A website for debloated open source alternatives | Hacker News Reader

A website for debloated open source alternatives

debloat.dev

412 pointsryanvogel131 comments

FWIW, this site is fast, works perfectly with a text-only browser, e.g., links (1.4MB static binary), elinks, or without a (bloated?) browser, e.g., retrieve all the pages, catenate and read as a single HTML file or convert it to, e.g., SQL, CSV or a plain text file

All the /p/ URLs are in the sitemap, all the site's pages can be retrieved over a single TCP connection

https://debloat.dev/sitemap.xml

This, i.e., retrieving all 200 /p/ URLs over a single TCP connection (using HTTP/1.1 pipelining), results in a 1.9MB HTML file comprising all the /p/ pages, including response headers. NB. This isn't "crawling". All URLs (cf. a set of "seed" URLs) are known before the connection is made

No cookies, no Javascript

Other than CSS, no bloat

NB. A 1.9MB sitemap would not be large. For example, here is a 2.9MB sitemap from AP

https://apnews.com/ap-sitemap-202607.xml

Also, the website seems to be largely ‘vibe coded’, at least given by the written copy and the style. If you ask Claude to use 1990s/2000s web design and HTML it’ll do something like this.

The sitemap "standard" was not introduced until 2005

Would a "1990's style" website offer a sitemap.xml file

Also, this site, like most today, uses "slugs" for its URLs. In the 1990's this was less common; URLs were often "uncool":

https://www.w3.org/Provider/Style/URI.html

NB. I'm not commenting on the website's content or how it was generated/sourced

I've been a big fan of https://alternativeto.net/ and using their "open source" and "self-hosted" filters. Even looking through the non-FOSS alternatives for a particular application/site can be helpful; sometimes a for-profit company will take a different approach to solving a problem and it will give me other ideas to consider.

Some of the information on there is useful, but I'd say that alternativeto.net is an almost perfect example of a bloaty & hostile website. OP's website is much more pleasant.

Seems it's always the same trade-off: either a commercial site with an opaque business model, or a weekend passion project which might go away when the hobbyist in question finds they have other (or more lucrative) things to do.

Alternativeto was a site I used to frequent, but after their redesign, it’s been pretty painful to use and a lot less valuable.

SaaSHub founder here. Have you tried https://www.saashub.com? It has both "open-source alternatives" and "EU Alternatives".

In most cases I’m looking for alternatives to local apps, rather than a SaaS offering.

The most bewildering thing about the redesign is that the button to visit an application's website -- arguably the most important one since one would assume people are on the site to try out alternatives -- is now hidden somewhere far below the fold.

i've found this site easier to use and have better results than alternativeto. https://openalternative.co/

I’d have given them bonus points if they’d used “alternativetoalternative.to”.

https://madeineurope.dk is by far the easiest to navigate imo, but all the alternatives are of course european given the website name (can also filter by open source).

Not able to access the site on Firefox "Error Code: SSL_ERROR_INTERNAL_ERROR_ALERT"

Bitdefender blocked it for being malware apparently

Cool site, though browsing through “most popular” I absolutely wouldn’t call Nextcloud “debloated.”

I think in general maintainers would be better served by declaring Performance is a Feature earlier in their projects and rejecting contributions that add surface area to the API at the expense of throughput on the existing surface area.

Wrapping modules to expose additional functionality for a niche in a problem domain is tricky. You will end up with some code duplication and some predictability issues with your timelines, but the end result will be better for it.

Long ago someone convinced me that 'Strangling a Service' could also be applied to APIs. The base module should be as simple as possible but no simpler, and more esoteric features should be shunted off to another module. And then in the case when features are antagonistic to each other, they can live in parallel in separate wrappers.

The tricky part there is writing unit tests in the base to defend the negative space that these features fit into. This feature depends on an invariant in the base API that will break everything if merged.

How is “bloat” measured here? Or is this simply a list of open source alternatives to commercial products?

The inclusion of Tailscale here was a little surprising to me. I'm betting on the latter.

I like Tailscale, but it's difficult to argue with a straight face that it's not seriously suffered from product creep.

This is the inherent problem with tech start-ups. Once they solve the problem they were founded to solve, then what? It's feature complete, but investors insist that the line must go up. Can you imagine if `sed` or `vim` was brought to market by a venture-backed company?

IMHO, capitalism and FOSS are fundamentally incompatible.

> IMHO, capitalism and FOSS are fundamentally incompatible.

A bold statement to make given the world we live in today contains plenty of both, working together, in harmony.

That’s [part of] what regulatory frameworks are supposed to accomplish: minimize and compensate for capitalism’s excesses. A free market can do a lot of very impressive things with surprising efficiency, but it fails on a number of things societies value, FOSS included. IMHO, there should be more robust publicly funded support for license enforcement, facilitated access to forming charitable/non-profit organizations, subsidized grants/loans, and something akin to the Public Broadcasting Service, but for the internet. Sort of like a national intranet, with federated social media, web archives, public service portals, email, digital asset storage, and software development and distribution resources. But that’s just my democratic socialist head-canon. No doubt such a system would be incredibly difficult to implement, and come with any number of unintended consequences that, for the purposes of this discussion, I chose to ignore. It’s also worth noting that there are several of those things in the USA’s federal government already: you can find open source software repositories for a number of federal agencies, like NASA and the NSA (which includes several fun projects).

Someone should add Kanboard for project management tools (like Trello): https://kanboard.org/

Really lightweight and nice!

I want a debloated (electric) car.

Taking the opportunity to plug my own recent project: canireplaceit.com

You can see some collection, and I'm building some advanced research capabilities to it.

For now, you can see if you can fully replace an app with a sub, with oss alternatives, or free alternatives.

Heavily vibe coded but all the apps out there didn't have what I needed, a list of features, what exactly is missing, if there is an sso tax or not, licenses, cheaper alternatives, if there are some tjat genuinely can't be replaced and why.

Lemme know what you think about it and what I should add to it

A lot of alternatives can be found at https://github.com/mayfrost/guides/blob/master/ALTERNATIVES....

suckless.org?

site is down. debloated too much

Category selection should be checkboxes, not a text box.

I first thought I’m on old eBay webpage lol

Same here. I’m guessing that’s not accidental.

Is there a marker to know if any of these tools are AI generated slop here?

That would be a good feature here.

How do you define "AI generated slop"?

There's also an implicit judgment in your question. Is well-designed, well-tested AI-generated code worse than poorly written, but hand-typed code?

Here we go: as soon as you ask for any kind of definition or rigor around AI development, people will work hard to shut you up.

You're asking an extremely reasonable question, IMHO, especially given these are projects with limited resources going up against generally larger, better-resourced projects. AI is an obvious point of leverage in such a situation, and so there is a lot of potential to reverse engineer using AI and get good results.

What all of these requests fundamentally come down to is "How are you filtering for quality?", which is an extremely difficult question given the state of, for example, the app stores.

> Here we go: as soon as you ask for any kind of definition or rigor around AI development, people will work hard to shut you up.

Heh. I was essentially run out of lobsters for repeatedly asking that question. Not even kidding. At some point I became the second most flagged user in the entire forum. The site itself invited me to delete my account.

After a while I simply asked Claude to define "slop" and posted the answer. To this day it's still the only coherent definition of "slop" in the entire site.

Software was never that different but everyone is definitely now realizing that, like all the other creative type things humans produce, the trick isn't finding one, it's finding a good one.

AI involvement is a possible heuristic but I seriously doubt it's all that useful for software (probably works a lot better for books)

Reply on news.ycombinator.com