Should we bring back gopher?

I'd just use mail to be honest. Cuts out low effort shit.

Is there an equivalent to ssl/tls for gopher?

Nice try, CIA niggers.

WTF is your malfunction? It's a gopher site, you access it in Lynx or other gopher client.

no, we should bring back usenet

Usenet and gopher serve different purposes. Everything on Usenet is ephemeral (based on your local server's retention policy). Well it's designed that way, even though jewgle archives many groups. Gopher is more like a cross between FTP and WWW site. Stuff is stored until the admin deletes them, which means it could be there for years or decades.

I don't know much about it, but the Gophernicus README has an example on setting up the server in combination with stunnel for SSL/TLS.

Thanks.
It looks like this board use type 7 to submit through queries. The other one use a seperate CLI HTTP based client.

IPFS is basically gopher (over a giant clusterfuck of p2p protocols). A port 80 web without pointless AJAX actually works pretty well, try it.

afaik ipfs doesn't have gopher menus which are a pretty important thing for plaintext sites.

It's halfway there. They have code for pure metadata nodes written so maybe (assuming the sun doesn't burn out first) one day it'll be able to show directories with manual ordering and inline descriptions.

It's not a bad thought. But we'll need some kind of encryption functionality. I guess we can set up gopher over ssh?

Gopher RFC: ietf.org/rfc/rfc1436.txt

RFC is odd as hell in that it appears to have been made for a 80x60 line resolution.

The first gopher clients and servers were ancient 68k Macs. I suspect much of the protocol weirdness can be traced back there.

Nigger you just summarized the entire history of programming

This makes me feel like a kid again. I just wasted the last 20 minutes typing "funny" into veronica. It's all text based and simple. I love this.

gopher://sdf.org/0/users/jba/Funny/engineer.txt

Looks like 56 lines, followed by a form feed (^L) to advance printer to a new sheet of paper.

I like this one..

Understanding Engineers - Take Seven

Normal people believe that if it ain't broke, don't fix it. Engineers believe that if it ain't broke, it doesn't have enough features yet.

How about us who are interested in gopher start an IRC channel? From there we can explore more of gopher and start setting up some gopher services.

Into the trash it goes.

Why is lynx still the only good browser?

Lynx doesn't try to be fancy, cater to normies, or take over the browser market.
Here's another simple terminal client:
openports.se/net/gopher

I don't see the point unless you want a website to be a secret club.

I actually have a paid sdf account. I don't even know what my username or password is anymore. That was about a decade or so ago. I still have the computer case I put my sdf sticker on.

It seems their paid memberships have changed. Back in the day to be a paid member you sent in like two bucks in cash through snail mail and they changed your account settings and sent you back a sticker.

I tried to setup an account on there around Y2K, or maybe ealier. Sent them $1 via mail like their website instructions said. Never got anything back, no email, no nothing. I was poor at the time too, so that kind of put me off.
Anyway gopher is pretty lean, and should run fine on a RPI board or similar. For hostname you can just use any of the dyndns services.

I think we'd probably do better to apply effort to making secure and simple web pages; avoid JavaScript or gratuitous CSS, keep the information high-quality, maintain it properly (avoid linkrot; links should live forever); send proper response codes, that kind of thing.
We should use HTTP for what it was originally designed: document publishing and retrieval.

The nice thing about gopher is it avoids the slippery slope problem. Someone who has web site will get people constantly telling him him to change it, make it more "modern" (crappy and bloated), and eventually that will happen with most sites. I even noticed some old school sites recently add links to youtube, etc. and/or get ugly modern design where the text content becomes relegated in favor of whatever it is the ugly ass modern sites try to do.
Gopher, being plain old simple text, doesn't have this problem.
And anyway it's good to have something completely different from HTTP/HTML, because it avoids monoculture. There should be other things too, not just gopher. But it's a start...

Lynx doesn't protect you from loading third-party resources at all. You just can't see it because Lynx doesn't have a gui to show it to you.

Wut?! Have you read the HTML5 standard? w3.org/TR/html5/ The big, center, dir, font, frame, strike elements are dropped just for the reason because they were design elements, not text.

HTML5 is about marking up meaning in documents. Those deprecated elements do not improve the semantics of document elements.

Can we have chan board on gopher. I need to shitpost peacefully on gopher site.

Shit looks comfy

Well then give an example, because I don't see how Lynx will automatically load external JS or CSS files, images, cookies, or even frames. When I point it to a URL, it downloads the page as the server sends it. There are no other separate requests from my client, unless I explicitely follow a link (unless it's a cookie and I told Lynx that it's ok to process cookies from that particular domain, but I avoid this).

Yeah right. Go compare that to the Gopher RFC and see which one is simpler. I bet you can even write a gopher client for an 8-bit micro, and this is the level of simplicity that's required to defeat the cia niggers. You want fancy shit? You'll just expose yourself more. Your code will have more bugs, and you'll need more powerful and complicated hardware.

You mean like gopherchan, which was already mentioned in

Oh shit. That's around when I did mine. That means it was more like two decades ago. I feel so old.

Sounds like a GNU project

I really ought to start using it again.

high jej

HTTP was that way too

You are right. HTTP is simple, but is useless to browse anything other than "Hypertext" documents.

Because menus are part of the Gopher protocol, it enables one to browse simpler content such as plaintext.

I've been making progress on a gopher browser. At the moment, it supports history and loading text documents. The next thing I want to work on will be support for more menu item types.

I also use the SDF. Good choice lad. Incidentally, I just finished watching the first season of Robotech on Sunday and was about to start the second season when I switched over to watching some Lone Wolf and Cub to take a break. Yes, I'm a weeb.

Another update for my gopher browser.

Switching between pages in your browsing history now saves your cursor position, so you don't have to rescroll through large documents
The status bar at the bottom tells what kind of menu item you've selected and how far through the document you are
Misc bugfixes

More bugfixes today, as well as cleaning up url handling. The next feature I want to add is the ability to submit queries to search items in a menu. I don't mind that, but I hate that I'll have to write the UI code to prompt the user for a string.

I also decided on a name: "Phrowse".

Nice job.. I'm following your progress.

I added support for search items in menus today. As you can see in this screenshot, gopherpedia.com is now very usable with Phrowse. I also added the ability for users to press 'g' at any time to go to any url they want.

You should try it out. Search for Phrowse on github and build it for yourself.

Let me know when adding proper TLS to Gopher isn't a complete shitshow.

This is the goddamn 21st century. Anyone sane is AT LEAST encrypting client-server comms.

gopher://bitreich.org/ suggests using TOR lol

Can someone, please, explain me Gopher? What's the difference between Gopher and HTTP? Why care?

Gopher is a suckless protocol very similar to HTTP/0.9 (and HTTP/1.0 without any bloat features like verbs and headers).
You open a TCP connection, you write the name of the resource you want followed by CR+LF, you get said resource in return and the server closes the connection. There's a defined format for directory listings and a few fixed file types. That's all there it.

en.wikipedia.org/wiki/Gopher_(protocol)#Protocol

The downsides is that there are no persistent connections as in HTTP/1.1 (because there's no indication of response size), which e.g. makes it crap to run over TLS , and in turn also no way of telling whether the response was cut short.

.onion

TLS is over-complex garbage, the idiots on the gopher mailing list have suggested various TLS and gophers:// schemes which are all insecure.

Using .onion gives you authentication (since the link is a public key) and secrecy by modifying 0 lines of code in your gopher client.

The simplicity of this is genius.

It takes a team of 100 people 15 years to build and maintain the millions of lines required for a web browser.

It takes one person one weekend to make a gopher client.

Gopher makes such a superior choice for an imageboard (or even textboard) that they have to fall back on HTTP for posting.
At least they could have used a hack, for example by percent-encoding the input like
/1/chan/post/tech/728065/Is%20this%20a%20meme%3F

gopher is not good for anything interactive.

web has become too complex so we need to learn from it and make something with gopher simplicity that supports interactivity.

You don't even need to encode anything when making a gopher request, since the resource selection string allows all characters (except tabs)

Gopher is meant for document retrieval. They should use a mailing list or an nntp server to host their forum. And then make an archive of their forum with a gopher interface, if they want it to be accessible via gopher.

A small update to phrowse today:

Just a bit of refactoring in preparation for new features. The next thing I want to support is downloading binary files. After that, I can't think of what else I should implement. Anyone have any suggestions?

I'm thinking that to submit a form with gopher, sites should be using the telnet pointer item and let users fill out fields in a telnet session.

One could probably also do file uploading, however that would be more complex and require a standard way of doing it that both the server and client agree on.

Do you have a link to compile Phrowse, or should we wait until it's more complete?

Tarball can be downloaded from here:
u.teknik.io/xCHnU.gz
Or you can search for phrowse on Github to clone it. The README has instructions for how to build it (all you should need is your distro's ncurses-dev package or equivalent).

Hey, thanks dude! However, when I try to compile it it says that I don't have ncurses.h even though I have the pagae installed. I guess I'll keep trying.

What distro are you running? You need the ncurses header files, and some distros package that differently from the base ncurses. For example, on debian or ubuntu, you want to install ncurses-dev. On a system like arch, however, the headers get included in the standard ncurses package.

No. You can make a simple, gopher-esque webpage and serve it over HTTP/S.

Normies are never going to use gopher.

Getting away from normies and the subversive megacorporations is kind of the entire point of using Gopher in 2017.
You really can't have one without the other. Anywhere there are hordes of normies, you'll find the subversive elements.

I am pleased to announce that Phrowse now supports downloading binary data menu items!

If anyone wants to try it for themselves, you can download the source from here:
u.teknik.io/4IlCf.tgz

As usual, make sure you have ncurses-dev installed. I look forward to hearing any feedback.

Give up on gopher and do something useful instead.

What's with all the jew wide web shills in this thread? We're being raided?

Well, HTTP is an in-between. Then JavaScript got invented. Who's to say we wouldn't end up with some client-side nigger dick up our asses as well?

Do you even know where you are?

You're right. I should start working on a logo

Fuck off back to Holla Forums already. You people are killing this board.

kys take your browser flamewars with you

I support all solutions which marginalize women and minorities, so any barriers to their participation in internet communication are welcomed.

Ok guys, the logo is done.

I think we should bring back HTTP. You know, HyperText Transfer Protocol. It was perfectly usable until Microsoft added fucking XHR and the zombie JS dev plague hit.

I believe your criticism is actually pointed at html not http. Overall yes I agree, we should go back to a simpler and saner model of web documents.
In particular, there should be absolutely no elements. The web has been utterly ruined by the idea that it could be used as a virtual machine to run "real" programs, using a language that was unfit to the purpose.
Maybe in practice we could make a browser and fork of webkit free of javascript and related insanity, and promote the resulting subset as a pseudo-standard of a simpler web. A web server could explicitly declare its conformance using non standard http headers (eg X-No-Fuckery: 1).
Another thing which comes to mind is Alan Kay's presentation "Normal considered Harmful" where I remember him shitting on Berners Lee for creating a web stardard which was vastly inferior to the proposal of Engelbart, also referencing the now defunct software Hypercard. He suggests reading the papers of Engelbarts research group but I have not done it yet.