A few voting statistics

With all of the, umm, debate over voting, you might find this interesting(or, you might not).

What I did was take the top 100 items and simply figured the average number of votes cast by each voter for that item. So if an item had two voters, and one cast 20 votes and the other 10 votes, the average would be 15.

I then totaled up the numbers and calculated the total average, median average, max average, and min average for the top 100 items, then the top 50, top 20, and top 10.

And, here is what I got.


Now, for some real fun, lets look at the top 20 items(I shorten a few titles to hopefully make it fit):


Funny thing is, the average votes for Maven(33.70) and AspectJ(31.31) are not that much different. Ironically, just above Maven is an item that sticks out like a soar thumb(with an averge of 58.90), yet has never been brought into this debate of voting(especially with the implications of ballot box stuffing being cast about so freely).

Finally, I am not a supporter of Maven(never used, and, at this time, likely never will). I am not a supporter of Clear Case. And, I am not necessarily against either.

But, in my personal opinion, if we are going to discuss items, imply wrong doing, and such, let's at least try to keep it fair.

For what it is worth, just a different perspecitive on things.

0

Jetbrains doesn't seem to use # of votes as a todo list either. they seem to view it as suggestions. nobody ever requested a C# version, or if they did it never had too many votes, but guess what, they are building it :)

0

I know, but there has been a lot of wailing and gnashing of teeth over features and voting and such. :)

And, there have been, umm, concerns about voting irregularities and such. :)

So, I took a few minutes to thorw together a few numbers, to, hopefully, put things into perspecitve.

But, I agree. I think they get ideas from the suggestions and work on the stuff they want to work on, to match their vision.

I just hope they don't loose sight of the fact that it is the small things that can make the real difference. The small, commonly used features and convieniences are just as important as the big press/ticket items.

Plus, that plugins are also important. It was because of some of the plugins created by others that I decided to used IDEA; along with some of its features.

0

These stats are pretty interesting, I'm sorry that I left out a thank you for doing that...sorry if it sounded ungrateful :)

it's funny (to me at least) I know nothing about maven, but because of all the fuss over it I don't like it :O

0

No thanks are needed; and I didn't think you sounded ungrateful. I did it because I wanted to...was a question that popped into my head. Figured I might as well share the answer.

And, I thought you brought up a valid and interesting point, in how Jetbrains appears to use the voting.

0

I thought that I'd chime in here since I'm one of the folks who voted
for Clearcase support (and dumped all 100 points into it).

The simple reason is that I really want decent VCS support. It amazes
me that Subversion has a higher total given that it's still an alpha,
but I assume that this is a side effect of it being free. (Not that
I've got anything against Subversion, I've heard nothing bad about it,
but I've also never used it).

I saw no point in voting for generics support because frankly Jetbrains
will have to do it regardless of what I vote simply to stay in the game,
and likewise I won't be able to do anything with it for a long while.

As for AspectJ support, well I've yet to meet anyone who's actually used
it. From what I've read I can't see that it would help me very much
since our project is well underway and many of the problems that it
appears to solve, we've already handled otherwise. I should probably
look into it more.

So for me I find that IntelliJ does what it does really well, and the
problems come when I'm relying on a plugin, every time there is a new
version it appears that the plugin API changes and finds a new and
exciting way for the Clearcase plugin to go wrong. VCS handling is
central to managing code, so when it's doesn't work well there is no way
to avoid noticing.

Aside from that. Thanks for the stats, they really are very interesting.

Guy

zmbs wrote:

With all of the, umm, debate over voting, you might find this interesting(or, you might not).

What I did was take the top 100 items and simply figured the average number of votes cast by each voter for that item. So if an item had two voters, and one cast 20 votes and the other 10 votes, the average would be 15.

I then totaled up the numbers and calculated the total average, median average, max average, and min average for the top 100 items, then the top 50, top 20, and top 10.

And, here is what I got.

               TOP    TOP    TOP    TOP
>  Votes        100     50     20     10
> ---------    -----  -----  -----  -----
> Average      29.99  30.63  29.31  29.33
> Median       26.47  27.43  27.69  29.54
> High         76.25  76.25  58.90  58.90 
> Low           5.09   7.27  12.69  12.69
> ]]>


Now, for some real fun, lets look at the top 20 items(I shorten a few titles to hopefully make it fit):

                                                     Total  # of    Avg.
>             Request Title                           Votes Voters  Votes
> AspectJ Support                                      3131   100   31.31
> Update IDEA's parser to support JSR14 - Generics     1964    65   30.22
> Reload class and continue to debug during debug...   1852    81   22.86
> Editor: Split function should be able to display...  1610    93   17.31
> subversion integration                               1241    43   28.86
> CVS "show history" as Local VCS to use "diff"...     1183    35   33.80
> Need adequate ClearCase support                      1178    20   58.90
> Maven Integration                                    1011    30   33.70
> Provide ability to record and play macros             787    62   12.69
> Make diff tool SMART                                  779    33   23.61
> IDEA's CVS integration is bad !!!                     753    16   47.06
> Support for RMI Compilation in IDEA                   742    28   26.50
> Tighter J2ME integration                              679    26   26.12
> Reformat code should be able to reorder methods       594    40   14.85
> If a ModificationAttemptListener changes the...       573    22   26.05
> Move method refactoring                               570    18   31.67
> add vi keymap                                         554    19   29.16
> add SSPI access method for CVS                        530     9   58.89
> support JSP reformat code.                            505    30   16.83
> Package View and Hide Empty Middle Packages           504    32   15.75
> ]]>


Funny thing is, the average votes for Maven(33.70) and AspectJ(31.31) are not that much different. Ironically, just above Maven is an item that sticks out like a soar thumb(with an averge of 58.90), yet has never been brought into this debate of voting(especially with the implications of ballot box stuffing being cast about so freely).

Finally, I am not a supporter of Maven(never used, and, at this time, likely never will). I am not a supporter of Clear Case. And, I am not necessarily against either.

But, in my personal opinion, if we are going to discuss items, imply wrong doing, and such, let's at least try to keep it fair.

For what it is worth, just a different perspecitive on things.


0

It amazes me that Subversion has a higher total given
that it's still an alpha.


Subversion is supposed to be the successor of the current nr. 1: CVS. Many of the original CVS developers have started this new project to fix the problems inherent to CVS' architecture. As such, it is understandable that many people want to test it. IMHO, we really need a comprehensive plug-in API for version control (which is also used for CVS support). Plug-ins like these should probably be open source, even when they are developed by Jet Brains. This allows new plug-ins to be derived from existing and will enable users to fix bugs. That will reduce the load on IDEA's developers and will reduce VCS feature requests.

I saw no point in voting for generics support because
frankly Jetbrains will have to do it regardless of
what I vote simply to stay in the game.


IMO that is a major reason why Generics has so few votes.

So for me I find that IntelliJ does what it does
really well, and the problems come when I'm relying
on a plugin, every time there is a new version it
appears that the plugin API changes and finds a new
and exciting way for the Clearcase plugin to go wrong.


Unfortunately, it seems that plugin API's are not IDEA's strongest point. Jet Brains should use it more often themselves probably, although it will reduce their coding speed. Still, I think that it is necessary since they cannot develop all the requested features themselves.

0

请先登录再写评论。