd claim it is because of their voting system and not in spite of it.
Well I don't agree with this, SO voting system isn't that good at all. It was already debatable on HN and in different medias too. For example, it's common to see that in different cases the most up voted answer is "not" the "best answer".
I don't think we disagree, you just got me wrong. People's votes aren't perfect, everyone knows that. I experienced it myself that not the accepted/highest-score answer, but the answer with the second or third highest score was most helpful to me. I never said SO's voting system is the best voting system there is, I said that SO *has* a voting system and SO (not its voting system) is generally considered the best resource for programming questions, which is an indication that a voting system on a programming forum might not be as harmful as people here expect. I have no idea what HN stands for.
Apr 01 2021
prev sibling parent Ola Fosheim =?UTF-8?B?R3LDuHN0YWQ=?= <ola.fosheim.grostad gmail.com> writes:
On Thursday, 1 April 2021 at 14:59:44 UTC, matheus wrote:
 Well I don't agree with this, SO voting system isn't that good 
 at all. It was already debatable on HN and in different medias 
 too.

 For example, it's common to see that in different cases the 
 most up voted answer is "not" the "best answer".
Yep. Often the most recent answer is the best one on Stack Overflow... What SO does reasonably well is classifying topics, but frequently the same topic is covered in multiple places, hard to fix. For languages that are used by people without formal education in programming (e.g. javascript) the most voted answer on SO tend to be not very good, and cute hacks would be promoted. You sometimes see this in the D forums too. Cute language hacks that makes code unreadable are popular in the D forums! For languages that are evolving (like C++) the most voted answer tend to be counter productive and convoluted... meaning you get an answer to solve something in a complicated way that there is a better to do in the most recent version of the language.
Apr 30 2021
prev sibling next sibling parent Timon Gehr <timon.gehr gmx.ch> writes:
On 01.04.21 16:33, Q. Schroll wrote:
 On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe wrote:
 Online voting is emotionally toxic and logically fallacious.
You're wrong, like really wrong. If that were true, [SO](https://stackoverflow.com) should be the pinnacle of emotionally toxic and logically fallacious posts when it comes to programming topics, but it's exactly the contrary: SO is one of the best if not *the* best resource for programming questions &mdash; and I'd claim it is because of their voting system and not in spite of it.
You are shifting the goal posts. His point was about voting on individual posts that are part of some debate.
Apr 01 2021
prev sibling next sibling parent tsbockman <thomas.bockman gmail.com> writes:
On Thursday, 1 April 2021 at 14:33:43 UTC, Q. Schroll wrote:
 On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe 
 wrote:
 Online voting is emotionally toxic and logically fallacious.
You're wrong, like really wrong. If that were true, [SO](https://stackoverflow.com) should be the pinnacle of emotionally toxic and logically fallacious posts when it comes to programming topics, but it's exactly the contrary: SO is one of the best if not *the* best resource for programming questions &mdash; and I'd claim it is because of their voting system and not in spite of it.
Forum voting tends to works OK when an issue is reasonably clear, objective, and testable, or low-stakes, or when there is no need nor intention to force the majority's opinion on the minority. If someone disagrees with the highest voted answer on a Stack Overflow question, they can just use a different one, or make up their own. By contrast, if someone disagrees with how their national government is run, they generally can't just go start a new one. Hence, people tend to be much more aggressive or manipulative in their use of voting in political discussions. Threads in D's "Learn" forum, most bug reports, and most pull requests are mostly the former sort of content, where voting isn't a problem. However, the long debates in the D's "General" forum concerning the future direction of the language are sometimes closer to the latter sort of content, where anyone who wants or needs to participate in the ecosystem just has to live with the consequences of whatever decision is ultimately made. Hence, passions run higher, and the risk of abusive voting is much higher.
Apr 01 2021
prev sibling next sibling parent Kagamin <spam here.lot> writes:
On Thursday, 1 April 2021 at 14:33:43 UTC, Q. Schroll wrote:
 On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe 
 wrote:
 Online voting is emotionally toxic and logically fallacious.
You're wrong, like really wrong. If that were true, [SO](https://stackoverflow.com) should be the pinnacle of emotionally toxic and logically fallacious posts when it comes to programming topics, but it's exactly the contrary: SO is one of the best if not *the* best resource for programming questions &mdash; and I'd claim it is because of their voting system and not in spite of it.
I think it's useful because it exists, is user generated and is searchable. Not sure what votes do there, they are time skewed, the best answer is routinely undervoted, the accepted answer is time skewed too and fails on top of that. As a result I end up reading all answers and comments to answers. Github issues are better for more focused information have no voting system, they too are useful because they exist, are user generated and are searchable.
Apr 02 2021
prev sibling parent deadalnix <deadalnix gmail.com> writes:
On Thursday, 1 April 2021 at 14:33:43 UTC, Q. Schroll wrote:
 On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe 
 wrote:
 Online voting is emotionally toxic and logically fallacious.
You're wrong, like really wrong. If that were true, [SO](https://stackoverflow.com) should be the pinnacle of emotionally toxic and logically fallacious posts when it comes to programming topics, but it's exactly the contrary: SO is one of the best if not *the* best resource for programming questions &mdash; and I'd claim it is because of their voting system and not in spite of it.
Well, yes and no. I've seen subreddit go into runaway mode where some content gets upvoted, which gets folks to self select and both feed into each other. The place turn to shit in no time. It can work, but it require heavy moderation, which I don't think anyone want to do here. You'll note that all successful example are heavily curated to begin with.
Apr 02 2021
prev sibling parent NonNull <non-null use.startmail.com> writes:
On Wednesday, 31 March 2021 at 21:40:15 UTC, Adam D. Ruppe wrote:
 I'm sure this wouldn't be terribly difficult to implement, but 
 I also don't think it is a good idea. Online voting is 
 emotionally toxic and logically fallacious. We're better off 
 without it.
Very well said.
Apr 02 2021
prev sibling next sibling parent Clint E. <clint gmail.com> writes:
On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei Alexandrescu 
wrote:
 I wonder how difficult it would be to add a voting system for 
 forum messages. Votes wouldn't be available in the NNTP 
 interface but would be accessible in the Web interface and 
 would allow sorting and filtering by votes.

 cc Vladimir :o)
If you want a voting system, just use old.reddit.com/r/dlang Clint.
Mar 31 2021
prev sibling next sibling parent reply 12345swordy <alexanderheistermann gmail.com> writes:
On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei Alexandrescu 
wrote:
 I wonder how difficult it would be to add a voting system for 
 forum messages. Votes wouldn't be available in the NNTP 
 interface but would be accessible in the Web interface and 
 would allow sorting and filtering by votes.

 cc Vladimir :o)
Here is an better idea: Why not allow us to edit our post to correct typos? -Alex
Mar 31 2021
next sibling parent Walter Bright <newshound2 digitalmars.com> writes:
On 3/31/2021 5:55 PM, 12345swordy wrote:
 Here is an better idea: Why not allow us to edit our post to correct typos?
Because it is implemented using NNTP. Editing posts means people will wind up with different versions of a post, which may be baffling. I would go for, at least with the DFeed, delaying posting a post for, say, 10 minutes to allow one to have second thoughts.
Mar 31 2021
prev sibling parent reply Clint E. <clint gmail.com> writes:
On Thursday, 1 April 2021 at 00:55:14 UTC, 12345swordy wrote:
 On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei 
 Alexandrescu wrote:
 I wonder how difficult it would be to add a voting system for 
 forum messages. Votes wouldn't be available in the NNTP 
 interface but would be accessible in the Web interface and 
 would allow sorting and filtering by votes.

 cc Vladimir :o)
Here is an better idea: Why not allow us to edit our post to correct typos? -Alex
The Non-editable feature is one of the best of this forum. I hope they keep that way. When you are able to edit, people say things reckless and times later they change their mind and edit, causing confusion most of time, and this happens a lot on reddit for example. We already have this great power to think before posting. Clint E.
Mar 31 2021
next sibling parent Jacob Carlborg <doob me.com> writes:
On 2021-04-01 05:32, Clint E. wrote:

 The Non-editable feature is one of the best of this forum.
 
 I hope they keep that way. When you are able to edit, people say things 
 reckless and times later they change their mind and edit, causing 
 confusion most of time, and this happens a lot on reddit for example.
 
 We already have this great power to think before posting.
When I use a platform that does support editing, most of my changes are correcting spell errors. Or when I've forgot to finish a sentence before starting a new one. I think GitHub hits a sweet spot. They allow editing comments but retain the full history which everyone can see. -- /Jacob Carlborg
Apr 01 2021
prev sibling parent 12345swordy <alexanderheistermann gmail.com> writes:
On Thursday, 1 April 2021 at 03:32:49 UTC, Clint E. wrote:
 On Thursday, 1 April 2021 at 00:55:14 UTC, 12345swordy wrote:
 On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei 
 Alexandrescu wrote:
 I wonder how difficult it would be to add a voting system for 
 forum messages. Votes wouldn't be available in the NNTP 
 interface but would be accessible in the Web interface and 
 would allow sorting and filtering by votes.

 cc Vladimir :o)
Here is an better idea: Why not allow us to edit our post to correct typos? -Alex
The Non-editable feature is one of the best of this forum. I hope they keep that way. When you are able to edit, people say things reckless and times later they change their mind and edit, causing confusion most of time, and this happens a lot on reddit for example. We already have this great power to think before posting. Clint E.
This isn't reddit though. -Alex
Apr 01 2021
prev sibling next sibling parent Kagamin <spam here.lot> writes:
Dunno, this attention economy all too easily devolves into dark 
patterns. Talking about youtube, an addiction-inducing 
recommendation feed improved engagement metric too. On top of 
that nobody figured out what the voted are or do, the feature is 
just there and everybody ends up using it unconsciously, which 
can't possibly go right.
Mar 31 2021
prev sibling next sibling parent Jacob Carlborg <doob me.com> writes:
On 2021-03-31 23:26, Andrei Alexandrescu wrote:
 I wonder how difficult it would be to add a voting system for forum 
 messages. Votes wouldn't be available in the NNTP interface but would be 
 accessible in the Web interface and would allow sorting and filtering by 
 votes.
GitHub has a new feature (still in beta), Discussions, which is kind of like a forum. I don't have much experience with it but it looks like it supports some form of voting. It also supports two form of discussions: open ended and questions. A question discussion can be marked as answered, which is pretty nice. -- /Jacob Carlborg
Apr 01 2021
prev sibling parent reply Vladimir Panteleev <thecybershadow.lists gmail.com> writes:
On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei Alexandrescu 
wrote:
 I wonder how difficult it would be to add a voting system for 
 forum messages. Votes wouldn't be available in the NNTP 
 interface but would be accessible in the Web interface and 
 would allow sorting and filtering by votes.

 cc Vladimir :o)
This idea did come up before, and is controversial as others have pointed out. My personal thoughts about this today are: - Downvotes are not nice to receive and in general carry bad vibes. - Sorting or filtering by votes creates unhealthy incentives - there is now a system which would be advantageous to "game". - Having a way to publicly communicate agreement or appreciation with a post without all the bulky noise associated with a reply might be nice. Good vibes only! - The only way I can think of to translate the votes to the NNTP/email medium is either converting them to messages which have nothing but a +1 (which is very spammy), or not doing it at all. Neither is ideal. - Different people engage with the community with different attitudes and expectations. I'm not sure how much sense it would make to flatten this to a single number. - In any case, at the moment we seem to be doing okay, at least as far as people intentionally participating in unconstructive ways.
Apr 01 2021
next sibling parent Jacob Carlborg <doob me.com> writes:
On 2021-04-01 10:40, Vladimir Panteleev wrote:

 This idea did come up before, and is controversial as others have 
 pointed out.
Someway to indicate a question is answered or a topic is resolved would be nice to have as well. It wouldn't be spammy for NNTP users since it would only be one message per thread. -- /Jacob Carlborg
Apr 01 2021
prev sibling next sibling parent reply Meta <jared771 gmail.com> writes:
On Thursday, 1 April 2021 at 08:40:10 UTC, Vladimir Panteleev 
wrote:
 - The only way I can think of to translate the votes to the 
 NNTP/email medium is either converting them to messages which 
 have nothing but a +1 (which is very spammy), or not doing it 
 at all. Neither is ideal.
Every time a new feature for the newsgroup comes up, and it gets shot down because it wouldn't work for people using email, I immediately think of https://xkcd.com/1782.
Apr 01 2021
parent Abdulhaq <alynch4047 gmail.com> writes:
On Thursday, 1 April 2021 at 13:37:01 UTC, Meta wrote:
 On Thursday, 1 April 2021 at 08:40:10 UTC, Vladimir Panteleev 
 wrote:
 - The only way I can think of to translate the votes to the 
 NNTP/email medium is either converting them to messages which 
 have nothing but a +1 (which is very spammy), or not doing it 
 at all. Neither is ideal.
Every time a new feature for the newsgroup comes up, and it gets shot down because it wouldn't work for people using email, I immediately think of https://xkcd.com/1782.
+1
Apr 01 2021
prev sibling next sibling parent reply Andrei Alexandrescu <SeeWebsiteForEmail erdani.org> writes:
On 4/1/21 4:40 AM, Vladimir Panteleev wrote:
 On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei Alexandrescu wrote:
 I wonder how difficult it would be to add a voting system for forum 
 messages. Votes wouldn't be available in the NNTP interface but would 
 be accessible in the Web interface and would allow sorting and 
 filtering by votes.

 cc Vladimir :o)
This idea did come up before, and is controversial as others have pointed out. My personal thoughts about this today are: - Downvotes are not nice to receive and in general carry bad vibes. - Sorting or filtering by votes creates unhealthy incentives - there is now a system which would be advantageous to "game". - Having a way to publicly communicate agreement or appreciation with a post without all the bulky noise associated with a reply might be nice. Good vibes only! - The only way I can think of to translate the votes to the NNTP/email medium is either converting them to messages which have nothing but a +1 (which is very spammy), or not doing it at all. Neither is ideal. - Different people engage with the community with different attitudes and expectations. I'm not sure how much sense it would make to flatten this to a single number. - In any case, at the moment we seem to be doing okay, at least as far as people intentionally participating in unconstructive ways.
Thanks for the thoughtful response. And also your recent work on post formatting is just amazing. Great stuff! I appreciate your nuanced take as I was a bit taken aback by the blanket statement that voting is bad, which I find anachronistic in a "don't use cars because accidents" way. Proposal got downvoted as it were :o). I agree that an upvote-only system similar to Twitter and Facebook would be nice to signal content that people find valuable. (Anecdote: A "dislike" button has been the top requested feature for Facebook for years. And not only casually by the users; entire business propositions have been made based on that. I'm glad they were not accepted.) As to the NNTP integration, I was thinking the voting could be a web-only feature. Which is nice because it encourages people to swap to the modern web interface (he wrote, typing furiously in his Thunderbird window). Perhaps do the converse instead, if a post is just a +1 with no other added content, convert it to a vote. No filtering or sorting by popularity would be needed, probably for years. Just a little heart out there with the number of votes would add great value. Thanks for engaging.
Apr 01 2021
next sibling parent Alexandru Ermicioi <alexandru.ermicioi gmail.com> writes:
On Thursday, 1 April 2021 at 14:25:09 UTC, Andrei Alexandrescu 
wrote:
 ...
It seems that we already have some sort of voting system in these forums, just look at all of these +1 replies, it is just not being used by web interface, to provide some useful insight. I think it is possible to have a voting system, based on +1 emails which people tend to reply when they see a message or a section of that message that they agree on, and simply have nothing more to say. Usually such messages start with an excerpt from previous message, followed by a +1 and optionally some text that expresses users opinion on that message (by enforcing or complementing it). If we do implement it based on such kind of emails, the NNTP clients would see them as another bunch of emails, however the web interface would be able to parse them, and then highlight original message with how many up-votes or down-votes (if any) that section of message has. It may even optionally show list of replies that up-voted/down-voted that section. The advantage of using +1 emails, is that user doesn't just up-vote or down-vote a message, he has to specify which part of message he up-votes/down-votes. The web interface may introduce also some restrictions to down-vote capability requiring user to type some reason to down-vote, otherwise it would not be counted. As for users who down-vote/up-vote ignoring restrictions found in web interface, web interface should just ignore them, and hide from any insight it may provide. Best regards, Alexandru.
Apr 02 2021
prev sibling parent reply Kagamin <spam here.lot> writes:
On Thursday, 1 April 2021 at 14:25:09 UTC, Andrei Alexandrescu 
wrote:
 I appreciate your nuanced take as I was a bit taken aback by 
 the blanket statement that voting is bad, which I find 
 anachronistic in a "don't use cars because accidents" way. 
 Proposal got downvoted as it were :o).
I thought about this a bit more and I think a decent voting system can be implemented in this particular forum :) 1. implement voting in the form of moderation (moting?). 2. the trick: reuse the flagging team to do moting; I assume if they can flag, they can also implement a moting policy. 3. the moting policy for a start: vote factual informative posts like Steven's description of slice structure, documentation clarifications, explanations of GC internals, helpful code samples; but not, e.g. a proposal of an interpolated string design, because it's not a fact, but a proposed design tradeoff competing with other proposed tradeoffs; rationale: upvotes are de jure positive, but de facto you can be surprised to discover dark corners of your psychics that never saw the light of day; facts should be sufficiently non-controversial for everyone. 4. possible UX: don't show votes in interface, only use them internally for sorting when one requests such sorting, the resulting order should be evident enough.
Apr 02 2021
parent Kagamin <spam here.lot> writes:
Downsides:
1. absence of karma communicates a message of egalitarianism, 
introduction of karma will make that message go away.
2. good karma system was never tried, which means all karma 
systems are failures, introduction of karma will communicate a 
message that this forum is yet another failure.
Apr 03 2021
prev sibling parent reply Q. Schroll <qs.il.paperinik gmail.com> writes:
On Thursday, 1 April 2021 at 08:40:10 UTC, Vladimir Panteleev 
wrote:
 On Wednesday, 31 March 2021 at 21:26:48 UTC, Andrei 
 Alexandrescu wrote:
 I wonder how difficult it would be to add a voting system for 
 forum messages. Votes wouldn't be available in the NNTP 
 interface but would be accessible in the Web interface and 
 would allow sorting and filtering by votes.
This idea did come up before, and is controversial as others have pointed out. My personal thoughts about this today are: - Downvotes are not nice to receive and in general carry bad vibes.
I agree that downvotes are not nice to receive. Keeping someone in a cushion can be worse in the long run than a clear cut message such as a 50%+ dislike ratio they need to hear.
 - Sorting or filtering by votes creates unhealthy incentives - 
 there is now a system which would be advantageous to "game".
I agree, especially if the sorted/filtered view is the default one. If you'd have to actively seek it, I guess it wouldn't lead to trouble. Defaults matter.
 - Having a way to publicly communicate agreement or 
 appreciation with a post without all the bulky noise associated 
 with a reply might be nice.
Yes. And having a way to publicly communicate disagreement or disapproval with a post without all the bulky noise associated with a reply might be nice, too.
 Good vibes only!
No. Bad feelings can be healthy. No *toxic* vibes, yes! I often got bad feelings engaging with people here. I'm always a little nervous about what people think when posting. That's okay.
 - The only way I can think of to translate the votes to the 
 NNTP/email medium is either converting them to messages which 
 have nothing but a +1 (which is very spammy), or not doing it 
 at all. Neither is ideal.
I cannot say anything about that.
 - Different people engage with the community with different 
 attitudes and expectations. I'm not sure how much sense it 
 would make to flatten this to a single number.
Constructive people engage constructively. Toxic people engage toxicly. It's really >99% the people (posters and moderation) and <1% the system. I have years of experience with two apps where people could post jokes, both with up- and down-votes, in both one could comment on posts and vote even the comments. * One had no manual moderation (posts were deleted on a flag vs votes basis) and a really toxic part-community. There were (separate) groups who posted spam, extremely offensive jokes, or falsely flagged posts of popular people. * The other has only manual moderation and almost no toxic posts. I was a moderator myself for roughly a year. When you get a lot of down-votes there, it's a clear indication that your joke was unfunny, really bad or stupidly offensive. Now here's the funny part: The second app was developed as a better and less toxic version of the first one, and it worked, apparently despite having down-votes.
 - In any case, at the moment we seem to be doing okay, at least 
 as far as people intentionally participating in unconstructive 
 ways.
Agreed. No one here thinks we desperately need votes. It's just a cost&ndash;risk&ndash;benefit analysis. Apart from the implementation effort, I think having two numbers below the name, photo, "Posted in reply to" in the box on the right will do very little harm at worst, but give people information.
Apr 01 2021
parent reply ag0aep6g <anonymous example.com> writes:
On 01.04.21 17:38, Q. Schroll wrote:
 cost&ndash;risk&ndash;benefit analysis.
Vladimir, I think something went wrong with the rich text thingy. This message shows "&ndash;" in Thunderbird. Q. Schroll, I think I accidentally sent this to your email before. Sorry about that. Thunderbird switched the buttons around.
Apr 01 2021
parent reply Vladimir Panteleev <thecybershadow.lists gmail.com> writes:
On Thursday, 1 April 2021 at 15:56:23 UTC, ag0aep6g wrote:
 On 01.04.21 17:38, Q. Schroll wrote:
 cost&ndash;risk&ndash;benefit analysis.
Vladimir, I think something went wrong with the rich text thingy. This message shows "&ndash;" in Thunderbird.
They are that way because the author typed them that way. It is probably the easiest way to type those characters on many systems. I added a lint rule which warns on HTML entities, and offers to fix it (by converting them into Unicode characters).
Apr 02 2021
parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 4/2/2021 12:58 AM, Vladimir Panteleev wrote:
  Vladimir, I think something went wrong with the rich text thingy. This 
 message shows "&ndash;" in Thunderbird.
They are that way because the author typed them that way. It is probably the easiest way to type those characters on many systems.
I cast about for a long time on the easiest way to add Unicode characters in microEmacs. I finally settled on placing the cursor over a character, and hitting ^X-U. It will scroll through related symbols. Put it over an a, and it will scroll through ä, etc. You don't need to remember anything but "it looks like an a". Of course this doesn't remotely handle all Unicode characters, but it gets the most used ones.
Apr 02 2021
next sibling parent reply Vladimir Panteleev <thecybershadow.lists gmail.com> writes:
On Friday, 2 April 2021 at 10:07:14 UTC, Walter Bright wrote:
 On 4/2/2021 12:58 AM, Vladimir Panteleev wrote:
  Vladimir, I think something went wrong with the rich text 
 thingy. This message shows "&ndash;" in Thunderbird.
They are that way because the author typed them that way. It is probably the easiest way to type those characters on many systems.
I cast about for a long time on the easiest way to add Unicode characters in microEmacs. I finally settled on placing the cursor over a character, and hitting ^X-U. It will scroll through related symbols. Put it over an a, and it will scroll through ä, etc. You don't need to remember anything but "it looks like an a". Of course this doesn't remotely handle all Unicode characters, but it gets the most used ones.
Here is how I do it: https://dump.cy.md/de83f239e276fa30d882c736f8fba7f3/2021-04-02%2010%3A09%3A45.mp4 😃
Apr 02 2021
parent reply Walter Bright <newshound2 digitalmars.com> writes:
On 4/2/2021 3:12 AM, Vladimir Panteleev wrote:
 Here is how I do it:
 
 https://dump.cy.md/de83f239e276fa30d882c736f8fba7f3/2021-04-02%2010%3A09%3A45.mp4
 
 😃
Sweet! Yours is obviously better. Here are the tables for mine: https://github.com/DigitalMars/med/blob/master/src/med/more.d#L350
Apr 02 2021
parent Imperatorn <johan_forsberg_86 hotmail.com> writes:
On Friday, 2 April 2021 at 20:39:43 UTC, Walter Bright wrote:
 On 4/2/2021 3:12 AM, Vladimir Panteleev wrote:
 Here is how I do it:
 
 https://dump.cy.md/de83f239e276fa30d882c736f8fba7f3/2021-04-02%2010%3A09%3A45.mp4
 
 😃
Sweet! Yours is obviously better. Here are the tables for mine: https://github.com/DigitalMars/med/blob/master/src/med/more.d#L350
https://blogs.windows.com/windowsexperience/2018/02/05/windows-10-tip-get-started-emoji-keyboard-shortcut/ Tldr: On Windows, press Win + . to bring up the emoji picker
Apr 30 2021
prev sibling parent reply =?UTF-8?Q?Ali_=c3=87ehreli?= <acehreli yahoo.com> writes:
On 4/2/21 3:07 AM, Walter Bright wrote:

 I finally settled on placing the cursor over a
 character, and hitting ^X-U. It will scroll through related symbols.
I use a similar idea in *regular* :) Emacs for only the Turkish subset=20 of Unicode. Because Turkish alphabet is based on the Latin alphabet, I came up with=20 the idea of typing a key twice to produce its Turkish relation. Copying=20 from my .emacs file: gg -> =C4=9F II -> =C4=B0 ii -> =C4=B1 etc. And there are escape sequences when I really want e.g. gg instead of =C4=9F= ,=20 and there are escapes for espaces. :) ("gg/" ["gg"]) ("gg//" ["=C4=9F/"]) ("gg/g" ["g=C4=9F"]) You can tell by the sound of my keyboard whether I'm typing an English=20 text or a Turkish text. :) Ali P.S. Here is the entire definition of the input method I named=20 'turkish-ali'. (Used with M-x set-input-method) (require 'quail) (quail-define-package "turkish-ali" "UTF-8" "TRa<" t "Turkish input method by Ali <key descriptions go here> " nil t nil nil nil nil nil nil nil nil t) (quail-define-rules ("A^" ?=C3=82) ("a^" ?=C3=A2) ("CC" ?=C3=87) ("cc" ?=C3=A7) ("GG" ?=C4=9E) ("gg" ?=C4=9F) ("II" ?=C4=B0) ("ii" ?=C4=B1) ("I^" ?=C3=8E) ("i^" ?=C3=AE) ("OO" ?=C3=96) ("oo" ?=C3=B6) ("SS" ?=C5=9E) ("ss" ?=C5=9F) ("UU" ?=C3=9C) ("uu" ?=C3=BC) ("U^" ?=C3=9B) ("u^" ?=C3=BB) ("A^/" ["A^"]) ("A^//" ["=C3=82/"]) ("a^/" ["a^"]) ("a^//" ["=C3=A2/"]) ("CC/" ["CC"]) ("CC//" ["=C3=87/"]) ("CC/C" ["C=C3=87"]) ("cc/" ["cc"]) ("cc//" ["=C3=A7/"]) ("cc/c" ["c=C3=A7"]) ("GG/" ["GG"]) ("GG//" ["=C4=9E/"]) ("GG/G" ["G=C4=9E"]) ("gg/" ["gg"]) ("gg//" ["=C4=9F/"]) ("gg/g" ["g=C4=9F"]) ("II/" ["II"]) ("II//" ["=C4=B0/"]) ("II/I" ["I=C4=B0"]) ("ii/" ["ii"]) ("ii//" ["=C4=B1/"]) ("i^/" ["i^"]) ("i^//" ["=C3=AE/"]) ("OO/" ["OO"]) ("OO//" ["=C3=96/"]) ("OO/O" ["O=C3=96"]) ("oo/" ["oo"]) ("oo//" ["=C3=B6/"]) ("oo/o" ["o=C3=B6"]) ("SS/" ["SS"]) ("SS//" ["=C5=9E/"]) ("SS/S" ["S=C5=9E"]) ("ss/" ["ss"]) ("ss//" ["=C5=9F/"]) ("ss/s" ["s=C5=9F"]) ("UU/" ["UU"]) ("UU//" ["=C3=9C/"]) ("UU/U" ["U=C3=9C"]) ("uu/" ["uu"]) ("uu//" ["=C3=BC/"]) ("uu/u" ["u=C3=BC"]) ("U^/" ["U^"]) ("U^//" ["=C3=9B/"]) ("u^/" ["u^"]) ("u^//" ["=C3=BB/"]) )
Apr 02 2021
parent Walter Bright <newshound2 digitalmars.com> writes:
On 4/2/2021 2:09 PM, Ali Çehreli wrote:
 On 4/2/21 3:07 AM, Walter Bright wrote:
 
 I finally settled on placing the cursor over a
 character, and hitting ^X-U. It will scroll through related symbols.
I use a similar idea in *regular* :) Emacs for only the Turkish subset of Unicode.
Nice! The trouble with my scheme is if you go too far, you have to "go around the horn" to recover. How I type Ç in email is to cut&paste it from your name :-)
Apr 02 2021