|
| I'll tweak the pattern matching again. Some cleanup I did yesterday has removed some of the previous options, plus there is an end-of-string error on a couple of the smileys I didn't catch. cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
Well, I was just browing through a few articles on CodeProject. I don't exactly remember which ones, but some of them were really good! They were worth reading and learning from.
However, I saw something which was really awkward. One of the articles had tons of votes of fives and fours, and a vote of one and two. After scrolling down to the bottom, the voters who voted one or two did not specify there reason for the low vote.
What I want to know is why exactly would anyone vote an article "1" (without any reason), when tons of people are voting the same article "5."
I just wanted this to be brought to attention.
*DISCLAIMER: I am not saying this because it happened to my article; I am saying this because this is slight fishy. Why is an article voted 1, when like 20 people vote it 5. Even if you want to vote 1, you should give a reason, so that the author can improve himself.
Thanks,
Harsimran Singh1) The good news about computers is that they do what you tell them to do. The bad news is that they do what you tell them to do. - Ted Nelson
2) Why is it drug addicts and computer afficionados are both called users? - Clifford Stoll
3) The real danger is not that computers will begin to think like men, but that men will begin to think like computers. - Sydney J. Harris
(Computer code: 00001111 - translation: Hello! :P )
REMEMBER:
"Computers are made for us, we are not made for th
|
|
|
|
|
This is a widespread problem at Code Project with no clear solution. Of course, people have varied motivations (finding out what they are would be like asking why "some people murder others when most other people don't murder them"). Here are what I expect to be some common reasons:
1) The voter wanted to see something in the article that wasn't there, and voted entirely based on the lack of that one small item.
-For example, somebody down-voted one of my articles because he thought I should have graphs and metrics and such.
2) The voter took a quick glance at the article and voted based on some immediate instinct (like "the article seems too long").
3) Some people have huge egos and, when they see something they interpret as a flaw (perhaps a simple typo), they will down vote it to attain a sense of power and worth.
4) Rival authors that want their article to be rated higher than yours.
5) Authors that just want all other articles to less popular than theirs (fame cravers).
6) Maybe you pissed them off by commenting on some forum post, so they are relaliating by voting your article down.
The list goes on and on... some people just do this. Though, the majority of votes will usually drown out these minority votes. Still, it would be nice if a coherent response was required for 1-voters, rather than those who just post "-" as their message body when they 1-vote. I think those posts can be reported as abuse and, if they get deleted, the 1-vote will also be removed from the article... not sure about that though.
|
|
|
|
|
I don't think it is as bad as you make it sound. See my message below.
 Luc Pattyn [Forum Guidelines] [Why QA sucks] [My Articles]
I only read code that is properly formatted, adding PRE tags is the easiest way to obtain that. All Toronto weekends should be extremely wet until we get it automated in regular forums, not just QA.
|
|
|
|
|
I basically just listed the motivations of why people 1-vote. I don't think the problem is too terrible (though it is common to most every article, so it is "widespread", as I say). See the bottom of MY message above, which includes some of the balances you speak of in your below message.
|
|
|
|
|
Yes I understand that, however the negatives took most of your text, and a reader with a slightly negative mindset would pick up all the reasons why such injustice may be committed, and end up with an even more negative mindset. I'm more of a positive thinker, so I felt a need to compensate and stress the positive things in this community; and doing so I probably illustrated my vote-balancing point myself).
Luc Pattyn [Forum Guidelines] [Why QA sucks] [My Articles]
I only read code that is properly formatted, adding PRE tags is the easiest way to obtain that. All Toronto weekends should be extremely wet until we get it automated in regular forums, not just QA.
|
|
|
|
|
|
yes, there is an occasional down-vote amidst a lot of high votes, often for no apparent reason. Which always is a pity, and may come as a shock to novices, especially on a new article, when there aren't many high votes yet.
But then one of many things tends to happen:
- people start to add high votes, even more so than they would have done if no low-vote were present;
- if the explanatory message seems inappropriate, people start to click "vote to remove"; when a sufficient number of people do this, the message holding the low vote gets automatically removed, and so does the associated vote.
- if the low-voter is recognized as a repeat offender, he will be signaled in the "suggestions and bugs" forum, possibly resulting in a lot or even all of his messages and votes being removed.
So in all the community tends to take care of it, with some delay.
The CP voting system itself also offers some protection as it uses weighted voting; new members have a vote weight of 1, silver/gold/platinum members have a vote weight of 2/4/8, and we can safely assume those aren't the ones casting malicious votes; so even if a moron votes you a 1, the next gold member voting you 5 brings your average up to (1+4*5)/(1+4) = 4.2
Luc Pattyn [Forum Guidelines] [Why QA sucks] [My Articles]
I only read code that is properly formatted, adding PRE tags is the easiest way to obtain that. All Toronto weekends should be extremely wet until we get it automated in regular forums, not just QA.
|
|
|
|
|
I think aspdotnetdev hit the nail on the head. Most of the people that do it are ding it maliciously. Sure, it's nice that you're a "positive thinker", but the reality is that most of the 1/2 votes are from people that you've managed to annoy/upset/piss off in some way. In fact, someone yesterday stated that "3 is the new 1", and I think he's right. voting a 3 doesn't require you to say why you voted a 3, but it's still an effective way to exact retribution on an article.
I still think numeric voting should be changed to "Is Useful", and maybe even a series of checkboxes that a user can click about an article to generate an overall "score":
Useful (the code or technique was used in my code)
Accurate (the description of the technique is accurate and reflects accepted practices as of the writing of the article)
Thorough (the code and technique are thoroughly discussed and presented in a meaningful manner)
Formatting (the article is formatted correct and adheres to the CP style)
Bookmarked (automatically checked if the user bookmarks the article)
Maybe the article's score could be calculated using the number of views compared to the number of users who marked those checkboxes.
Of course, membership status might play a big part in calculating the overall article metrics.
EDIT ==============
Of course, there's no real way to bring current articles from the current voting system to the one I propose....45 ACP - because shooting twice is just silly
-----
"Why don't you tie a kerosene-soaked rag around your ankles so the ants won't climb up and eat your candy ass..." - Dale Earnhardt, 1997
-----
"The staggering layers of obscenity in your statement make it a work of art on so many levels." - J. Jystad, 2001
modified on Friday, February 12, 2010 7:40 AM
|
|
|
|
|
John,
one of my early posts in the suggestions forum was to replace the 1-to-5 vote for articles by a set of 1-to-5 votes, covering different quality aspects like you suggest now. That was long before a message became 'necessary' to enter a 1- or 2-vote.
The basic idea was the overall score would be the average of the votes entered, each of them being optional. So you could up-vote the text of the article, not vote the formatting, and down-vote the accuracy, offering some real feedback without really entering any textual reply at all.
The idea wasn't picked up then.
Luc Pattyn [Forum Guidelines] [Why QA sucks] [My Articles]
I only read code that is properly formatted, adding PRE tags is the easiest way to obtain that. All Toronto weekends should be extremely wet until we get it automated in regular forums, not just QA.
|
|
|
|
|
John Simmons / outlaw programmer wrote: Maybe the article's score could be calculated using the number of views compared to the number of users who marked those checkboxes.
FYI, I'm pretty sure the popularity of an article already uses the number of views as part of the calculation. It uses the number of views and the article rating, so viewing the most popular article will get you pretty close to what you describe.
|
|
|
|
|
I've wondered if something are clueless and think 1 is good despite the up and down thumbs. I know various finger gestures mean different things in different cultures.  You measure democracy by the freedom it gives its dissidents, not the freedom it gives its assimilated conformists.
|
|
|
|
|
Just for a fair debate, I've personally seen articles get tons of 5s from people who vote without reading the entire article and making sure it's all accurate. Then someone comes along who recognizes that the article has several flaws and down-votes it. At this point a casual observer would see several 5s and a solitary 1 and would think the 1 vote is unfair (when truth would be the opposite). Just playing devil's advocate here.
And because 1 votes are non-anonymous, some people vote a 3 so they won't have to risk getting revenge votes on their articles.
Me personally, I now ignore low votes I get - in fact I usually ignore 5 votes too. I do reply to comments on my articles whether good or bad, but the votes - they are just one indicator of whether people find the article useful or not. Another very good indicator would be the view count - articles with a high view count are usually good enough for most people. And then comments - they tell you how people are using it.
|
|
|
|
|
Nishant Sivakumar wrote: Then someone comes along who recognizes that the article has several flaws and down-votes it.
If he takes the trouble reading the article and finding problems/shortcomings, I'd expect him to vote and explain his vote in a message. I still believe a message should be a prerequisite for all kinds of votes, marks or whatever expression of appreciation.
And of course you are right, voting isn't all that important, messages are much more relevant than numbers or flags.
Luc Pattyn [Forum Guidelines] [Why QA sucks] [My Articles]
I only read code that is properly formatted, adding PRE tags is the easiest way to obtain that. All Toronto weekends should be extremely wet until we get it automated in regular forums, not just QA.
|
|
|
|
|
Luc Pattyn wrote: If he takes the trouble reading the article and finding problems/shortcomings, I'd expect him to vote and explain his vote in a message.
Yeah but I don't think everyone would want to do that. Sometimes a 1 is not meant to help the author, it's meant to warn other people to stay away
Also many authors react *very* negatively to criticism. While they all ask for "feedback", they usually only want to hear that their article rocks - anything else results in lots of fuss and complaining!
|
|
|
|
|
Nishant Sivakumar wrote: many authors react *very* negatively to criticism
That is too bad, but it will not stop me, nor will it make me agree with a voting system that only allows for positive judgement.
If I feel the need to provide feedback, I will; and if it happens to contain criticism, then so it is. Every honest comment is an opportunity to learn, both for the author and for other readers. Whoever does not understand that, has a problem he should take care of. If the author gets angry, I might use different words a next time; if need be, I'm quite capable of using positive terms only and still tell others it is not OK at all! And if he acts like a jerk, I'll have no trouble ignoring him, as others would do too.
Luc Pattyn [Forum Guidelines] [Why QA sucks] [My Articles]
I only read code that is properly formatted, adding PRE tags is the easiest way to obtain that. All Toronto weekends should be extremely wet until we get it automated in regular forums, not just QA.
|
|
|
|
|
Whatever the reason is for the "1" vote, I think it is important to provide feedback.
Why?
So that the author can improve himself. I mean, if that happened to me, which hasn't yet happened, I would most likely look for feedback on what I did wrong. So, I can improve it next time. I mean, we are all here to help each other, right?
If you don't want to give feedback, might as well not vote. You are not helping in any way. Just making it worse.
I think if a Vote of 1 is provided, the page should ask whether the voter wants to provide feedback.
Thanks,
Harsimran Singh1) The good news about computers is that they do what you tell them to do. The bad news is that they do what you tell them to do. - Ted Nelson
2) Why is it drug addicts and computer afficionados are both called users? - Clifford Stoll
3) The real danger is not that computers will begin to think like men, but that men will begin to think like computers. - Sydney J. Harris
(Computer code: 00001111 - translation: Hello! :P )
REMEMBER:
"Computers are made for us, we are not made for th
|
|
|
|
|
Nishant Sivakumar wrote: Another very good indicator would be the view count - articles with a high view count are usually good enough for most people
Yes, that usually points to people coming back to reread the article more than once. But what I would like to see is a download count. How many people actually download the code after reading the article would seem to me to be a better metric.
The best metric, if it were even possible to track, would be a count of how many applications actually used the code.
|
|
|
|
|
Since you guys recently asked in the lounge about improving QA, I just had an idea:
Questions that received an answer that is not sufficient kind of down when they are neither "very active" nor "recent". That's a any similar sites (if they'd exist at all ) would have - how to distinguish an abandoned from an unanswered question?
Suggestion: The user asking can signal (e.g. link or checkbox) s/he is "still waiting for input", which would include the question in the "unanswered" list (or a separate category) - anythign that signals helpers "input wanted". This state expires automatically after a given time - maybe 24 hours. Something that is short enough so abandoned questions don't clog the system, but long enough not to annoy askers.
Also, definitely "friendlier" than a "reject answer" I thought of first.
I'm not sure if this has been tried before and if it would work as intended, some mechanism against scripters might be required (captcha?), possibly some other tweaking, too. Like, clicking the link could be replaced by ending in some beer and hamster food.
What do you think?
|
|
|
|
|
Perhaps allow them to "bump" the question (i.e., move it to the top of the questions list) without actually modifying the question? Could be done on an exponential scale, to prevent abuse. So, the first bump could be done after 24 hours, the second after 2 days, the third after 4 days, and so on. Maybe combine that with your 24 hour window. So, when it gets bumped, it gets temporarily placed in the unanswered questions list, for 24 hours.
Or...
Currently, the answer count shows the number of total answers to a question. Allow the user to accept/reject answers. When an answer is added, it counts +1 toward the number of answers. When it gets rejected, it counts -1 toward the number of answers. So, if the author rejects all the answers, the question gets added back to the unanswered questions list. So, answers could have 3 possible states: unattended, accepted, and rejected. Unattended and accepted answers would count 1 toward the answer count, and rejected answers would count 0 toward the answer count. Would probably want a good way to visualize rejected answers, so those answering questions don't think the question was already satisfactorily answered.
|
|
|
|
|
You are right, that's identical with "bumping". 
I prefert my wording, though, it's less childish-compulsive, more professional  Also, the way I suggested it, makes it more "ON/OFF". However, for a list "no answers or still looking for answer", you DO need a sort condition, I'm not sure that "Date posted/bumped" is the best solution.
Reject Answer: I wouldn't do that. It builds tension, it's a personal attack: "I reject your answer". "I still need more information" is not directed at the user trying to help. Should I really reject an answer that is actually ok but doesn't apply to me because of a constraint I forgot to mention? Do you know what happens when someone can't tell a GIT from a ROT, but rejects a JSOP answer?
|
|
|
|
|
Interesting.
What about really, really poorly worded questions? What if we only allowed a question to be bumped if it had a score over 3.0 (or no score)?cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
The nasty details are left as an exercise for the implementer, of course.
A "poorly worded" question (nice way to put it ) probably will be voted down especially if bumped repeatedly. So yeah, allow bumping unvoted questions, but not badly voted ones might work fine.
|
|
|
|
|
I've been thinking about this more and I worry that this feature will only be used by a very small set of members.
Those who ask good questions typically take care in asking, and those questions will typically get an answer if an answer is possible.
Those who ask, ahem, poorly worded questions often do not take the care necessary. They are then also unlikely to take the care to bump their question. More likely they will just repost.
The remaining questions that will get bumped? Reasonably posed and very tricky.
Hmmm. Actually...
Maybe this would be a good idea...
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
I'm using Chrome, I like the fact that your profile page no longer has plenty of space in between each of your profile links. However, the job title and the rest are still too far below.
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
