(j3.2006) Fortran user interest survey

Tom Clune Thomas.L.Clune
Fri Jul 8 14:31:17 EDT 2016


My immediate response is to keep it as short and sweet as possible.   I am routinely frustrated by well intentioned surveys that ask the same question in 5 different ways and for which I can only give vague responses.   And think carefully about what metadata we would actually use.   It is interesting to correlate features with orgs/vendors/disciplines, but is it actually useful for planning?     But ?discipline? would seem to be useful.

It would also be useful if some/most/all of the questions were in the form of  ?What things do you find (unnecessarily) difficult??  Rather than ?What things do you want to see?"




> On Jul 8, 2016, at 1:55 PM, Damian Rouson <damian at sourceryinstitute.org> wrote:
> 
> All,
> 
> I believe the committee voted to entertain new features no sooner than 2017.  I offered to assist with developing a strategic plan to inform new feature considerations.  I  A sample table of contents and summary from a plan I wrote is available here: http://bit.ly/29ts0Xt <http://bit.ly/29ts0Xt>.  A useful first step is surveying stakeholders. I welcome suggestions for survey content and to conduct the survey soon or closer to 2017.
> 
> I suggest including an open-ended question.  Would it also be useful to survey interest in broad concepts such as generic programming and programming by contract?  Would it be useful to include demonstrations of the form such concepts might take?  Would it be useful to have links to past papers that committee members plan to propose again?  Do we want an optional question on user characteristics such as profession, employer, market sector, job title, application domain, and platform choice?  I will collect suggested questions and place them in a template on SurveyMonkey.com <http://surveymonkey.com/> for committee review.
> 
> As I noted earlier, I detect a disconnect between user interests and what implementors report hearing from users. For example, several vendors say they have not heard interest in coarray Fortran (CAF). Yet OpenCoarrays releases have been downloaded a minimum of 936 times in the first year of release (http://bit.ly/29nkCez <http://bit.ly/29nkCez>), not including repository clones and installations via package management, which would probably add several hundred more.  Also, survey data from 110 respondents over 8 training course sites shows 75% of respondents see learning CAF as somewhat or very important prior to attending the course (http://bit.ly/29nnjgm <http://bit.ly/29nnjgm>).  Anecdotally, short-course attendees found CAF of sufficient interest that they suggested moving the CAF material from the end to the beginning of the courses.   Similarly, although the graduate course I teach at Stanford is focused on object-oriented design, I revised it in the most recent offering to use CAF features from the first lecture onward and the students gave the course strong reviews. Somehow the user interest in at least this feature isn?t reaching vendors.
> 
> Lastly, I hope a survey will illuminate user interests and help prioritize feature considerations, but I also think it?s important to lead.  Technology history is littered with examples of products that became very popular even though most users weren?t requesting or even conceiving of the product in advance. 
> 
> 
> Best Regards,
> _________________________________
> Damian Rouson, Ph.D., P.E.
> President
> <B16688D4-3FF1-4793-A7EF-AF71537900DD-1_1_1_1_1.png> <http://www.sourceryinstitute.org/>
> A California public-benefit nonprofit corporation
> +1-510-600-2991 (mobile)
> 
> 
> _______________________________________________
> J3 mailing list
> J3 at mailman.j3-fortran.org <mailto:J3 at mailman.j3-fortran.org>
> http://mailman.j3-fortran.org/mailman/listinfo/j3 <http://mailman.j3-fortran.org/mailman/listinfo/j3>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.j3-fortran.org/pipermail/j3/attachments/20160708/4a87b0ce/attachment.html 



More information about the J3 mailing list