|
I posted a reply to Pete on this thread and didn't want to repeat myself in Nish's thread.
Is this discussion useful? You have access to a lot more information than any of us!
Nick
|
|
|
|
|
Nicholas Butler wrote: The voting on articles works well in the sense that most articles have enough
votes to indicate a general consensus. This is a really useful indicator.
Actually, I disagree. In the old days of the voting system, it was common for some authors to vote 1 to push articles out of the "latest best picks" articles so that their articles would be at the top. Now, a single vote of 1 is enough to push an article out of site, and consequently it never gets read.
Also, your approach presumes that there are enough selfless people willing to vote 5 to compensate. One of Nish's recent articles was viewed over a 1000 times before anybody bothered to vote at all; something that greatly irritated me because the article was an outstanding one. The standard deviation filter is useless in cases like this.
So, I vote no to the anonymous voting.
|
|
|
|
|
Pete O'Hanlon wrote: Actually, I disagree. In the old days of the voting system, it was common for some authors to vote 1 to push articles out of the "latest best picks" articles so that their articles would be at the top. Now, a single vote of 1 is enough to push an article out of site, and consequently it never gets read.
It is annoying when someone votes on an article for reasons other than the merit of the article, but that's just the way of the world. People do all sorts of strange things for strange reasons. When I write articles, I accept that this is going to happen to a certain extent, but my experience is that most people vote on merit and that the overall score is a pretty good indicator of quality.
When mandatory comments were introduced, it reduced the number of low votes and I think that is a bad thing. It reduced the scale from 1-5 to 4-5, which made it harder to differentiate articles by quality and also skewed the ratings of all new articles relative to ones that were rated under the previous rules.
Pete O'Hanlon wrote: Also, your approach presumes that there are enough selfless people willing to vote 5 to compensate. One of Nish's recent articles was viewed over a 1000 times before anybody bothered to vote at all; something that greatly irritated me because the article was an outstanding one. The standard deviation filter is useless in cases like this.
Again, my experience is different. I wouldn't call people who vote "selfless", I would call them polite. It takes a lot of effort to write a decent article and I think the least a reader can do is form an opinion and press a button. The problem has always been that there is not enough incentive for the reader to vote. Perhaps this needs some attention, but that's another discussion. The point here is that making it harder for people to vote is exactly the wrong thing to do.
If you look at existing articles, most have about 1 vote per 1,000 views. That's not great, although I don't know how many of the "views" are by site spiders. It might not be immediate, but over time a consensus is reached and with the new standard deviation filter ( a fantastic idea ), any anomalous votes are discarded. I think this is more valuable than the risk of being knocked off the "Latest Best Picks" by a single low vote. Even then, it doesn't take many 5's to filter out a 1-vote.
Pete O'Hanlon wrote: So, I vote no to the anonymous voting.
I vote for anything that increases the number of votes cast, whether up or down. Mandatory comments put people off voting and also cause retaliatory voting - which can result in skewed scores for the targeted articles.
No system open to the public will be perfect, but I think mandatory comments cause more problems than they solve.
Cheers,
Nick
|
|
|
|
|
Nicholas Butler wrote: It might not be immediate, but over time a consensus is reached and with the new
standard deviation filter ( a fantastic idea ), any anomalous votes are
discarded
While the sd filter is a good idea, it only works if an article gets sufficient votes. My point is that a 1 vote can push a new author so far down the ratings that their article is never read; and this means they won't have a chance to recover. Consider how offputting this is for a new author who has spent an age slaving over an article that they are rightfully proud of, (and no, it might not be the greatest quality, but at least they have had the guts to have a go), only to see one selfish vote demolish their article.
In this, I think that mandatory comments have been a good thing. More importantly, they give other users a chance to disagree with the vote (and in cases where it's obvious that the 1-vote has no value, they can vote to remove the vote). Again, this is a positive thing.
I say, let's keep it with the mandatory comments. They do serve a purpose (even if it is only to identify HWSNBN's sock-puppet accounts).
|
|
|
|
|
Pete O'Hanlon wrote: My point is that a 1 vote can push a new author so far down the ratings that their article is never read
I don't know the stats about how articles are found and read, but I suspect being on the "Latest Best Picks" isn't as important as you assume.
The list of all new articles is on the front page, there is an RSS feed of all articles and CP articles rank pretty highly on Google.
Nick
|
|
|
|
|
Pete O'Hanlon wrote: a 1 vote can push a new author so far down the ratings that their article is never read
With the volume of articles that come through, the listings on the homepage really are only there as a way of saying thank you to the top authors: they don't influence readership, long term, that much because most views of an article come via searches, through the newsletter, or through direct links.
The complaint here is that top articles get kicked off the homepage due to a single downvote.
I have an idea...
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
Pete O'Hanlon wrote: My point is that a 1 vote can push a new author so far down the ratings that their article is never read
I've made a change so that the top 5 articles on the homepage are now ordered by number of upvotes, and not by actual score. This means that the set of articles on the homrepage is now at the mercy of vote-stuffers, but are now less likely to be affected by spurious down voters.
We'll see how this goes.
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
Nicholas Butler wrote: When mandatory comments were introduced, it reduced the number of low votes and
I think that is a bad thing
In my experience, it mostly reduced the malicious and immature down-votes. Sub-par articles still get down-votes and from regular/senior members.
Nicholas Butler wrote: If you look at existing articles, most have about 1 vote per 1,000 views.
The views don't really make for a good stat unless you know how many of those were from guests and how many were from members. Guests can view, they cannot vote or comment. I requested a new feature to see separate veiw-counts for this here[^]
Nicholas Butler wrote: It might not be immediate, but over time a consensus is reached and with the new
standard deviation filter ( a fantastic idea ), any anomalous votes are
discarded. I think this is more valuable than the risk of being knocked off the
"Latest Best Picks" by a single low vote. Even then, it doesn't take many 5's to
filter out a 1-vote.
The way the site receives tons of new articles these days, an article gets visibility for about 2-6 hours after it's posted. After that it goes off the front page and there is a steep decline in view count. The really good articles (usually when authors have put enormous effort into it) make the Top-5 list thus giving them more visibility for a few more days (often up to a week). It's these articles that are adversely affected when a malicious 4 is cast to knock it off the front page. Once that happens, it's extremely unlikely that enough people will view the article for enough 5s to be cast to kill the effect of the 4 vote (assumption here is we are talking about a high quality article that deserves front page attention).
At the personal level this is quite frustrating to the authors. Talk to Pete or Sacha or Marcelo and you'll see that they've all experienced this many times here. From CP's and thus Chris's perspective good/interesting articles that improve the utility value of the website are being knocked into oblivion by one single 4 vote. I am sure Chris will want to take steps to prevent that.
Hope this made sense.
|
|
|
|
|
Hi Nish - thanks for your reply 
If the "Latest Best Picks" isn't working, that is a separate problem to anonymous voting, surely?
If an author down-votes someone else's article to promote his own, then his article will attract more attention. If his article is not most excellent, it should be voted off the list by people voting properly - on merit.
I think the underlying problem is that not enough people vote properly. Mandatory comments exacerbate that.
Nick
|
|
|
|
|
Nicholas Butler wrote: If the "Latest Best Picks" isn't working, that is a separate problem to
anonymous voting, surely?
It's related. The issue with best-picks surfaces only when someone wrongly casts a vote to knock an article off the list.
Nicholas Butler wrote: If an author down-votes someone else's article to promote his own, then his
article will attract more attention. If his article is not most
excellent, it should be voted off the list by people voting properly - on
merit.
It's not just authors who do this. Sometimes people just do it because they don't like a particular author (mostly because of completely unrelated comments the author may have made in say a Lounge thread on politics or sport or some such thing).
Nicholas Butler wrote: I think the underlying problem is that not enough people vote properly.
I agree to that. I don't agree that this is because of mandatory comments (btw you don't need to actually write a comment, the site auto-posts a message with a default subject of my vote of X).
There needs to be better incentives to voting.
|
|
|
|
|
Isn't this supposed to be polled every hour? It's running a bit late.
|
|
|
|
|
I've given it a kick - please let me know if you don't see anything within a few hours.
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
I can find the reference link to the article "Low Latency Audio using ASIO Drivers in .NET" by either searching for the article or by navigating via Multimedia > Audio and Video > Audio .
In either case, I find the link to this article, but when I click on it, I get a "Unfortunately the page you requested was not found" message.
This message occurs from within the Code Project environment.
I ask that someone else verify this bug so I can rule out any problem specific to my system (unlikely, I think).
If this is a real bug, I ask that a site administrator look into this. Thank you.
Logan
|
|
|
|
|
This article: Low Latency Audio using ASIO Drivers in .NET[^]? It's working fine for me.
If the request fails again, could you please look at the very bottom of the screen, left hand side, for "WebXX". This will tell you which server you're on which could help (though from what it's telling me, all servers are currently in sync)
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
Chris,
Thank you for your reply. I am still consistently unable to access this one article. Every other article I link to works perfectly. In fact, I was able to access the article in question several time before it stopped working for me. Very curious!
My latest attempt utilized WEB04 server and several more attempts using WEB22 server.
FYI ... I am using Firefox on XP SP3.
Regards,
Logan
|
|
|
|
|
Logan Durst wrote: FYI ... I am using Firefox on XP SP3.
What version? I just tested FF 7.0.1 and can access the article in question without any problems. Can you navigate to the URL via the link Chris posted?
|
|
|
|
|
Randor,
I too am using FF 7.0.1. Chris's link would not work for me until I cleared history, cookies, etc. It works fine now.
Regards,
Logan
|
|
|
|
|
Chris,
I did some experiments and found that if I try to access the ASIO article via "multimedia > audio and video > audio", I am directed to the address "www.codeproject.com/script/common/404.aspx?aspxerrorpath=/kb/audio-video/asio_net.aspx", which does not work.
If I then modify the address to be www.codeproject.com/KB/audio-video/asio_net.aspx", I can then access the article successfully.
Could it be that my browser is being hijacked?
Edit:
Just discovered that by clearing my Firefox history and cookies, everything works perfectly. Clearly the problem is exclusively on my end. Sorry for the false alarm.
Regards,
Logan
|
|
|
|
|
Thanks for the update, Logan.
www.codeproject.com/script/common/404.aspx?aspxerrorpath=/kb/audio-video/asio_net.aspx
Is the "Page not Found" handler, meaning that when you clicked on the correct link it simply sent you straight to the "Can't find it page". As to why that would happen before, and not after, you cleared your history is a total mystery because your browser should know nothing about what to do in case of a missing file redirect. My only guess is that, once, the file could not be found due to a network issue or a file sync issue our end, and your browser stubbornly cached the information that "This link goes to that 404 page" without bothering to have another actual check on the state of things.
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
Check this users article list; http://www.codeproject.com/Articles/yaxW[^]
One of his articles is posted twice, in 2 different sections, and you can see they have 2 different view counts.......
[edit: one of them is actually in a composing state?? and is basically empty, and just needs deleting]
modified 15-Oct-11 15:40pm.
|
|
|
|
|
All sorted.
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
I created this question[^] to try and add ASP.NET to the Q&A tags but is still does not take. There must be something odd about this tag.
|
|
|
|
|
I wonder if periods are disallowed in tag names.
Somebody in an online forum wrote: INTJs never really joke. They make a point. The joke is just a gift wrapper.
|
|
|
|
|
Probably not as there is tags like C#3.5 and .NET4 available.
0100000101101110011001000111001011101001
|
|
|
|
|
Should be good now. Was choosing the set of attributes from the wrong site.
cheers,
Chris Maunder
The Code Project | Co-founder
Microsoft C++ MVP
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
