Ask HN: How far could you get Coding without Google?

If I expect to be disconnected from the net I download as many manuals as I can. Back in the 1980s we had lots of paper manuals.

I don't know, but don't take IRC from me. Then I'm really screwed.

Depends what I'm currently doing.. The past week I've implementing some trivial feature in javascript and C and so haven't had to look up anything.

But weeks ago, I had to implement something in DQL, which I'd never used before, their documentation is terrible, and google didn't help that much either.

Docs is all offline with Dash, no problem :)

It wouldn't be a problem - as you state in the question - coders used to do that before the web.

This is a rare (?) occasion where old dogs would likely adapt more quickly due to having done it before. Not only will their old muscle memory kick in, they will be more comfortable with the situation.

I would need a few physical books but I think I would be OK. I would be more likely to use vanilla JS, HTML and CSS as a result.

As 80's programmer, definitly.

Get hold of the necessary books.

All the way through the 00’s!

Had a lot of dog eared O’Reilly books, now I haven’t bought a technical book in over a decade!

Btw, You could cut out google and go straight to stackoverflow, just saying.

I block outgoing ports 80 and 443 quite regularly when I need to get shit done, coding-wise. Takes some getting used to, but it's no major obstacle.

   $ sudo iptables -A OUTPUT -p tcp --dport 443 -j REJECT
   $ sudo iptables -A OUTPUT -p tcp --dport 80 -j REJECT

   $ sudo iptables --flush

that's exactly how i worked for 10+ years at the beginning of my career... with a bunch of ( Oreilly's ) books on my desk.

Same as before Google. Books, text files, and collected code snippets. During school, the SWAG collection for Pascal was eagerly downloaded from BBSs when new editions came out.

http://www.retroarchive.org/swag/index.html

I’d get a copy of Dash and download the Docksets you probably needed.

[1]: https://kapeli.com/dash

I routinely work on things that are dead in the "real" world and so are not Google-able.

If I can cheat a little, OpenAI Codex has been more useful than Stack Overflow.

Also assuming you mean Google the search engine, not the company, because as an Android dev, it would be tough without any documentation.

Used to know somebody who contracted with national security and had to do this. They said it wasn't so bad with paper manuals! The worst part for them was not being able to use outside libraries without going through extremely extensive checking processes

Have an offline copy of devdocs.io which is all I need.

Although I would like instant access to Stack Overflow for various snippets. Things like ‘How to sort an array of numbers’

1. https://www.kiwix.org/en/ : see content at https://library.kiwix.org/

2. Info manuals, info browser

3. Self-documenting programs & environments (e.g., emacs but also Python with docstrings etc)

4. Look at the source code on your machine

I was a programmer in the 80s. We had MS-DOS, Turbo Pascal (with it's damned fine documentation, in an actual book, and help files).

The hardware sucked, but we didn't have to worry about anything other than putting characters on a screen, eventually VGA graphics if we got fancy. There was one platform, the IBM compatible PC, and that was the only one I worried about.

We didn't have GIT, we had PkZip, and a stack of floppy disks with labels like "Source Backup v42 2/3/89". If we made a mistake, we had to manually revert.

I wrote "OverSeer" - a program that managed the inspection of Fire Extinguishers, using hand-held computers and barcode scanners from the Norand corporation of Iowa. Eventually it was adapted to other uses. It was written in Turbo Pascal. All the libraries were my own, including one to do cooperative multi-tasking.

Our customers were in Northern Illinois, and I was the programmer/tech.

As me anything. ;-)

[Edit - Extended description of the stuff I wrote]

In the good old *nix days the system provided pretty much everything you needed: C, perl, man pages, and well documented header files. Throw in a couple of good books on programming, and you could do a lot.

That's still the case, but it feels like today's software development reality is different. There are endless frameworks, apis, devops methods, scalability concerns, and n+1 standards with which to integrate and keep up to date on. Knowing your language is only about 25% of the battle, and few projects are developed in a vacuum.

It's not possible to keep track of all those things, so endless internet searches are the end result.

maybe farther, because i wouldn’t be served garbage content and be distracted all the time

Depends on what I'm doing. If I'm working on my usual BE stuff, I'd probably go for years - I can't remember the last time I googled something about this. If you ask me to do frontend work I'd probably get stuck within a few days.

I'm not sure which is harder to imagine:

1. Programming without google

2. Programming (a web app) without webdev tools

When I first started learning development in the early 2000s, it was php, jquery, html, and css and very few search results when googling.

When Firebug came along for Firefox, my entire world changed and it was as if lights from the heavens had been turned on.

Yeah, a couple good books and possibly some source of my older projects and I'm set. I still regularly buy reference material, not as much nowadays but its nice to have it available in a pinch.

I used to write pseudo-code that didn't compile for one pass, working out logic, then work through turning it into code on a second pass. Generally, stuff I pseudo-coded seemed much less buggy than stuff I wrote straight in as compilable, presumably because I didn't have much besides program-logic to concern myself with. Translating it to something compilable was pretty mechanical, since I wrote something in the ballpark; I just didn't worry about syntax at all.