Three years and counting

[Brendan Tracey]

According to the membership list, today marks three years since I signed up for the gonum-dev mailing list. I was looking at one of my codes early this week, and a subset of the imports are

    "github.com/gonum/blas/blas64"
    "github.com/gonum/blas/cgo"
    "github.com/gonum/floats"
    "github.com/gonum/matrix/mat64"
    "github.com/gonum/optimize/functions"
    "github.com/gonum/plot"
    "github.com/gonum/plot/plotter"
    "github.com/gonum/plot/plotutil"
    "github.com/gonum/plot/vg"
    "github.com/gonum/stat"
    "github.com/gonum/stat/distuv"
    "github.com/gonum/stat/samplemv"

And for the other major code I'm working on at the moment

    "github.com/gonum/floats"
    "github.com/gonum/matrix/mat64"
    "github.com/gonum/quad"
    "github.com/gonum/stat/combin"
    "github.com/gonum/stat/distmv"   
    "github.com/gonum/stat/distuv"

That's an impressive set of tools. Not only that, but I can think of active work underway on almost all of our repositories.

When I started coding in Go, it seemed every time I wanted to explore a new idea there was yet another basic functionality without a go version. Thanks to all of our efforts, I feel that more and more rarely. We don't yet cover the full breadth of scipy/numpy, but we're a lot closer than one might think.

I'm very proud of all that we've built. Rather than take the traditional path of linking in Fortran libraries, we've instead taken a fresh look at implementations and APIs. I believe we are succeeding in having legible, performant and correct code; code that can demystify algorithms for users who care to look.

On a meta level, I'm fascinated with how we've managed to accomplish all of this. We've had contributions from at least 4 continents and at least 8 different time zones. Very few of us (if any?) knew each other before contributing to gonum. Yet, we've managed to have constructive and productive collaborative development. For me, this is a shining example of what the modern web can enable.

Thank you all for being a part of this endeavor, and I hope we continue to thrive!

[Daniel Skinner]

I'm not a contributor, but I would like to thank everyone involved for their efforts. I've used various parts of gonum to learn, confirm, and test my understanding of certain algorithms, act as a stand-in for various ops until I write something more specialized for my use-cases, and explore lots of new (to me) ideas that's served as reading material from just browsing the docs. I also actively use the plotter in place of octave for better integration testing and really appreciate that package's availability.

Thanks,

Daniel

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

[Dan Kortschak]

Thanks Brendan. I'd like to echo those sentiments. being involved in this has been very enjoyable and extremely satisfying.

Dan

[Seb Binet]

On Thu, May 5, 2016 at 9:52 PM, Dan Kortschak <dan.ko...@adelaide.edu.au> wrote:

Thanks Brendan. I'd like to echo those sentiments. being involved in this has been very enjoyable and extremely satisfying.

me too.

I have really enjoyed the ride so far: I think Go and gonum made a better s/w engineer of me.

I hope we can maintain resolve and continue to work so that eventually gonum will percolate through science fields and enable better, faster and more reproducible science.

-s

[Jonathan Lawlor]

I've to to echo what Seb said, it has made me a substantially better software engineer.  Now when I collaborate at work it feels like I'm coming down from a higher league.

re: seb's reddit post - I think gonum is reaching the point where it is useful to ease adoption, via a combination of in depth examples and some kind of manual.

When I started working with go (and in particular in gonum) my wife and I were expecting a baby, and I was thinking about how important role models are to children.  I've had some really great ones - and in particular I was thinking, "by this time Jon Bentley had made significant contributions to help other people, and all I'm doing with my life is going to work and making money."  Well hopefully baby Charlotte has a better role model now.

As an aside, I was lucky enough to know him as Scoutmaster Bentley, or Mr. Bentley, and while I knew he was a smart guy working at Bell Labs, I didn't know much more.  About 10 years later I found myself programming so much at work that I decided I needed to be good at it.  I go through book after book on programming, and I come to the well recommended "Programming Pearls" by Jon Bentley.  And I'm like, "wait, what? the guy who was extremely patient with me as a teenage smart-ass was a well known computer scientist?"  And I thought of the opportunity that I didn't even know that I had, wasted.  But I don't think it was quite so much anymore.

[Vladimír Chalupecký]

It seems to be a common effect that gonum and Go make people better software engineers. It certainly has been so in my case. gonum has changed not only how I think about numerical programming but also how I perceive certain areas of mathematics. For example, before I would have never thought that matrix algorithms could be interesting and fun. Or, thanks to reviewing for the graph package I got to study algorithms from an area that is completely outside of my expertise. And it's not only the technical aspects, our small community is just as big (if not bigger) reason why I'm glad for being involved. There still a lot of work to do and I intend to contribute as much as I can.