Mozilla has been part of the problem for ages now. Even their language Rust has been using Discord. Not to mention that their browser has been the pioneer on telemetry.
So strange they chose Slack and Discord when they had a perfectly fine IRC server (that now is deprecated). Instead of trying to improve IRC (either the protocol or any clients), they chose to go with closed-source hosted platforms. Damn shame.
It pops right up for me with #/room/#rust:matrix.org at the top of the list.
Now try searching for Firefox.
There's a Russian Firefox community on Page 2 of the search results. There are beta and nightly groups on page 3. There's various other noise. But the maintainers have hidden the real Matrix room for Firefox so well, it doesn't even seem to be in the search index.
The choice was public. The move has not been public at all.
It’s hard for Mozilla to fight for privacy on the internet since their main customer is Google. Firefox needs to have enough tracking in it, in order to send data to Google. Otherwise Mozilla would earn no money.
And there are no other major customers willing to pay Mozilla. Pretty much the only other options are Yahoo/Verizon (they are currently in a legal fight With Mozilla about a breach of contract that happened in 2017), and Bing (but Microsoft is currently on a mission to promote new Edge, and even included various pop-ups in Windows 10 to discourage people from using Firefox, so I don’t think they are interested in being Mozilla’s customer at the moment).
Mozilla has to stay in bed with google, and this is going to muddle any kind of privacy-related initiatives and communication from them.
I thought the Mozilla Google partnership ended several years ago when they changed to Yahoo for Default search and signed a MASSIVE mutiyear deal with Yahoo.
I know they terminated it (and kept the money) when Yahoo was sold to Verizon but I had not heard they signed a new deal with Google, I assume they were surviving off the Yahoo Money still
Rust's decisions about chat platforms has nothing to do with Mozilla.
Mozilla chose Matrix, not Discord, for example.
(Also, while the popular version is "Rust chose Discord," it's actually more complex than that: each team decides what they want. Many chose Discord. A bunch chose Zulip. The embedded working group is on Matrix.)
Mozilla was heavily involved in the early development of Rust (and since there is no Rust Foundation, is the holder of the Rust trademarks etc.).
But at this point however you can't really call it a Mozilla project as active development is no longer dominated by Mozilla employees. Holding the trademark is just a formality owing to there not being a better way to do that without starting up a new Foundation.
> Mozilla chose Matrix but they advertise Slack on the linked page.
Gotcha, I missed this. Very odd.
> Is Rust not a project of mozilla?
Mozilla was the first sponsor of the project, and still sponsors a bunch of stuff to this day. But they don't own the project, and cannot make decisions on its behalf.
There are quite a few different groups within Mozilla that are accustomed to and comfortable with different communication platforms. For chat, it's mostly Slack vs Matrix. Everyone around me will roll their eyes and grumble if you suggest using Slack for anything but very very clearly company-internal stuff. But other groups "just use Slack" and don't think twice about it, so are a bit confused when we demand to know why they're only on Slack.
I guess few large-ish companies have managed to keep their communication platforms rationalized and consistent. It's more important for Mozilla to stay open than it is for most places, though.
This is total BS. If anything that needs a fix is Mozilla. They have been taking money in 100s of Millions of dollars from the same company that tried to kill their flagship product.
Nothing is wrong with Internet. But a lot wrong with Web. Mozilla cannot fix Web until they start attacking FAANG, which I very much doubt they will.
That would be nice but without that money they are dead in the water. Mozilla is currently the only thing standing against a total browser monoculture.
There's Apple as well, and with the continued decline of Firefox market shares (both absolute and relative), their importance increases. Everywhere where Firefox runs, Chrome can also be installed, so you can just tell users to install Chrome, while it's not possible to do so with iOS devices.
The rendering engine of Safari is similar to blink but not the same. And more importantly, Apple has the money to finance WebKit coming forward, as well as the willpower as they love to implement things themselves. I'm not so sure about how well Mozilla will do in the future compared to Apple.
Apple doesn't care about the open web. They've repeatedly tried to push encumbered media formats and not supported open ones, and they've held back implementing standards that pose any threat to the App Store. The fact that iOS doesn't allow other browsers at all is detrimental to the web.
They do care about the open web, but in their own way. Of course their own profits are most important. Same goes for Google. I still hold the kicking out of flash to Apple's credit. Google followed them on mobile devices and this was the start of the end of flash on the web (the end of the end is soon to come!).
As for the encumbered media formats point, it's been mainly a Mozilla and Google thing, as both distribute their browsers gratis. Apple already pays lots of money for patents on other aspects of their devices, like the wireless connectivity parts. But Apple has been warming up to open formats slowly. They've implemented opus support in Safari (although sadly no ogg/opus). They've joined AOM and maybe one day they'll implement AV1 in Safari as well.
In general, there is a ton of momentum behind H.264 and its successors which is hard to direct towards open alternatives. Outside of browsers, media formats like VP9 are practically dead. Most video cameras use H264 or H265. DVB-S2 uses H264.
> They do care about the open web, but in their own way
Apple's own way is to cripple it to force people to use the App Store. Google's and Mozilla's incentives at least mostly align with those who want an open web.
> As for the encumbered media formats point, it's been mainly a Mozilla and Google thing, as both distribute their browsers gratis.
No, an open web means anybody can contribute and consume without worrying about licensing gatekeepers. By pushing for encumbered formats in web standards, Apple explicitly took a stand against this idea.
> In general, there is a ton of momentum behind H.264 and its successors which is hard to direct towards open alternatives.
All the other browsers supported VP8/9. Apple stuck with H.264 and H.265 until licensing issues forced its hand.
> Most video cameras use H264 or H265. DVB-S2 uses H264.
Most videos created by people are on phones, which save them in the formats that their phones support. It was up to Apple to make its phones support an open format for use on the web, and it refused.
FAANG got their monopoly by having tech stacks able to do basic Web 2.0 features better than open source alternatives.
Once upon a time AOL was the FB of the day and Steve Case was the Mark Z. What happened? The Web Browser happened. HTTP was an open protocol and people could run their own domain foo.com instead of “keyword Foo”
Do you think FAANG could have gotten their start on top of AOL, MSN, Compuserve? The previous FAANG?
What we need is simply to create an open source alternative to FAANG that is permissionless and built on top of the open Web.
And despite being a non-profit company they have been putting millions of dollars in the pockets of their executives. There have been quite a few threads here on HN about that.
Only tiny part of Mozilla is non-profit. It’s called Mozilla Foundation.
The rest, called Mozilla Corporation (I.e. the maker of Firefox) is very for-profit, including all the usual things like profit-driven layoffs and insane bonuses for executives, etc.
I was under the impression that the Mozilla Corporation was owned by the Mozilla Foundation. Would that not mean that the profit from the corporation would go to the foundation? Or is this a legal trick to allow them to profit from the foundation?
They need to keep them financially and legally separated, but Corporation is still allowed to make some money transfes to Foundation (but not the other way around).
I’m actually surprised they are now allowed to have the same CEO for Corporation and Foundation. They might be inviting an IRS audit this way.
> I’m actually surprised they are now allowed to have the same CEO for Corporation and Foundation.
You're spreading misinformation. AFAIK Mitchell Baker is still the chair of the Foundation, but the Foundation's equivalent of a CEO is Executive Director Mark Surman.
> Mozilla cannot fix Web until they start attacking FAANG
They already attack Facebook (they put Facebook in Facebook purgatory by default, don't they?) and I'm not sure what they need to go after Apple, Amazon or Netflix over, I think you are really just mean Google here.
In which ways has Mozilla attacked Facebook? One way I'm aware of is their "Facebook Container" extension, which is basically the same as their Multi-Containers extension, with the key difference of being preconfigured for Facebook. Last I checked there is a similarly preconfigured extension for Google, but that one is from a 3rd party extension developer, not Mozilla (I have a very hard time recommending 3rd party extensions to other people since I do not know if those extensions will remain trustworthy...)
This suggests to me that improving the UX of the generic Multi-Container extension should be a higher priority. As it stands, doing something as simple as creating a new container for a site and then having that site always open in that container takes numerous less-than intuitive user actions.
1) Click "+" (tooltip: create new container)
2) Fill out a name for the container, and choose from a very limited choice of colors and icons (only 8 colors and 12 icons; 96 total combinations)
3) Open a new tab with that container, and navigate back to the website
4) Select "Always open in [container name]"
5) Open a new tab, nagivate back to the website
6) Select "Remember my decision"
7) Select "Open in [container name]"
Step 2 should not have such inane artificial restrictions. Step 3 should happen automatically. Step 6 and 7 should not happen at all because my choice in Step 4 should be respected.
They have a list of domains to exclude from their so-called tracking "protection" and that list contains Google and Facebook. I wouldn't call that attacking by any means.
The Facebook Container is absolutely pointless when you account for fingerprinting and IP address tracking.
Standards-compliant web browsers are impossibly complex to build and maintain, which is why the handful which exist are all under powerful commercial influence.
If you want to "fix" the internet, spend some time enriching open, federated protocols that are designed to be human-scale. The internet is here to stay, but we can do better than the web as it presently exists.
Drew DeVault has a good blog article about this problem. Choice quote:
> I used wget to download all 1,217 of the W3C specifications which have been published at the time of writing, of which web browsers need to implement a substantial subset in order to provide a modern web experience. I ran a word count on all of these specifications. How complex would you guess the web is?
> The total word count of the W3C specification catalogue is 114 million words at the time of writing. If you added the combined word counts of the C11, C++17, UEFI, USB 3.2, and POSIX specifications, all 8,754 published RFCs, and the combined word counts of everything on Wikipedia’s list of longest novels, you would be 12 million words short of the W3C specifications.
This seems exceptionally misleading. Playing with the filters on https://www.w3.org/TR/, we find out filtering to latest documents (to exclude cases where there are multiple versions of a single document) brings us down to 953, excluding non-normative notes brings us down to 573, we can likely exclude all semweb-related specs to bring us down to 480 specs.
And that's still including 199 working drafts, not all of which have any implementation (and some are unlikely to ever have any at this point, essentially long-abandoned).
The web platform arguably has problems with scale, but it's nowhere near that extreme.
Also if you were starting a browser development nowadays you could probably safely strip out most of the XML specs as not that relevant.
Keep XML, XML Namespaces, drop XSL-T (despite how much I love it, I would not implement if I were building a browser), XSL-FO (never had browser implementation tht I'm aware of), XML Schema and probably some other stuff I'm forgetting.
It's flexible enough that you could write a variety of interesting services in it. The primary limitation would be the amount of information you can feed in with URLs capped at 1024 bytes. Mostly-read-only APIs like fetching maps are still quite practical.
With INPUT, REDIRECT, and the mechanism for creating sessions via transient certificates, field-at-a-time forms are possible. You could even build a pretty friendly experience around it, since the server can validate and re-prompt each field as you go.
I've been thinking that a text/gemini browser could recognize link lines that end in .png/.gif/.jpeg and offer a button next to them to manually expand the imaages inline. It would be a nice balance between being able to see figures along with the rest of a document without bloating page fetches. Perhaps other data types would have sensible inline interpretations, too- CSV files become inline tables if you expand them, etc.
I understand the protocol is simple, but there's no reason the client has to be. Taking your example:
> CSV files become inline tables if you expand them, etc.
Why not only make them inline tables, but make them searchable, pivotable and graphable? Embed a bunch of statistical tooling directly in the interface. It's something I dream would happen on the Web - if Google^H^H^H^H^H^Hweb standard bodies could agree on a new tag for tabular data, say <datatable>, that would - by default - come with aforementioned capabilities, and with enough styling and hooks to lure in the webdevs, then maybe all the tabular data that's on webpages could be finally made explorable in an efficient fashion...
I hear you, and I definitely imagined something along those lines eventually wrt. basic analysis tools in-browser.
In my mind, the key idea is that instead of inventing a new standard or extending an existing one (a new HTML tag, for example), what we need to do is compose formats which are each individually simple.
CSV is far from perfect, but it has an RFC, it's simple, and there are already loads of tools that can work with it. Many times I have wanted to work with data that was displayed in a tabular fashion on a web page, and I have variously used copy-and-paste and manual cleanup or writing a scraper to get it into a workable form. Imagine how much easier my life could have been if those tables simply were CSV files that my browser happened to display nicely, and I could have fetched them directly at any time!
A format like text/gemini is an opportunity to start from scratch with this kind of compositional thinking in mind. Smart browsers could do all sorts of convenient things, but even "dumb" browsers and scripts benefit from having data and content in accessible formats.
Those powerful commercial influences should use their muscle to force the use of protocols other than TCP rather than the silliness of piling everything onto port 80/443.
They do. HTTP/3 is over QUIC, which uses UDP (on ports 80 and 443, as it should). That effort has been spearheaded by Google.
Edit: Also, putting everything on port 443 is a major security and privacy boon. Internet service providers can no longer just block ports for services they don't like.
Is there an actual problem with routing SCTP over the Internet? ISTM that the main issue would be enterprise middleboxes being generally unable to deal with it, in which case QUIC may turn out to be more of a stopgap solution.
This might be naive, but it bugs me that protocol ports are visible to intermediaries. HTTPS encrypts the URL path. Why don't TCP and UPD encrypt (or otherwise scramble) their port numbers? (Yes, I know, those protocols don't include cryptographic provisions. Maybe they should be updated?)
For better or worse, one of the main drivers behind piling everything onto ports 80 and 443 is the silliness of corporate security theatre. If your product/service/idea communicates by any other port, regular employees most likely won't be able to use it, and you'll have harder time selling it, as dealing with IT is often too much of a hassle for the interested manager.
I love the idea behind Gemini, and on a similar note, I'm wondering: is there a name for the back-to-basics, web 1.0 philosophy? Like, a name for the "sites that look like https://motherfuckingwebsite.com " ?
No, because the only traction it has ever had and will ever have is with techies who are capable of recognizing (and issuing) a description of "HTML documents with only very minimal client-side JS, often without any CSS whatsoever".
I don't think a case can be made for Wikipedia being on http://motherfuckingwebsite.com/ style level. Craigslist--maybe the listing pages, sure. And that's a fair point! It has more penetration than I was allowing. But I don't think it's going to get a designy sounding name because designy sorts of people don't feel the same nostalgic affection for fairly raw HTML.
It looks alright but to me the simple-web ideal is rather like.. no design. Like, you don't design an essay, or a comment, you just write one. And so was the web. Initially you did not do "web design", you wrote hyperlinked documents with some markup to support formatting (not design!) and embedding images; how it looks in the end is up to the user agent (and transitively to the user who could pick and configure their UA).
Then web design happened and it all went to hell. Web "designers" started doing "cute" things and business owners started wanting more and more branding and fashionable stuff.. and here we are. I've said it before: https://news.ycombinator.com/item?id=22608293
HN looks simple, but its under-the-surface table-based layout is ... actually something of a mess. Especially if you want to parse or reformat it readily (e.g., by parsing the structure, not just reskinning CSS).
I feel like I've heard of this world of exclusively hyperlinked simple web pages referred to as "the document web".
But in my brief searches I can't find another source for that term. Maybe I'm inventing it in my head right now, but regardless, it feels like a good name.
> Gemini is a new, collaboratively designed internet protocol, which explores the space inbetween gopher and the web, striving to address (perceived) limitations of one while avoiding the (undeniable) pitfalls of the other.
I had to look up what gopher is and still don't get the pitch. What are the perceived limitations of gopher? What are the undeniable pitfalls of the web?
If the author ever reads this: Why not write something along the lines of "Gemini is an internet protocol that avoids {{ insert pitfalls of the web here }} while being {{ insert opposite of perceived limitations of gopher here }}."?
Gopher was an early competitor to HTTP. Gopher documents ("gophermaps") have a simple tabular presentation. The types of data and links which can be placed in a gophermap are limited and non-extensible, as Gopher predates MIME types. Gopher also has no facility for encrypted communication.
The web (both HTTP(S) and its preferential presentation format, HTML) is built on standards which are obscenely complex, which have features that do not have user privacy at heart (cookies, user-agent strings, css lazy-loading resources, etc., etc.), and these standards are, as I noted above, heavily guided by profit-motivated actors.
While I agree that the top-level pitch might benefit from speaking to a broader audience, the FAQ and standard document linked from the main page provide answers to many of your questions.
Genuinely, if they want to help fix the internet, they should sack everyone not doing core browser work and spend every penny they have spare on marketing to increase their market share. Busses, billboards, TV, radio, Google ads, whatever. Spend it all. Millions a month, internationally. With balance in the browser market comes accountability for Google, which is sorely lacking. Everything else is just mucking about, imo.
Mozilla is in a infinite game. An all-in approach like that could be fatal. They are in this for the long term.
Specially because they can be out-spent massively. If Google, Microsoft, Apple, any of the competitors want, they can outspend Mozilla during their "burn all money" campaign and drown them.
Also, Mozilla doesn't exist to "make Google accountable". Mozilla is there to ensure the web stays open, and it's been doing a damn-fine job at that. Not perfect; but damn good.
Update: Also, how does balance in browser market make Google accountable? Google is far bigger than Chrome.
I find it very hard to believe that Mozilla would have any luck with any amount of advertising. Firefox offers no features that are useful to non-technical people that Chrome doesn't have. If anything, it's worse for the average person, since there are quite a few sites that don't work well in Firefox.
It's not crazy fast like Chrome was compared to Firefox when Chrome came out. Privacy is difficult to quantify or feel, so it's not really an effective marketing technique to market to a large number of people.
Can you expand? I did a little research and found on Wikipedia:
> The Mozilla Foundation is an American not-for-profit organization that exists to support and collectively lead the open source Mozilla project... Unlike the Mozilla Foundation, the Mozilla Corporation is a tax-paying entity, which gives it much greater freedom in the revenue and business activities it can pursue. [1]
> The Mozilla Corporation is a wholly owned subsidiary of the Mozilla Foundation... Any profits made by the Mozilla Corporation will be invested back into the Mozilla project. There will be no shareholders, no stock options will be issued and no dividends will be paid. The Mozilla Corporation will not be floating on the stock market and it will be impossible for any company to take over or buy a stake in the subsidiary. [2]
So seems like the for-profit part is just to give it more flexibility in achieving the non-profit aims?
I'd love to know why you think "their end goal is profit". There's nothing in the Wikipedia pages about any substantiated criticism there. So just curious if there are real reasons for thinking this, or not.
Easy. Start by releasing a browser with built-in uBlock Origin and a bunch of other extensions that make the internet better (strip utm_* crap from URLs, etc).
You know, actually do something instead of merely pretending like they currently do. Their current "tracking protection" is an absolute fraud when you look at the list of domains it whitelists explicitly, and both Google and Facebook are in there. The opt-out telemetry (illegal under the GDPR, should be opt-in instead) as well as the telemetry for those who opt-out of telemetry (yeah you heard that right) is just the icing on the cake.
>(illegal under the GDPR, should be opt-in instead)
Anonymized telemetry that cannot be correlated to an individual user, and which contains no sensitive information, is perfectly legal under the GDPR.
Mozilla telemetry isn't tracking your movements across the web, it's measuring things like "how many crashes are being experienced per 1000 hours on Nvidia graphics cards" and "what percentage of users have addons installed".
Anonymized datasets have been deanonymized in the past so there's no telling how anonymous the telemetry actually is and when in doubt I'd err on the side of caution.
The IP address is also being transmitted as part of the telemetry request, and while they can claim they don't log it there is nothing technically preventing them (or an attacker) from logging it which could be a privacy violation especially if you correlate it with the telemetry data.
Following Conway's Law, the problem with the internet is not so much the internet but the destruction of IRC and the centralization of email. The stagnation in open-source email software and security, the embrace-extend-extinguish of AMP for email, and the decline in quality independent email providers. Fix email (and revive IRC) and we can start fixing the internet.
Probably less than 1% of browser users have any idea what IRC or SMTP are. For your average user chat is whatever FAANG puts on their device / screen and email is whatever address one of the several services you signed up for when getting Internet gave you.
The Internet is considered broken by Mozilla because corporate centralization has turned the web experience into at most a dozen websites per user down from the potential infinity of domains possible.
> Join us on Slack to be a part of the Builders community.
I know it's a bit petty, but I find it quite odd that Mozilla is promoting (or at least using) Slack, when Slack has a large feature—video conferencing—which only works in Chrome due to using non-standards-compliant WebRTC. And when asked by Mozilla, Slack reportedly stated they weren't interested in fixing it.
IPFS isn't http. Web browsers as software are meant to access and render html obtained via http. Content addressing on IPFS is entirely different to how dns resolution works and as such you shouldn't just glue on an entirely different name resolution system to existing software. You want an ipfs browser that is UX optimized for actually browsing ipfs content.
Back in the day browsers supported alternative protocols like ftp but in practice those are largely abandoned to single function as an http browser.
And what in the world would adding PGP to Firefox do? Its not an email client. Its not an IM client. A website that wanted to can already use PGP via wasm. Firefox is not an SMTP client, its an HTTP one.
The goal should be to render HTML documents not control where they come from. The server/client model or the master/slave model is only interesting if the clients are mediocre computers with limited bandwidth.
Think of it like managing a bunch of children. It's a great formula until they are old and mature enough. If you continue to control them as if children they effectively become your slaves.
I have tons of bandwidth and disk space. This is all that is required. If you unreasonably insist I should also be the one to serve every request it's game over for me. Not by technical limitation but by force.
Oh, I can rent my own web space? I haven't money for this and it has been tried before. Those sites are all gone. Everything is gone. I have tens of thousands of blog postings talking about stuff that doesn't exist anymore.
It is certainly not perfect, for example IPv4 has a scaling problem, but it still works damn well. It can be improved so that it can fit the needs it didn't have originally (ex: with IPv6), but it is not fixing, it is an adaptation.
None of the points in the "fix-the-internet" are things that are broken. They are new needs. Everything privacy-related for instance. The base internet protocol are all clear text, then we started encrypting secrets like passwords, then personal data, then full end-to-end encryption. Decentralization went from providing a reliable link between two computers, to services that are fully accessible even when part of it is offline, to services that simply can't be shut down.
There is no fix here, it is just building stuff on top of an internet that works better than I ever hoped for. Good, but calling it "fixing" is a bit pretentious.
When is this nonsense going to end? How about just “Founders Pitch Practice”? If you want to fix something, Mozilla, maybe start by fixing your bigotry.
When a critical mass of people start treating women (as a group) with the same respect and dignity as men are treated (as a group) would be a good start, no? Or are we post-sexism now?
Hey everyone, cool to see this being shared again. I'm Patrick, one of the mentors for this Incubator. Lots of chatter in this thread already, but just want to reiterate what we're offering / looking for in this Incubator:
We have 3 different offerings for this Summer. All incubator style, meaning you meet weekly or biweekly with mentors and we really try to help drive you from point A to point C.
1. $75k investment in a startup. MUST be serious about wanting to build something awesome and put in the hard work it takes to do so.
2. $16k funding in a much earlier stage project (idea stage / MVP stage). MUST be serious about commitment it takes to get to launch.
3. OPEN LABS: these are open to the entire community and you have access to the mentors. 10 min checkins each week & peer sessions. We've had TONS of amazing projects for our Open Labs in the Spring and we hope to see TONS more for the Summer.
In terms of MISSION and what we're looking for:
We started this new incubator out of Mozilla in order to work with & invest in developers, startups, and technology enthusiasts who are building things that will shape the internet and have a positive impact without needing to hyper focus on the bottom line. Projects, apps, & technologies that will be huge and a big part of the internet ecosystem while also being in line with Mozilla's mission of an open web & ethos of "Privacy over Profit".
Comment below if you have any questions about the programs and we'll be happy to answer them.
This probably won't be popular but, does anyone take Mozilla seriously? For me they've always seemed well intentioned, but completely oblivious to reality. For example, does a browser with 5% marketshare really need its own engine? Can a browser with 5% marketshare even change anything if they wanted too?
Personally I don't think so, & also why I don't take them seriously. They seem to be more concerned with waving their "For The People" flag, than actually trying to change anything. If they were serious about "fixing the internet" they'd swallow their pride, transition FF to Chromium, & essentially become something akin to an activist shareholder within Chromium.
They did not always have 5% market share. iOS and Android pushed WebKit/Blink hard, and Mozilla did not quickly respond; the ecosystem has suffered as a result.
That's all true, but overall they've been steadily falling since 2010.
However I disagree with your view of the ecosystem. The web needs stability and consistency more than anything else. FF switching to Chromium would help with that. So many people have this knee jerk reaction of Chromium = Chrome = Google having total control. But don't understand the only reason Google has had this much control over Chromium is because no other major vendor used it. They were the biggest kid on the street. But now that MS moved in a few doors down, that's no longer true. Google has to acknowledge MS in a way they never did with Opera, Vivaldi, Brave, etc.. And the same thing would happen if FF switched to Chromium.
Idk about you, but having 3 of the 4 biggest vendors all being forced to collaborate and implement solutions supported by at least 1 of the others, is 1000000x better than having each do their own thing. You effectively go from a monarchy to some form of democracy.
I used Ff since the early Phoenix days. Sadly, some ten years ago, the organization was taken over by the wrong people. And now, it's a sjw group that cares more about politics than anything else. It's really a shame, because the world very much needs a second major browser engine. Just that these people and their political games destroyed the one organization that could have developed it.
A few ideas off the top of my head. Note that I mostly work on the backend, so my frontend knowledge is at least 5 years out of date. My terms might be wrong, but I think the concepts align closely with how I projected the web would work back in 1995 when I first saw it. Pretty much everything has gone the opposite direction though, and we’ve recreated all of the pitfalls of desktop programming on the web.
- A one-shot, partially Turing complete scripting language would be nice. We should be able to include external scripts, position elements, perform basic computation, and then have a guarantee that the script has finished and will no longer run and use resources. Javascript is fine, but the user would have to give permission for it to run setTimeout(), setInterval() and similar ongoing operations. Discussion needed about scripts running post-load, like for clicks are keeping elements positioned (possibly a user-settable instruction, memory or time limit).
- I'd vote to largely retire frontend model-view-control (MVC) patterns. The browser is already the view, so I find most of the approaches introduced by frameworks like Angular.js to be at the very least superfluous, but potentially misguided or even hazardous. See Intercooler.js for declarative alternatives to what we have now.
- Instead of hand-rolled scripted pages, browsers need a basic component or window metaphor, in which all GUI elements descend from a base element, can be nested, and have a full life cycle that can be customized at each step. The user would have to give permission to override builtin elements like <body>, <input>, etc, but elements could descend from them with overrides. Discussion is needed about security. For what it's worth, I just don't think that the various approaches to building components have panned out. I feel like there is a common abstraction here behind things like React that could be generalized.
- On that note, the way we include remote resources on the web is broken. We need a way to choose whether the HTML for a component is written in-place, or as a URL (or other address) that includes the remote definition. Similar to if inline data: base64 images correspond to our current markup, then we could replace a block of text with an href to allow reuse. Like <div href=“example.com/mydiv.html”>, and mydiv.html would hold <div>something<span>something</span></div> or whatever. See server side includes (SSI) for an early failed example of this.
- Individual requests need to go away and be replaced by a content-addressable caching system where the hash is the address, so we can go back to including individual global sources of truth for things like jQuery or Bootstrap files, rather than maintaining copies of those includes on our own server. See subresource integrity (SRI) as a starting point.
- It may be time to scrap build systems. I'd like to see the world take a step back, and consider architecting a web where more protocols and formats like UDP or Markdown "just work" and don't need to be transpiled to whatever brittle/proprietary-infested conventions are available. Think of this as how compilers can provide generators by altering the code to work in runtimes that don’t have generators. That concept could be extended to everything, so that the server/browser is able to provide features that aren’t yet mainstream, by perhaps running portions of our build pipelines directly under the hood. Basically a turing complete language should be able to run anything, but right now we have wide swaths of computer science blocked to us for security and other reasons, but usually because a proprietary player like Microsoft doesn’t want to implement something. We should be able to run Firefox within Edge, conceptually. And figure out what fundamental capabilities must be provided to allow for something so general. It could be as basic as a general-purpose sandbox where the user decides which system calls and data to al...
Is there a way to see people that have won / applied to this grant, and are looking for help? I suspect many of the projects in this space are ones I'd be interested in working on.
"Fix-the-internet" and "shape the internet" have different meanings, and you have used both... so I am not sure what you are offering!
Where are the problems to be fixed? I can see glossy bios' of mentors and their song-and-dance about how great this is, but it seems a PR gimmick from Mozilla.
Why not make webpages that are not accessible by any of x in x = {Chrome, Firefox, IE, Edge, Opera, ...} and only accessible via simplistic web interfaces?
(Yes, IRC is like that.)
The cool factor can be as much "everyone is doing it" as "the cool kids are doing it".
118 comments
[ 4.4 ms ] story [ 206 ms ] threadDoubt.
It is not deprecated. It is completely dead. Try connecting to irc.mozilla.org — you can’t because it’s permanently offline.
It pops right up for me with #/room/#rust:matrix.org at the top of the list.
Now try searching for Firefox.
There's a Russian Firefox community on Page 2 of the search results. There are beta and nightly groups on page 3. There's various other noise. But the maintainers have hidden the real Matrix room for Firefox so well, it doesn't even seem to be in the search index.
The choice was public. The move has not been public at all.
And there are no other major customers willing to pay Mozilla. Pretty much the only other options are Yahoo/Verizon (they are currently in a legal fight With Mozilla about a breach of contract that happened in 2017), and Bing (but Microsoft is currently on a mission to promote new Edge, and even included various pop-ups in Windows 10 to discourage people from using Firefox, so I don’t think they are interested in being Mozilla’s customer at the moment).
Mozilla has to stay in bed with google, and this is going to muddle any kind of privacy-related initiatives and communication from them.
I know they terminated it (and kept the money) when Yahoo was sold to Verizon but I had not heard they signed a new deal with Google, I assume they were surviving off the Yahoo Money still
https://www.computerworld.com/article/3487825/mozilla-in-tro...
I wish people didn't instinctively downvote when someone is wrong on a technicality, but this is a site filled with nerds after all lol
EDIT: i guess he isnt even technically wrong.
Mozilla chose Matrix, not Discord, for example.
(Also, while the popular version is "Rust chose Discord," it's actually more complex than that: each team decides what they want. Many chose Discord. A bunch chose Zulip. The embedded working group is on Matrix.)
> Rust's decisions about chat platforms has nothing to do with Mozilla.
Is Rust not a project of mozilla?
But at this point however you can't really call it a Mozilla project as active development is no longer dominated by Mozilla employees. Holding the trademark is just a formality owing to there not being a better way to do that without starting up a new Foundation.
Gotcha, I missed this. Very odd.
> Is Rust not a project of mozilla?
Mozilla was the first sponsor of the project, and still sponsors a bunch of stuff to this day. But they don't own the project, and cannot make decisions on its behalf.
That's patently false. It pales in comparison to the telemetry sent by Edge:
https://arstechnica.com/information-technology/2020/03/study...
Regardless, I said that it was the pioneer, not that it has the most telemetry.
I guess few large-ish companies have managed to keep their communication platforms rationalized and consistent. It's more important for Mozilla to stay open than it is for most places, though.
It used to be a company with great engineering. Now it’s a company with excellent cultural and diversity management.
The rendering engine of Safari is similar to blink but not the same. And more importantly, Apple has the money to finance WebKit coming forward, as well as the willpower as they love to implement things themselves. I'm not so sure about how well Mozilla will do in the future compared to Apple.
As for the encumbered media formats point, it's been mainly a Mozilla and Google thing, as both distribute their browsers gratis. Apple already pays lots of money for patents on other aspects of their devices, like the wireless connectivity parts. But Apple has been warming up to open formats slowly. They've implemented opus support in Safari (although sadly no ogg/opus). They've joined AOM and maybe one day they'll implement AV1 in Safari as well.
In general, there is a ton of momentum behind H.264 and its successors which is hard to direct towards open alternatives. Outside of browsers, media formats like VP9 are practically dead. Most video cameras use H264 or H265. DVB-S2 uses H264.
Apple's own way is to cripple it to force people to use the App Store. Google's and Mozilla's incentives at least mostly align with those who want an open web.
> As for the encumbered media formats point, it's been mainly a Mozilla and Google thing, as both distribute their browsers gratis.
No, an open web means anybody can contribute and consume without worrying about licensing gatekeepers. By pushing for encumbered formats in web standards, Apple explicitly took a stand against this idea.
> In general, there is a ton of momentum behind H.264 and its successors which is hard to direct towards open alternatives.
All the other browsers supported VP8/9. Apple stuck with H.264 and H.265 until licensing issues forced its hand.
> Most video cameras use H264 or H265. DVB-S2 uses H264.
Most videos created by people are on phones, which save them in the formats that their phones support. It was up to Apple to make its phones support an open format for use on the web, and it refused.
They do care so much about the "open web" that they embed a patented H264 decoder.
FAANG got their monopoly by having tech stacks able to do basic Web 2.0 features better than open source alternatives.
Once upon a time AOL was the FB of the day and Steve Case was the Mark Z. What happened? The Web Browser happened. HTTP was an open protocol and people could run their own domain foo.com instead of “keyword Foo”
Do you think FAANG could have gotten their start on top of AOL, MSN, Compuserve? The previous FAANG?
What we need is simply to create an open source alternative to FAANG that is permissionless and built on top of the open Web.
I’m doing it, as are many other projects: https://qbix.com/QBUX/whitepaper.html#Distributed-Operating-...
Also: Matrix, Inrupt, etc.
I’m actually surprised they are now allowed to have the same CEO for Corporation and Foundation. They might be inviting an IRS audit this way.
You're spreading misinformation. AFAIK Mitchell Baker is still the chair of the Foundation, but the Foundation's equivalent of a CEO is Executive Director Mark Surman.
They already attack Facebook (they put Facebook in Facebook purgatory by default, don't they?) and I'm not sure what they need to go after Apple, Amazon or Netflix over, I think you are really just mean Google here.
This suggests to me that improving the UX of the generic Multi-Container extension should be a higher priority. As it stands, doing something as simple as creating a new container for a site and then having that site always open in that container takes numerous less-than intuitive user actions.
1) Click "+" (tooltip: create new container)
2) Fill out a name for the container, and choose from a very limited choice of colors and icons (only 8 colors and 12 icons; 96 total combinations)
3) Open a new tab with that container, and navigate back to the website
4) Select "Always open in [container name]"
5) Open a new tab, nagivate back to the website
6) Select "Remember my decision"
7) Select "Open in [container name]"
Step 2 should not have such inane artificial restrictions. Step 3 should happen automatically. Step 6 and 7 should not happen at all because my choice in Step 4 should be respected.
The Facebook Container is absolutely pointless when you account for fingerprinting and IP address tracking.
The former being something they actively support defending against with privacy.resistFingerprinting?
If you want to "fix" the internet, spend some time enriching open, federated protocols that are designed to be human-scale. The internet is here to stay, but we can do better than the web as it presently exists.
Gemini is worth looking into: https://gemini.circumlunar.space
> I used wget to download all 1,217 of the W3C specifications which have been published at the time of writing, of which web browsers need to implement a substantial subset in order to provide a modern web experience. I ran a word count on all of these specifications. How complex would you guess the web is?
> The total word count of the W3C specification catalogue is 114 million words at the time of writing. If you added the combined word counts of the C11, C++17, UEFI, USB 3.2, and POSIX specifications, all 8,754 published RFCs, and the combined word counts of everything on Wikipedia’s list of longest novels, you would be 12 million words short of the W3C specifications.
https://drewdevault.com/2020/03/18/Reckless-limitless-scope....
And that's still including 199 working drafts, not all of which have any implementation (and some are unlikely to ever have any at this point, essentially long-abandoned).
The web platform arguably has problems with scale, but it's nowhere near that extreme.
Keep XML, XML Namespaces, drop XSL-T (despite how much I love it, I would not implement if I were building a browser), XSL-FO (never had browser implementation tht I'm aware of), XML Schema and probably some other stuff I'm forgetting.
Those are some wordy specs we would be dropping.
As far as interacting with the WWW vs "Gemini space" it's night and day, or should I say nightmare and day.
With INPUT, REDIRECT, and the mechanism for creating sessions via transient certificates, field-at-a-time forms are possible. You could even build a pretty friendly experience around it, since the server can validate and re-prompt each field as you go.
I've been thinking that a text/gemini browser could recognize link lines that end in .png/.gif/.jpeg and offer a button next to them to manually expand the imaages inline. It would be a nice balance between being able to see figures along with the rest of a document without bloating page fetches. Perhaps other data types would have sensible inline interpretations, too- CSV files become inline tables if you expand them, etc.
Just a few ideas I'm knocking about.
> CSV files become inline tables if you expand them, etc.
Why not only make them inline tables, but make them searchable, pivotable and graphable? Embed a bunch of statistical tooling directly in the interface. It's something I dream would happen on the Web - if Google^H^H^H^H^H^Hweb standard bodies could agree on a new tag for tabular data, say <datatable>, that would - by default - come with aforementioned capabilities, and with enough styling and hooks to lure in the webdevs, then maybe all the tabular data that's on webpages could be finally made explorable in an efficient fashion...
In my mind, the key idea is that instead of inventing a new standard or extending an existing one (a new HTML tag, for example), what we need to do is compose formats which are each individually simple.
CSV is far from perfect, but it has an RFC, it's simple, and there are already loads of tools that can work with it. Many times I have wanted to work with data that was displayed in a tabular fashion on a web page, and I have variously used copy-and-paste and manual cleanup or writing a scraper to get it into a workable form. Imagine how much easier my life could have been if those tables simply were CSV files that my browser happened to display nicely, and I could have fetched them directly at any time!
A format like text/gemini is an opportunity to start from scratch with this kind of compositional thinking in mind. Smart browsers could do all sorts of convenient things, but even "dumb" browsers and scripts benefit from having data and content in accessible formats.
Edit: Also, putting everything on port 443 is a major security and privacy boon. Internet service providers can no longer just block ports for services they don't like.
Then web design happened and it all went to hell. Web "designers" started doing "cute" things and business owners started wanting more and more branding and fashionable stuff.. and here we are. I've said it before: https://news.ycombinator.com/item?id=22608293
But in my brief searches I can't find another source for that term. Maybe I'm inventing it in my head right now, but regardless, it feels like a good name.
--
[0] - https://en.wikipedia.org/wiki/Resource_Description_Framework
I had to look up what gopher is and still don't get the pitch. What are the perceived limitations of gopher? What are the undeniable pitfalls of the web?
If the author ever reads this: Why not write something along the lines of "Gemini is an internet protocol that avoids {{ insert pitfalls of the web here }} while being {{ insert opposite of perceived limitations of gopher here }}."?
The web (both HTTP(S) and its preferential presentation format, HTML) is built on standards which are obscenely complex, which have features that do not have user privacy at heart (cookies, user-agent strings, css lazy-loading resources, etc., etc.), and these standards are, as I noted above, heavily guided by profit-motivated actors.
While I agree that the top-level pitch might benefit from speaking to a broader audience, the FAQ and standard document linked from the main page provide answers to many of your questions.
Specially because they can be out-spent massively. If Google, Microsoft, Apple, any of the competitors want, they can outspend Mozilla during their "burn all money" campaign and drown them.
Also, Mozilla doesn't exist to "make Google accountable". Mozilla is there to ensure the web stays open, and it's been doing a damn-fine job at that. Not perfect; but damn good.
Update: Also, how does balance in browser market make Google accountable? Google is far bigger than Chrome.
It's not crazy fast like Chrome was compared to Firefox when Chrome came out. Privacy is difficult to quantify or feel, so it's not really an effective marketing technique to market to a large number of people.
Their end goal is profit. Same for their execs: maximize their own profit individually.
Once this is understood it's easier digest PR bulshit.
> The Mozilla Foundation is an American not-for-profit organization that exists to support and collectively lead the open source Mozilla project... Unlike the Mozilla Foundation, the Mozilla Corporation is a tax-paying entity, which gives it much greater freedom in the revenue and business activities it can pursue. [1]
> The Mozilla Corporation is a wholly owned subsidiary of the Mozilla Foundation... Any profits made by the Mozilla Corporation will be invested back into the Mozilla project. There will be no shareholders, no stock options will be issued and no dividends will be paid. The Mozilla Corporation will not be floating on the stock market and it will be impossible for any company to take over or buy a stake in the subsidiary. [2]
So seems like the for-profit part is just to give it more flexibility in achieving the non-profit aims?
I'd love to know why you think "their end goal is profit". There's nothing in the Wikipedia pages about any substantiated criticism there. So just curious if there are real reasons for thinking this, or not.
[1] https://en.wikipedia.org/wiki/Mozilla_Foundation
[2] https://en.wikipedia.org/wiki/Mozilla_Corporation
Easy. Start by releasing a browser with built-in uBlock Origin and a bunch of other extensions that make the internet better (strip utm_* crap from URLs, etc).
You know, actually do something instead of merely pretending like they currently do. Their current "tracking protection" is an absolute fraud when you look at the list of domains it whitelists explicitly, and both Google and Facebook are in there. The opt-out telemetry (illegal under the GDPR, should be opt-in instead) as well as the telemetry for those who opt-out of telemetry (yeah you heard that right) is just the icing on the cake.
Anonymized telemetry that cannot be correlated to an individual user, and which contains no sensitive information, is perfectly legal under the GDPR.
Mozilla telemetry isn't tracking your movements across the web, it's measuring things like "how many crashes are being experienced per 1000 hours on Nvidia graphics cards" and "what percentage of users have addons installed".
The IP address is also being transmitted as part of the telemetry request, and while they can claim they don't log it there is nothing technically preventing them (or an attacker) from logging it which could be a privacy violation especially if you correlate it with the telemetry data.
The Internet is considered broken by Mozilla because corporate centralization has turned the web experience into at most a dozen websites per user down from the potential infinity of domains possible.
I know it's a bit petty, but I find it quite odd that Mozilla is promoting (or at least using) Slack, when Slack has a large feature—video conferencing—which only works in Chrome due to using non-standards-compliant WebRTC. And when asked by Mozilla, Slack reportedly stated they weren't interested in fixing it.
Back in the day browsers supported alternative protocols like ftp but in practice those are largely abandoned to single function as an http browser.
And what in the world would adding PGP to Firefox do? Its not an email client. Its not an IM client. A website that wanted to can already use PGP via wasm. Firefox is not an SMTP client, its an HTTP one.
HTTP is a slave protocol, where you need a master.
It's time to retire it, and design the web with an uncensorable protocol that does not loose memory.
The slave design of a web server need to go.
Think of it like managing a bunch of children. It's a great formula until they are old and mature enough. If you continue to control them as if children they effectively become your slaves.
I have tons of bandwidth and disk space. This is all that is required. If you unreasonably insist I should also be the one to serve every request it's game over for me. Not by technical limitation but by force.
Oh, I can rent my own web space? I haven't money for this and it has been tried before. Those sites are all gone. Everything is gone. I have tens of thousands of blog postings talking about stuff that doesn't exist anymore.
https://news.ycombinator.com/item?id=23182232 (AmA)
https://news.ycombinator.com/item?id=23194178 (TC article)
It is certainly not perfect, for example IPv4 has a scaling problem, but it still works damn well. It can be improved so that it can fit the needs it didn't have originally (ex: with IPv6), but it is not fixing, it is an adaptation.
None of the points in the "fix-the-internet" are things that are broken. They are new needs. Everything privacy-related for instance. The base internet protocol are all clear text, then we started encrypting secrets like passwords, then personal data, then full end-to-end encryption. Decentralization went from providing a reliable link between two computers, to services that are fully accessible even when part of it is offline, to services that simply can't be shut down.
There is no fix here, it is just building stuff on top of an internet that works better than I ever hoped for. Good, but calling it "fixing" is a bit pretentious.
When is this nonsense going to end? How about just “Founders Pitch Practice”? If you want to fix something, Mozilla, maybe start by fixing your bigotry.
When a critical mass of people start treating women (as a group) with the same respect and dignity as men are treated (as a group) would be a good start, no? Or are we post-sexism now?
An organization that uses that kind of violence against political opponents will never "fix" the internet.
We have 3 different offerings for this Summer. All incubator style, meaning you meet weekly or biweekly with mentors and we really try to help drive you from point A to point C.
1. $75k investment in a startup. MUST be serious about wanting to build something awesome and put in the hard work it takes to do so.
2. $16k funding in a much earlier stage project (idea stage / MVP stage). MUST be serious about commitment it takes to get to launch.
3. OPEN LABS: these are open to the entire community and you have access to the mentors. 10 min checkins each week & peer sessions. We've had TONS of amazing projects for our Open Labs in the Spring and we hope to see TONS more for the Summer.
In terms of MISSION and what we're looking for:
We started this new incubator out of Mozilla in order to work with & invest in developers, startups, and technology enthusiasts who are building things that will shape the internet and have a positive impact without needing to hyper focus on the bottom line. Projects, apps, & technologies that will be huge and a big part of the internet ecosystem while also being in line with Mozilla's mission of an open web & ethos of "Privacy over Profit".
Comment below if you have any questions about the programs and we'll be happy to answer them.
Personally I don't think so, & also why I don't take them seriously. They seem to be more concerned with waving their "For The People" flag, than actually trying to change anything. If they were serious about "fixing the internet" they'd swallow their pride, transition FF to Chromium, & essentially become something akin to an activist shareholder within Chromium.
However I disagree with your view of the ecosystem. The web needs stability and consistency more than anything else. FF switching to Chromium would help with that. So many people have this knee jerk reaction of Chromium = Chrome = Google having total control. But don't understand the only reason Google has had this much control over Chromium is because no other major vendor used it. They were the biggest kid on the street. But now that MS moved in a few doors down, that's no longer true. Google has to acknowledge MS in a way they never did with Opera, Vivaldi, Brave, etc.. And the same thing would happen if FF switched to Chromium.
Idk about you, but having 3 of the 4 biggest vendors all being forced to collaborate and implement solutions supported by at least 1 of the others, is 1000000x better than having each do their own thing. You effectively go from a monarchy to some form of democracy.
Mozilla destroyed itself due to terrible and politicized leadership.
I just hope that a Rust foundation is created soon, before the language goes down together with Mozilla.
- A one-shot, partially Turing complete scripting language would be nice. We should be able to include external scripts, position elements, perform basic computation, and then have a guarantee that the script has finished and will no longer run and use resources. Javascript is fine, but the user would have to give permission for it to run setTimeout(), setInterval() and similar ongoing operations. Discussion needed about scripts running post-load, like for clicks are keeping elements positioned (possibly a user-settable instruction, memory or time limit).
- I'd vote to largely retire frontend model-view-control (MVC) patterns. The browser is already the view, so I find most of the approaches introduced by frameworks like Angular.js to be at the very least superfluous, but potentially misguided or even hazardous. See Intercooler.js for declarative alternatives to what we have now.
- Instead of hand-rolled scripted pages, browsers need a basic component or window metaphor, in which all GUI elements descend from a base element, can be nested, and have a full life cycle that can be customized at each step. The user would have to give permission to override builtin elements like <body>, <input>, etc, but elements could descend from them with overrides. Discussion is needed about security. For what it's worth, I just don't think that the various approaches to building components have panned out. I feel like there is a common abstraction here behind things like React that could be generalized.
- On that note, the way we include remote resources on the web is broken. We need a way to choose whether the HTML for a component is written in-place, or as a URL (or other address) that includes the remote definition. Similar to if inline data: base64 images correspond to our current markup, then we could replace a block of text with an href to allow reuse. Like <div href=“example.com/mydiv.html”>, and mydiv.html would hold <div>something<span>something</span></div> or whatever. See server side includes (SSI) for an early failed example of this.
- Individual requests need to go away and be replaced by a content-addressable caching system where the hash is the address, so we can go back to including individual global sources of truth for things like jQuery or Bootstrap files, rather than maintaining copies of those includes on our own server. See subresource integrity (SRI) as a starting point.
- It may be time to scrap build systems. I'd like to see the world take a step back, and consider architecting a web where more protocols and formats like UDP or Markdown "just work" and don't need to be transpiled to whatever brittle/proprietary-infested conventions are available. Think of this as how compilers can provide generators by altering the code to work in runtimes that don’t have generators. That concept could be extended to everything, so that the server/browser is able to provide features that aren’t yet mainstream, by perhaps running portions of our build pipelines directly under the hood. Basically a turing complete language should be able to run anything, but right now we have wide swaths of computer science blocked to us for security and other reasons, but usually because a proprietary player like Microsoft doesn’t want to implement something. We should be able to run Firefox within Edge, conceptually. And figure out what fundamental capabilities must be provided to allow for something so general. It could be as basic as a general-purpose sandbox where the user decides which system calls and data to al...
https://htmx.org
Where are the problems to be fixed? I can see glossy bios' of mentors and their song-and-dance about how great this is, but it seems a PR gimmick from Mozilla.
(Yes, IRC is like that.)
The cool factor can be as much "everyone is doing it" as "the cool kids are doing it".