You do have to choose which packages you want to install at some point. When you do `npm install && npm start` it will still work as planned. The problem here address is for when you need to deal with a new issue.

If it were a handful of extra kb, I'd agree. Of the top off my head I can think of following file formats "somehow" suited for configuration: json, cson, properties files (java), yaml, xml, conf-files (Apache), ignorefile syntax, ini files.

Supporting all of them could be a reasonable goal to the community (looking at the packages in npm: it is an actual goal already).

https://en.wikipedia.org/wiki/Configuration_file

Use js files for config. They support comments, syntax highlighting will work out of the box, and you can require them.

A problem that I have with my article is that readers seem to get hung up on the config loading bits when I tried to use it as an example for a more-general problem.

To fix this I added a section to the article: https://t.co/sXJhTYXQvy

(Using executable files for configuration can be an enormous problem for some uses.)

I think it's a poor example. I don't want to choose between 20 different config file formats. Give me a config file in a format I know.

Yes, the missing dependencies would show up at runtime. Which is a valid and important concern, however: It would be possible (though probably inadvisable) to implement an automatic download of missing dependencies.

Which is as terrible idea. You have some code deployed, and suddenly this code downloads random things from internets without any control whatsoever.

Even worse if the application was deployed behind firewall and can't access internet freely.

It is definitely a bad idea when working on a server, but for a cli tool or some editor that might install it anyways it could be arguably worth a try.

Yes, a package.json is good to get the required dependencies. This article is about possible dependencies that help in cases where you all the possible would mean a huge load on the system.

If you can use a package without them then they aren't really dependencies are they?

True, how should those things be called?

A good example is Consolidate, which provides a consistent API to various template libraries. You wouldn't want to force installation of all the unused libraries.

https://github.com/tj/consolidate.js

Will add a link to the "other problems" list, thanks!

How is it a huge load on the system given that node can lazy load modules? Do you mean disk space?

Personal experience doesn't always reflect averages: According to the EPI Japan is #30: http://www.ef.edu/epi

I think this was on HackerNews before?

Are you thinking of this project instead? http://richzhang.github.io/colorization/

Yes - how do those two projects differ in their approach ?

Thanks, I think that is what I meant.

Love to be able to attend, unfortunately kind of hard to fit in the schedule. Side question: Is wednesday a good day for a conference?

Hey leichtgewicht,

Its a fair question. It's unknown right now, we thought long and hard about the day so lets see if its a success.

Ian

To me this article relates to the idea of pair-programming. In pair-programming you should also pair lower-skilled with higher-skilled people to lift the beginner's skill. Empathy and Patience required. The problem with teamed-up programming is that it is very much relying on the groups culture and switching between cultures might turn into a hard challenge.

I'd suggest to call an authority other than Homeland Security and tell them a suspicious big box is standing there. Lets see how that plays out.

So I only see two major downsides: 1) I basically I have to verify/transact every "printed" money the moment I get it - preferably using a smartphone - in case someone else has a copy of it and "takes" it first. Copies could be made with any high-resolution camera. 2) If the money travels from A to B to C in paper shape and the digital transaction will be from A to C then A will know that B has done business with C.

Yeah. Right now, it is largely based on trust. We're hoping to implement some kind of one-time "seal" and new token that a user can verify after receiving a bill to ensure that someone else can't activate it with a picture or by saving the token. That obviously comes with the downside of making it more difficult to transfer.

As a proof of concept, and being based on Venmo, I think it is and will be largely based on user trust as long as the bill creator keeps the funds (Venmo allows no bot accounts).

Unless the note changes hands additional times between B and C.

Point taken. It then just becomes a statistical probability of how many steps between A and N. Considering that everybody using this system better have 0 steps (former point made) 1 step is about the furthest I would hope. But you are right: there could be more steps.

Oh! I got to try the InstanTRIP PAD 02 a while ago but it never seemed to have made Hacker News. Thank you Sacha for posting. The Pad02 is a interesting step forward from the 01 version http://www.amazon.com/InstanTRIP-PAD01-4-5-5Inch-Smartphones... , a lot cheaper, a lot lighter. The only problem I see left with the device are that the lenses are fixed size (only working with devices of a limited screen size). Else the product concept is very nice imho. I would love to forward (translate) the HN communities opinion to Cocoloworks.

(Disclaimer: I know the creators - Cocoloworks - in person . They are members of the same coworking space as I am and I helped a little with the technical concept of the Pad02)

"I helped a little with the technical concept"

"Hold the smartphone firmly in your hands"

Seriously?!

Why didn't you help more?

Head Mount setups - as you point correctly out - are required for a long-time VR consumption. But they come with caveats: - They make the devices heavier & bulkier. - You need a good mechanism to fastly switch back to use as a mobile phone (i.e. when you receive a phone call while in VR) - You need to make sure that the phone is tightly in place. If you move your head considerable force is applied to the casing.

Using your hand allows InstanTRIP Pad 02 to be tiny. It is folded together so tiny that you can put it in the back-pocket of jeans and forget you put it in there. Other VR boxes have to be carried in a bag and consume a lot of space in there.

Other head-mount-less "cardboxes" cost at least 20% more than InstanTRIP (for most you pay double the price) and the setup time is considerable.

There have been trade-off's made but I don't think that they are all bad (feel free to correct me)

If this thing is so tiny and light, why can't they simply staple some rubber bands on it?

The rubber bands would hold the case in place but the case is not strong enough to sustain the device. In other words: move your head to fast and it flies away. Hands prevent that.