Discussion:
[Swig-devel] swigsurvey data
Olly Betts
2016-09-27 21:34:48 UTC
Permalink
The "survey" which users are asked to fill in before downloading has
been hooked up to a CGI script on my site since ~2008. I've just
migrated that site to a new server, and looking at the directory of CSV
files, there are close to a quarter of a million entries. Some are no
doubt from spam bots randomly filling in forms in the hope of ... who
knows what. But looking at recent the last month I don't see much sign
of that - IME those bots don't like leaving a text field empty, and I
don't see random junk in the text fields.

Anyway, nobody's wanted to analyse this data for at least 8 years, so I
wonder if it's time to stop asking downloaders to provide it?

I also notice it's not been updated for more recently added languages
- for example, Lua, Go, Scilab.

Cheers,
Olly


------------------------------------------------------------------------------
William S Fulton
2016-11-05 22:01:09 UTC
Permalink
Post by Olly Betts
The "survey" which users are asked to fill in before downloading has
been hooked up to a CGI script on my site since ~2008. I've just
migrated that site to a new server, and looking at the directory of CSV
files, there are close to a quarter of a million entries. Some are no
doubt from spam bots randomly filling in forms in the hope of ... who
knows what. But looking at recent the last month I don't see much sign
of that - IME those bots don't like leaving a text field empty, and I
don't see random junk in the text fields.
Anyway, nobody's wanted to analyse this data for at least 8 years, so I
wonder if it's time to stop asking downloaders to provide it?
I also notice it's not been updated for more recently added languages
- for example, Lua, Go, Scilab.
I thought I published an analysis a year or two after it first went
live, but unfortunately can't find it now. I would quite like to do
that analysis again, to see how usage has changed. Could you make the
data available? Once we have some sort of analysis we can probably
decide what to do next.

William
William S Fulton
2017-05-10 18:02:14 UTC
Permalink
Post by William S Fulton
Post by Olly Betts
The "survey" which users are asked to fill in before downloading has
been hooked up to a CGI script on my site since ~2008. I've just
migrated that site to a new server, and looking at the directory of CSV
files, there are close to a quarter of a million entries. Some are no
doubt from spam bots randomly filling in forms in the hope of ... who
knows what. But looking at recent the last month I don't see much sign
of that - IME those bots don't like leaving a text field empty, and I
don't see random junk in the text fields.
Anyway, nobody's wanted to analyse this data for at least 8 years, so I
wonder if it's time to stop asking downloaders to provide it?
I also notice it's not been updated for more recently added languages
- for example, Lua, Go, Scilab.
I thought I published an analysis a year or two after it first went
live, but unfortunately can't find it now. I would quite like to do
that analysis again, to see how usage has changed. Could you make the
data available? Once we have some sort of analysis we can probably
decide what to do next.
<snlp>
Thanks for the swigsurvey csv files Olly. I wrote a Python script to clean
up the data a bit. I've committed the scripts to the www repo and created a
survey results page which can be seen by anyone at
http://swig.org/surveyresults.html. I don't think there are any real
surprises, but I added in a bit of commentary. Does anyone else have any
interesting thoughts about it or suggestions to improve the data published?

Olly, could you send me the latest files and I can update the data? We
should try and update it regularly (well at least once a year). I'd also
like to update the list of possible target languages. Olly, you and I would
need to co-ordinate this to deal with the additional columns in the
resulting .csv files.

William

Loading...