3 ms·

So, are we seeing a fall of microsoft from the developers' point of view? Are windows phone and asp.net mvc not enough to try to get back the hackers in the pla

by yulaow 13y ago

So, are we seeing a fall of microsoft from the developers' point of view? Are windows phone and asp.net mvc not enough to try to get back the hackers in the platform?

Phlarp 13y ago
This was my interpretation, the largest downward movers across the board were the Microsoft stack technologies. Perhaps, given this data is from a Q and A site, these languages have just gotten so amazing in the last three years that developers are collectively asking fewer questions because they're running into fewer critical / main sequence blocking issues. Maybe these languages have moved to only being deployed by established and professional shops, leading to less of the entry level developer "noise" questions on SO [e]: sibling comment explains these trends, quite convincingly, as site demographics. I'd love to see this data compared to a similar set from github. Although public repo stats would suffer the same selection bias as SO posts.
throwit1979 13y ago
No. SO was founded by two windows developers, and build on a windows stack. They had a podcast detailing the process of development prior to public release of the site, followed primarily by windows developers. The initial audience of the site was, ergo, primarily windows developers. As popularity grew, more developers from the broader community showed up, driving the proportion of windows questions down.
- jaredmcateer 13y ago
  Also as the knowledge base of questions for a tag grows, room for new questions become fewer. Duplicates will be asked but I wonder how many times a person has a problem, does a search, finds an answer and never has to ask their own question on the topic.
  - viraptor 13y ago
    Unfortunately breakdown for the close votes is not available like for the flagged answers. But my experience with the review queue is that 1 in every 4-5 votes is due to a duplicate. Considering there are 58.1k close votes to review at the moment... that's loads of dupes.
vyrotek 13y ago
No. Stackoverflow goes though these "land grab" phases with languages and frameworks. Because C# was one of the early languages it is very difficult to keep up inertia of new questions. My personal experience has been that I rarely need to ask new C# or MVC related questions these days. I really have a hard time coming up with something has hasn't been discussed already. I think Stackoverflow eventually hits a saturation point and people begin to consume more than contribute.
- eterm 13y ago
  It has got to the point with SO that if a question is obscure enough that there isn't already a question for it, I know when I ask I'll likely not get an answer.
nwilliams 13y ago
Tags on new questions is a pretty bad way to measure that. SO traditionally had a huge .NET audience, so many questions are already answered. Total number of questions tagged would be a somewhat better metric. An even better way would be to see the views on each question. Probably the best metric publicly available would be the trend of the number of job openings on Dice or Indeed. Anyone got that data?

by yulaow 13y ago

So, are we seeing a fall of microsoft from the developers' point of view? Are windows phone and asp.net mvc not enough to try to get back the hackers in the platform?

Phlarp 13y ago

This was my interpretation, the largest downward movers across the board were the Microsoft stack technologies. Perhaps, given this data is from a Q and A site, these languages have just gotten so amazing in the last three years that developers are collectively asking fewer questions because they're running into fewer critical / main sequence blocking issues. Maybe these languages have moved to only being deployed by established and professional shops, leading to less of the entry level developer "noise" questions on SO [e]: sibling comment explains these trends, quite convincingly, as site demographics. I'd love to see this data compared to a similar set from github. Although public repo stats would suffer the same selection bias as SO posts.

throwit1979 13y ago

No. SO was founded by two windows developers, and build on a windows stack. They had a podcast detailing the process of development prior to public release of the site, followed primarily by windows developers. The initial audience of the site was, ergo, primarily windows developers. As popularity grew, more developers from the broader community showed up, driving the proportion of windows questions down.

jaredmcateer 13y ago
Also as the knowledge base of questions for a tag grows, room for new questions become fewer. Duplicates will be asked but I wonder how many times a person has a problem, does a search, finds an answer and never has to ask their own question on the topic.
- viraptor 13y ago
  Unfortunately breakdown for the close votes is not available like for the flagged answers. But my experience with the review queue is that 1 in every 4-5 votes is due to a duplicate. Considering there are 58.1k close votes to review at the moment... that's loads of dupes.

vyrotek 13y ago

No. Stackoverflow goes though these "land grab" phases with languages and frameworks. Because C# was one of the early languages it is very difficult to keep up inertia of new questions. My personal experience has been that I rarely need to ask new C# or MVC related questions these days. I really have a hard time coming up with something has hasn't been discussed already. I think Stackoverflow eventually hits a saturation point and people begin to consume more than contribute.

eterm 13y ago
It has got to the point with SO that if a question is obscure enough that there isn't already a question for it, I know when I ask I'll likely not get an answer.

nwilliams 13y ago

Tags on new questions is a pretty bad way to measure that. SO traditionally had a huge .NET audience, so many questions are already answered. Total number of questions tagged would be a somewhat better metric. An even better way would be to see the views on each question. Probably the best metric publicly available would be the trend of the number of job openings on Dice or Indeed. Anyone got that data?