💾 Archived View for bbs.geminispace.org › u › mediocregopher › 15136 captured on 2024-03-21 at 19:19:41. Gemini links have been rewritten to link to archived content
-=-=-=-=-=-=-
Re: "Which parts of the web want saving? Which can we do without?"
We could cut out a huge swath of security issues (and complexity associated with remediating those security issues) by disallowing loading of resources from third-party domains. No more CORS, CSRF, or CSP, amongst many others.
This wouldn't get rid of ads/tracking, but it would make them harder to implement, so that'd be another positive.
I would also love to see browser support for document formats beyond HTML and PDF. Markdown and gemtext come to mind immediately. By supporting these natively web browsers can tailor to users who don't wish to deal with the complexity of HTML/CSS/JS.
Feb 19 · 5 weeks ago
🌲 Half_Elf_Monk · Feb 19 at 22:47:
I like the simplicity of gemtext, but I get why minimalist aesthetics aren't for everyone. We're in this mess because human psychology is exploitable. Perhaps an inversion of the question would be helpful: What if all cross-site resource links had to be loaded through an advert:// protocol? If a site still wants to serve me giant intrusive ads, they've gotta do the work on THEIR servers/connections. I can choose to block/limit the advert:// protocol as my router sees fit.
advert:// is a good idea, but can you convince advertisers to use it? the whole purpose of advertising is to make anything to make people, who don't even want to see it, see it.
btw, i was surprised to learn that some folks want to see ads. they tell me how they love that google or facebook fingerprints them since it shows them relevant ads.
i was thinking nobody loves ads.
@mediocregopher i agree, but in today's web most of the web pages will stop working. they load shitload of scripts from different domains.
so the problem is indeed a human problem. i think we can get more people to know about gemini. i didn't know for years when gemini already existed.
and i consider myself curious.
🐐 satch [OP] · Feb 20 at 16:53:
Yes @norayr some people have complained to me when watching YouTube (on their browser with no adblock) that I skip ads when they want to watch them... very interesting to think about why
unbelievable.
we don't have youtube ads here in armenia, our region is not (yet?) interesting to advertisers, but i force vpn from our apartment via openwrt wifi router, so my wife has to tolerate youtube ads, as well as has one less reason to tolerate me, which i don't know why she still does.
🍺 mrrobinhood5 · Feb 20 at 18:24:
@norayr that's so funny. but for reals most YT ads are.really bad and not entertaining to watch. I remember TV ads (commercials) and they were light, original, creative, and interesting to watch.
🦉 ResetReboot · Feb 20 at 21:35:
@norayr @mrrobinhood5 With how many examples that a well thought advertising campaign can install itself into the social collective conscience, advertisers decided to go the annoying route.
Given too, that nowadays anyone can really launch an ad campaign. So that show in the quality of those and the things sold...
👻 mediocregopher [...] · Feb 21 at 07:45:
@norayr
in today's web most of the web pages will stop working. they load shitload of scripts from different domains.
True, but I consider this a good thing... people will have to bear the bandwidth burden of their bloated JavaScript monstrosities themselves, rather than outsourcing to some cdn which is selling analytics data and centralizing the web besides.
🐐 satch [OP] · Feb 21 at 14:24:
Loading JavaScript from CDNs is bad and unnecessary. But using an iframe to embed? Not so clear.
🦉 ResetReboot · Mar 02 at 21:43:
I was thinking about the "throwing the distributed application system" and I have a few handful reasons to throw it away.
1- Ownership. It is something being talked about a lot in videogames, with everything turning into digital downloads tied to an account on a service. The day the company goes belly up... you lose what you paid for. Same principle here, and I will cite Google Reader and its untimely demise.
2- The freedom to know what your computer is running. A coworker, sysadmin, asked the frontend team "Can we see what's exactly the code that is running on the client". The reply was "You doin' drugs, mate?". Not even the devs can really say
(Goes on)
🦉 ResetReboot · Mar 02 at 21:46:
3- Overcomplicated system to develop: Try to start a new JS application. Do a simple hello world. You have a lot of frameworks (phased out in mere months!) with a lot of dependencies and requirements. With other programming languages, you can get the compiler, create your hello world and in a minute, you are up and running. There's a lot of well seasoned developers that sigh and reach to see how to start a new JS project.
4- (And this is a personal take!) Starting your base on a language that was only thought to be a glue with Java applets and make a couple nice effects. Then drop a ton of patches to make it more usable.
☕️ tenno-seremel · Mar 19 at 17:23:
I’m with no-scripters. Once you allow scripting it is only a matter of time before it will be extended. Exactly as it already happened. Forms, which can be validated client-side, should stay, though, so every website didn't have to do things this one had to do. And they should be useable without spending lots of time to drag them to present times.
btw there is now this new technology called nextjs which runs js only in backend and only gives the browser prerendered html and css but i doubt corporations will give up browser side scriytss, they need their tracking scripts in your machine.
Which parts of the web want saving? Which can we do without? — Let’s have an unrealistic hypothetical here. Suppose the http web were replaced entirely by a new protocol (or maybe set of protocols?) which allowed for client side scripting, unrestricted styling, and all the other aspects of the web which make it distinct. Except this time, everything was being built from scratch, with person-first values in mind, like privacy, accessibility, ease of use, etc. The designers of this protocol are...