September 29, 2011 at 11:21 pm
Comments posted to this topic are about the item Closing Out Replies
September 29, 2011 at 11:25 pm
I would say the way pioneered by http://www.StackOverflow.com is the best way, where responses are voted for, but you can only vote for a response if you have "credibility".
So then the person who finds the question later has a good indication of which the correct responses. I have found this works very well, and in fact have yet to find a question posted on Stackoverflow, which has voted answers, that didn't work as per the most voted answer!
They have a api called StackExchange which allows anyone to set up the same type of site, perhaps that should be looked into...
if you don't have the time to do it right, when will you have the time to do it over ?
September 30, 2011 at 12:17 am
Matthew Joughin (9/29/2011)
They have a api called StackExchange which allows anyone to set up the same type of site, perhaps that should be looked into...
You mean like this: http://ask.sqlservercentral.com/
The problem with that is that it's the popular answer that gets the votes, not necessarily the right one. I've seen SQL-related questions where the most votes were for a simple-looking, elegant solution that would work fine on a couple hundred rows and die a slow, agonising death on a few hundred thousand. There were answers warning about the performance implications, but they had few-no upvotes and even some downvotes (probably because they were more complex)
Gail Shaw
Microsoft Certified Master: SQL Server, MVP, M.Sc (Comp Sci)
SQL In The Wild: Discussions on DB performance with occasional diversions into recoverability
September 30, 2011 at 7:41 am
Matthew Joughin (9/29/2011)
I would say the way pioneered by http://www.StackOverflow.com is the best way, where responses are voted for, but you can only vote for a response if you have "credibility".So then the person who finds the question later has a good indication of which the correct responses. I have found this works very well, and in fact have yet to find a question posted on Stackoverflow, which has voted answers, that didn't work as per the most voted answer!
They have a api called StackExchange which allows anyone to set up the same type of site, perhaps that should be looked into...
Which of course has problems with the definition of "credible". I'm probably pretty credible on coding and performance tuning, since I do a lot of that, but I'm far from credible on a number of other things in SQL Server, and partially credible on other things.
So, does that mean I would get to "vote" only on certain posts? Would Gail's votes count more than mine on some posts (since she's definitely more credible on certain subjects)? How about Joe Celko? He's got credibility on certain things, and is dead wrong on others, even ignoring his tendency to do his best to be rude and arrogant. What does he get to vote on? How much do his votes count? Jeff Moden is highly credible on a number of SQL subjects, but I've seen him "vote", in effect, against SSIS solutions that were perfectly valid, just because he personally dislikes SSIS. Same for CLR solutions. He's come around a bit on those recently, but check some of his posts from last year even, on either of those subjects.
Which brings up, what happens if some credible expert votes for something, and five minutes later is told a better solution by someone who's completely new to the site, tests it, and finds out their solution is better? I've seen that happen numerous times.
The Stack Overflow solution has positive factors, but it also has negative ones, and some of them are eggregiously negative. There are a lot of high-vote solutions on there that are valid but far from optimum. Some because they were optimum at the time they were voted for, but no longer are, others for other reasons.
- Gus "GSquared", RSVP, OODA, MAP, NMVP, FAQ, SAT, SQL, DNA, RNA, UOI, IOU, AM, PM, AD, BC, BCE, USA, UN, CF, ROFL, LOL, ETC
Property of The Thread
"Nobody knows the age of the human race, but everyone agrees it's old enough to know better." - Anon
October 3, 2011 at 7:55 am
Matthew Joughin (9/29/2011)
I would say the way pioneered by http://www.StackOverflow.com is the best way, where responses are voted for, but you can only vote for a response if you have "credibility".So then the person who finds the question later has a good indication of which the correct responses. I have found this works very well, and in fact have yet to find a question posted on Stackoverflow, which has voted answers, that didn't work as per the most voted answer!
They have a api called StackExchange which allows anyone to set up the same type of site, perhaps that should be looked into...
Gail and Gus have some good points. StackOverflow works great in some ways, not so great in others. The answers and suggestions are disjointed, and there is a lot of information in the comments because the format doesn't lend itself well to many questions. Works great for some, not for others.
It also doesn't necessarily help someone learn, or tease out why it's a good answer. The reordering of items also makes it hard to come into the thread later and figure out what happened, especially as items get edited and comments/answers no long apply.
We are considering implementing some ideas that are similar to SO, but not it exactly. The StackExchange API isn't open, nor could we use it here. Right now there are two large projects (blog changes, Simple Talk migration) that are holding up some work, but voting/rating/etc are probably next. Likely something we do and deploy in early 2012.
October 4, 2011 at 12:13 pm
GSquared (9/30/2011)
Which brings up, what happens if some credible expert votes for something, and five minutes later is told a better solution by someone who's completely new to the site, tests it, and finds out their solution is better? I've seen that happen numerous times.The Stack Overflow solution has positive factors, but it also has negative ones, and some of them are eggregiously negative. There are a lot of high-vote solutions on there that are valid but far from optimum. Some because they were optimum at the time they were voted for, but no longer are, others for other reasons.
Well stated Sir.
The probability of survival is inversely proportional to the angle of arrival.
Viewing 6 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply