Stepping down as a maintainer of SciRuby/daru

[Victor Shepelev]

Hi all.

This is an official notification, that I can not anymore consider myself as a part of SciRuby initiative, or a maintainer of daru or any other libraries.

I don't want this event to be seen as anybody's personal fault.

It is a result of a long-thought and hard for me decision.

It was awesome several years together, and I am grateful to all colleagues and students I met online and offline and had an opportunity to work together.

I assume I owe you at least some explanation.

As some of you may remember, my level of participation dropped significantly circa summer 2017: I did a lot of preparatory work for GSoC that year and was a designated mentor for formally one, but actually two projects; but then, around mid-summer, it became really hard for me to maintain pace, and I want to sincerely apologize to Athitya and Shekhar, who both are brilliant students and, fortunately, succeeded with the projects nevertheless.

The main reason of my detachment of that time was purely personal (clinical depression), that's why when my condition got better at autumn, I tried to make up for the lost summer, contribute to Daru refactoring and further development, but mostly just made plans and promises.

The thing is, after several (personal) events of that year, I had to review and rethink a lot of stuff and to understand what is important and what I should do with my life. I still had hopes to somehow integrate my SciRuby work into the new picture (that's why I participated in GSoC 2018, with unsatisfactory results), but it just didn't work.

Thing is, I self-identify with all my OSS work a lot, and it just became impossible for me to deeply self-identify with SciRuby goals. I do not want to discourage those who are staying, so to say, but it seems clear that wider Ruby community is not adopting SciRuby achievements, and all the work stays as a "thing in itself". Maybe my step is just an act of cowardice (avoiding all the hard work that should be done to "catch with other languages"), but I currently want to dedicate my free time and passion to my own work on Molybdenum/Reality[1] project, and to general Ruby language evolution[2].

I want to express deep respect to all old and new participants of the community, and deep regret for all the false promises I made and former enthusiasm I can't feel anymore.

My special gratitude and special great apology should go to Sameer, who was the person that invited me into this community and trusted me with our common goals and libraries. It should not be considered his fault that I turned out to be not as wise and energetic person he saw in me.

Thanks for everything.

V.

1: https://github.com/molybdenum-99/reality

2: https://bugs.ruby-lang.org/users/710

[Pjotr Prins]

Dear Victor,

Just wanted to say we really appreciated your contribution - and you
made significant ones. I also very much enjoyed our interactions the
times we had video conferences with the GSoC team :). Don't beat
yourself up about leaving this project, I know your contribution
will be valuable wherever you go and really happy to see you want to
continue with Ruby in general!

Thanks again, and hopefully you'll share some of your wisdom with us
once in a while.

Pj.

[John Woods]

Victor,

Thanks so much for your contributions over the last few years. I'm sorry to hear about the depression, and applaud you for your willingness to discuss mental health publicly. We all face such struggles, and denial of that simple fact serves no one.

I do think it's worth noting that 51 people have joined sciruby-dev so far this year (suggesting we should have about 70 new by the end of the year). There are 674 people in the group, which has existed since 20 December 2011, so we expect about 96 new members per year.

While 70 < 96, I'm not sure it follows that SciRuby's contributions aren't being adopted by the wider community. It probably does follow that more scientists are getting picked up by SciPy and NumPy, given the relative popularity of Python and Ruby. 

I continue to believe that a smoothly functioning and easy-to-learn plotting library is the main thing preventing wider adoption by scientists. I myself had to stop using Ruby at work because it was taking too long to plot results of experiments.

It may be tempting to think we're "losing" to Python or to SciPy/NumPy. I think this view is flawed. When the barrier to adoption is sufficiently low (when we solve plotting), I think Ruby's many advantages will put it on excellent footing to compete with Python.

Moreover, I'm not sure it matters. Even if you're right and people decide they don't like NMatrix, daru, etc. and don't want to use them, these libraries are open source. People can learn from them/us and make better libraries, even if they don't want to reuse much of the code. (I was super happy to hear of Sameer's hiring to work on a project that will eventually replace NMatrix, for example.)

Anyway, you're always welcome here, and again — thank you so much for your contributions.

John

--
You received this message because you are subscribed to the Google Groups "SciRuby Development" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sciruby-dev...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Sameer Deshmukh]

Dear Victor,

It is sad to hear this news. I was hoping to keep working with you for much longer, but I guess all good things

come to an end. Daru really needed someone as capable as you and your absence will be greatly noticed.

However, I respect your decision and sincerely hope that your work makes a positive impact on the Ruby community.

You're always welcome to help us on any SciRuby project anytime you wish :)

All the best!