Ask HN: What did you work on in 2017?

We wanted to take our own shot at building a reliable open service with new tech. We're also much more focused on providing open and easy access to the data we have as conveniently as possible, so that people can build cool stuff with it. Right now, we provide a Google BigQuery dataset [1] that you can run any queries on. It has around 70M listens from 1300 users.

[1] https://blog.musicbrainz.org/2017/08/05/listenbrainz-data-is...

Thanks. I've learned to accept that I can't do anything about most of what's happening around the world but I can do something about my daily decisions.

If you’re an alcoholic and drink daily, it can be very dangerous to stop cold turkey. If this is more of a lifestyle change, where you drink a few times a week but want to cut it out, I’d say (with zero scientific justification whatsoever) it’s sort of up to the person. Dietary and habitual changes for me have been easiest when made dramatically, but I know for many others that reduces quality outcomes. This is probably one of those scenarios where you need to know which side of the spectrum works best for you from past experiences.

How much do you drink a day? a week?

What made you want to stop?

I can help you out based on your responses.

It's doable, the first few days can be rough for the body but after that it's mostly in our heads. I've learned that people who are enthusiastic about drinking and getting others drunk usually have some other issues going on under the surface. Some people can just drink a few times a year or month, not me. I'm an all or nothing person, life got easier once I accepted that.

As for how I did it, it boils down to just making two columns on a sheet of paper and listing how alcohol has helped and how alcohol has hurt you. It became an easy decision after I did that.

It can't do a lot. The language 'specification' is just a buch of lined paper sheets. What I want my language to do, is a layer between the core of a C++ program, and the user-customizable side.

I'm using it to call a zoom function in my skeleton image program.

  void toysanta() {  
    toy	*nbzoom = nullptr;  
    nbzoom = isVerb(L"setzoomlevel");  
    if(nbzoom) {  
        if(nbzoom->hasClaus && nbzoom->sled == toy::integer) {  
            bool	k = false;  
	    POINT	pt = {0,0};  
            toy *xt = getToy(L"x");  
            toy *yt = getToy(L"y");  
            if(xt && yt && xt->canNumber() && yt->canNumber()) {  
                pt.x = xt->getInteger();  
                pt.y = yt->getInteger();  
                k = true;  
            }  
            setzoomlevel(nbzoom->claus.integer, k, pt);  
        }  
    }  
  }

The toysanta() member function is from the VerbCommand class. From the zoom widget, there's

  void ZoomWidget::sendzoom_to_target(bool k, POINT pt) {  
    if(target) {  
        std::wstring	cc = L"setzoomlevel ";  
        int		zoom = realZ;//(zPos-32)/3;  
        cc += std::to_wstring(zoom);  
        if(k) {  
            cc += L" x " + std::to_wstring(pt.x);  
            cc += L" y " + std::to_wstring(pt.y);  
        }  
        target->receive(&cc);  
    }  
  }

The receive() member function calls the toysanta() function from the target.

What I plan, is for it to just be perfect, to be honest. I hope it may help others create better software one day.

I have worked on some for the Rift, the Gear and the Daydream. I think the reason why not a lot of people are in the space is that customers aren't there yet. It's one of those things that doesn't live up to the hype, but it is getting better fast.

I don't mind going against the current, I'm a contrarian by nature. :)

I hope you find it useful, and good luck in 2018.