This is the version with local translation of web pages. It looks like they translate between Bulgarian, Dutch, English, French, German, Italian, Polish, Portuguese, and Spanish.
I tried a French language page, and the translation to English seemed excellent. It was fun to watch the text change on the page as it was translated.
Finally, that extension for Google Translate was really clunky and it was weird that this is coming so late while Chrome had a polished integration for such a long time already.
Google absolutely has qualm about you using Firefox. Yes, they pay Moz money to keep them as the preferred search, but that's because it's the better of two options (no data at all from you, or at least your search data).
But if you need evidence that Google has qualms about you using other browsers, look at how poorly supported other browsers are for google services.
I don't claim to read minds or have inside information, but there's no doubt in my mind that this is largely a $450mil annual hedge against getting antitrust'd by the DOJ.
Goggle payments have done nothing but go up as FireFox market share has done nothing but go down --- to the point of insignificance, now around 3% world wide --- less than Opera.
Paying a huge sum of money for a miniscule market share doesn't provide any reasonable hedge --- as evidenced by the fact that Google is in the middle of an anti-trust lawsuit and FireFox has played virtually no role.
I would expect Google aren't paying based on market share, they're paying based on the number and quality of users they get access to. If market share goes down but total market size rises, and the cohort of users selecting Firefox have more disposable income (hence are more profitable for search advertising), paying more for them over time would be a rational decision.
Goggle payments have done nothing but go up as
FireFox market share has done nothing but go down
Your reasoning, then, is that Google is paying Firefox strictly for business reasons but they've decided to increase payments as Firefox's share has decreased. Because they're bad at business? Or some other strictly business-y reason(s)? I mean, okay, but please be explicit about this.
FireFox has played virtually no role [in Google's
antitrust case]
That suggests it's working.
The DOJ has antitrust beef(s) with Google, but monopolizing the browser market is not one of those beefs. We can't read the minds of the DOJ, but it strongly suggests Google's support of Firefox is working within that context.
It does not, obviously, give them global blanket immunity against all conceivable antitrust shenanigans -- were you mistakenly believing this is what I was suggesting?
Not sure where you’re getting your Opera figures from, but I’m seeing 4x Firefox than Opera, even with Firefox typically shimming away Google Analytics.
Right, but Firefox’s enhanced tracking protection mode blocks Statcounter’s JS (see https://github.com/disconnectme/disconnect-tracking-protecti... which is where, I believe, Mozilla source their ETP blocklist from). Given the userbase for Firefox these days, I’d be very surprised if Statcounter is even vaguely accurate.
It is very well possible that the preferred boxes to tick include you using Chrome and paying $450 million / year to Mozilla to recoup some of that money. Chrome is one of the more powerful planks in Google's control over the web, without Chrome they'd be fighting a continuous rearguard action, now they get to be the dictators. Think of the $450 million as a license to operate. Any users that end up using the product created using that $450 million are additional loss.
Yeah, the ideal outcome for Google would be paying Mozilla whatever amount they need to stay afloat while still having no market share. As you said, 450 is just the cost of doing business. Though I wonder how effective the monopoly "shield" would be if Firefox continues slipping in market share. Maybe they just need it to exist?
A difference that doesn’t matter to most people most of the time.
GP wants a translation service built in to the browser. Even if they appreciate it being local-only, coming back with “no, here’s some largely irrelevant technical difference that makes this better!” misses the point.
> A difference that doesn’t matter to most people most of the time.
This seems to be about the first time I see Mozilla knowing its own users better than a random HN user know Mozilla's users.
Mozilla/Firefox is meant for users who care about privacy, it's literally the first taglines you see when visiting the landing pages for Mozilla or Firefox.
So, with this newfound knowledge about who is using Firefox and who is using other browsers like Google Chrome, maybe now it makes sense why Firefox wanted local translations instead of sending them off somewhere else?
Are you implying that your stock Edge is feeding you altered text?
If so, you should speak to a professional, and I don’t mean a computer professional.
EDIT: someone who isn’t me should probably report the literal death threats in your bio.
Following the links, it might be using the Marian NMT framework under the hood? I'm wondering where specifically the model weights are coming from, and it would be interesting to read more about the training process[0].
I'm excited to see a more open translation system that seems to perform pretty well and runs locally.
It definitely needs to add support for more languages, and based on other models I've seen recently... I have to wonder if building dedicated models for each pair of languages is still the best choice. I believe SeamlessM4T just uses a single model (available in different sizes), and I have definitely seen that Whisper only uses a single model for multiple language pairs as well (although it was only specifically trained to translate into English). Similarly, virtually all of the LLMs are multilingual. It seems like a single model is able to learn shared insights that apply across languages, reducing the amount of training data needed for each additional language (or increasing the accuracy with the same amount of data), but I admit that I could be wrong. This has just been my perception of how things are developing.
> and it could make sense, since "watered-down" can be considered redundant.
I don't think that makes sense and isn't correct, a translation should be a translation without subjective copy editing. It may be redundant or not, but it's not the job of the translation tool/service/person to decide.
This is a fairly clear case, but there can be more subtle issues. For example, consider this sentence (that I made up):
We got in the car and drove across the island.
Google translate renders this in Spanish as follows:
Nos subimos al coche y cruzamos la isla.
"We got into the car and crossed the island."
This preserves the 'across' part but not the 'drove' part, which is left implicit. That is not an accident: in contrast to English, Spanish tends to be 'verb framed' rather than 'satellite framed' (https://en.wikipedia.org/wiki/Verb_framing). In this particular case the translation is arguably fine because it is obvious from context that the manner of crossing the island was by driving. I am not a native Spanish speaker, but I think that there would be no way to make the 'driving' part explicit in Spanish without it sounding redundant and odd.
I think it sounds a bit weird to use the verb 'conducir' like that (which is why Google translate chooses not to do so). In any case, there are many other such examples of Spanish tending to encode path in the verb rather than manner, even if this particular one doesn't work.
Beyond the simplest cases, all translations have “subjective copy editing”. If two languages were 1:1 with each other, they would be the same language. For an extreme example of how not 1:1 translation can be, go read up on Toki Pona, but this is a fundamental property of translation.
Just think about all the idioms that we use in English that would be totally baffling if translated literally.
I know not everything can be translated 1:1. But if in the cases it can, I don't think removing things "that could be considered redundant" is the right approach if you want to have correct translations.
But it actually can be correct to remove redundancy. In Spanish, a double negative is the correct thing to do in many cases, but it would translate poorly to English, even though it can be literally translated as such.
These decisions naturally crop up all over the place during translation. It feels pointless to make blanket statements like you did about the removal of “watered down”. Without knowing that specific language intimately, I can’t say that it is wrong, and this is a very simple case.
Good quality translation will always be somewhat subjective.
Yeh! A lot of words have dozens of meanings; many carry cultural meanings and shades, which can further depend on context. Some can be positive or pejorative depending on pronunciation or emphasis... etc. it's endless!
Do you speak/understand more than one language? Honestly wondering, because if you do, I don't think this would be something you would think to be possible.
You're confusing a transliteration with a translation. And your standard readily breaks down when vernaculars & idioms are used. If a certain idiom doesn't exist in language A do you substitute it for the equivalent in language B? Or do you persist in writing nonsense no language B speaker will ever understand?
> "Ursula von der Leyen elogia il compromesso sulle emissioni dei veicoli attenuato"
> Note: "Watered-down" is translated as "attenuato" in this context, which means "weakened" or "diluted." The term can be subjective, so ensure the tone and implication match the intent in the original context.
ChatGPT 3.5:
> "Ursula von der Leyen elogia il compromesso sulle emissioni dei veicoli ridotto"
And as so often with Mozilla has the translation-feature an awful user experience. It's not showing itself, unless you are on a site with a supported language. And supported means, the language-pack is installed. How are people supposed to discover this?
Additionally, the auto-discovery seems to appear only once per domain? I tested it on Wikipedia, and after the first article, it never shows up again. But I guess this is ok, after all that dialog is Fu*ing huge. Couldn't they choose something more subtle? It's their own browser, they should have all the ability to make proper integrations for something like this.
This is overall a really great feature, which could have great use for people, but as it's now, it's more likely to be removed again in a while, because "nobody is using it", like so often in the past.
You can disable that the dialog shows up, then it's just a subtle icon.
That it often doesn't show up at all is indeed a bit confusing. I can translate german to english or french to german, but english to anything doesn't seem to show up on english websites even thought it's an option when I'm on a non-english website.
Did they ever add actual vertical tab support? When they migrated to webextensions TreeStyleTabs was pretty hamstrung because it couldn't replace the actual tab UI anymore.
Using a different size for elements matching a :hover selector is how a lot of CSS gets this done.
The first result on Google (https://gist.github.com/olmstadfm/d6b4d37219e957d9cdcdb3ec88...) also seems to employ the :hover selector to accomplish hiding the tabs. I don't know how effective this particular script is, but it makes sense to use such a system.
Yep! This and a few userchrome.css modifications is all it took for me. A few css transitions and hidden elements and viola, it functions better than edge does.
You can always add vertical tab support. What you can't do is turn off the normal tabs across the top. So any side tab solution (I use TST) also needs some userChrome.css.
No, you still need to "hack" the tabbar away on your own. Back then the claim seems to be security-reasons, so no malicious addon can replace the tabbar for you, and manipulate what you see. Which seems fairs, removing the tabbar on your own is not very complicated.
And Tree Style Tabs still works very well today, the early bugs are all fixed now. In the meanwhile there appeared also some other vertical tabbars/trees, but none are really on TSTs level, so not sure how popular they are. So over all, there is little reason for Mozilla to bring a vertical tabbar on their own, especially as they are still missing support for a second sidebar.
How many tabs do you have open at the same time? I was at a few hundreds (with an auto-unload add on as well). I often get there while doing research on a new subject - open everything relevant as a tab, and slowly go through them.
(Yes, it’s a horrible workflow, but I haven’t found one that works better for me, or even remotely as well)
I don't know what's fast or slow is for you. But I can say in general I have no performance-problems with TST. I tested the new version of Sidebery, the last days, and performance-wise it seems around the same. Some difference is the animation-speed, which out of the box on TST is a bit slower than with Sidebery. But that's a configuration-detail.
I'd like to see that too, but even with Sidebery I have been using Safari more because of its tab groups (and I have found the extensions lacking). It seems Mozilla is not interested in that idea any more. Without tab groups, its a stark difference that feels like the olden days when you had IE (no tabs) and FF (tabs). You just can't go back if they are useful to you.
> The visibility of fonts to websites has been restricted to system fonts and language pack fonts to mitigate font fingerprinting in Private Browsing windows.
I tried EFF's version* and it also said my Firefox browser was unique: Your browser fingerprint appears to be unique among the 179,689 tested in the past 45 days.
However the fonts listed looked to be only system fonts so there's that.
The EFF version also tells you for each metric how (dis)similar you are from other users/devices. Or at least it’s supposed to - for one of them which was supposed to be a true/false boolean test, it said my browser revealed -1.00 bits of information and shared its value for that test with 1 out of every 1 device they saw.
> Your browser fingerprint appears to be unique among the 180,370 tested in the past 45 days.
Close down the window, switch VPN endpoint, and retry, and once again I'm unique.
> Your browser fingerprint appears to be unique among the 180,381 tested in the past 45 days.
Looking at the values it seems they are identical. How do they know that the two tests are the same user
If I try outside of private mode I see differences in DNT (set to true in private mode, false in not private mode) and in Hardware Concurrency (4 in private mode, 2 in non-private mode). The rest of the entries are the same.
Nethertheless
"Our tests indicate that you have strong protection against Web tracking."
That sounds about right. US+Canada population is about 5% of the world, and eastern time is just a slice of that. People in ET are more likely to be internet users than people in some other parts of the world, so that apparently brings the percentage back up to around 5%.
However it also says that 42% of users are on Firefox, which definitely suggests that this isn't a totally representative sample.
I'm guessing that most people who visits a website called "My browser fingerprint" with a domain of "amiunique.org" has at least a fleeting interest in privacy and making choices around that. And that group is most likely biased to using Firefox instead of Google Chrome, at least compared to a general section of the population.
It just says that 27% of the entries in their database are Linux. You could infer that Linux users are more likely to use the website. Which matches my understanding
Not allowing JS cuts them off from a ton of specific data points in exchange for one big general one. Honestly I think the trick is to present a different unique fingerprint each time. It doesn't matter if you standout so long as each session can't be linked to all your previous ones.
That seems a lot easier to pull off than trying to blend in perfectly with the crowd. I wouldn't bet on my efforts to account for everything that might uniquely ID me in a never ending game of Whac-A-Mole with fingerprinters. They seem to find new tricks all the time.
I'm pretty sure my system fonts are a near-unique set. https://www.amiunique.org/fingerprint indicates that 0.0% of the net has the same set of system fonts.
OTOH, I know that installing user fonts (e.g.: non-system wide) is supported on fontconfig (Linux, BSD, etc), and I've done it in the past... but is this even common enough to make a difference?
Edit: my set of fonts is unique and I only have 6 fonts installed.
I disable DoH on Firefox immediately because I am my own sysadmin (and network admin) and believe apps should use the DNS settings I’ve configured my device with. As such, archive.is has worked perfectly for me on all my devices :)
I wonder what the deal with Google Meet was. Did Firefox fix this? Did Google team fix this? If this is retroactively fixed for the last 3 versions it must be a google-side fix right?
Still crashing like crazy on my system. It started in 114 or 115. I've sent at least 50 crash reports, with detailed descriptions of what I had been doing in the first few. Not sure what else to do. After using Firefox exclusively since around 2006, I had to switch to another browser (which I am surprisingly happy with), because getting 10-20 crashes per day really interferes with the work.
I've heard Linux support was down to like one guy who is doing half of the work in his free time out of goodness of his heart. Even if it's not true, sure feels like it.
I'm guessing Manjaro (my immediate guess) or nVidia (much less likely). Which actually makes me really curious which distro it is, because the maintainer for the distro presumably should have caught it.
> I've heard Linux support was down to like one guy [...]
Linux support is down to you. It's down to all of us. Install rr (https://rr-project.org/) and capture a crash with it.
Then you can replay the crash, find out that it's actually crashing in your closed-source graphics driver, which will motivate you to switch to an open source driver and fix your issue.
Also, while you're at it, update your linux kernel and wayland. They've both had bugs that could manifest as random firefox crashes in the last several months.
The Firefox bug tracker is full of thousands of bugs with some sitting there for 10-15 years. I’d like to see way more focus on cleaning these up than new features and UI changes.
With translations built in, I have one less reason to use Chrome.
There are a few old open bugs, the oldest Firefox bug seemingly being 9 years old, and is about pdf.js not automatically rendering hyperlinks as clickable. There are a few UI/usability bugs in the list, but most old bugs seem to be meta bugs linking several blocking issues, or features being completed part by part.
Firefox may lack engineering power to keep up with Chrome (especially after firing all of those Rust devs), but in terms of bugs I don't think they're doing all that badly. It's not like fixing "Consider to process DOMLinkAdded event using animation frame callback + setTimeout(, 0)" is going to help them gain any market share.
When I tried to look, the results maxed out at 10,000 and the bug tracker stopped being able to respond or sort by date. I wouldn't call 10,000+ "a few" it sounds like you're downplaying it for some reason.
Should releases include hundreds of closed bugs? No. But there could be a couple dozen. Mozilla has the resources to do that. Look at their release notes. A handful of security fixes (good) and otherwise nothing about bug fixes. These are not the release notes of an organization prioritizing software quality.
Just discovered this by chance today in my FF/Linux, I'm still on 117
> Video Effects and background blur are now available to Firefox users on Google Meet! (Note: These effects have also been released retroactively to support Firefox versions back to Firefox 115.)
Now, does everyone know where the issue actually was? Giving that all of a sudden is retro-compatible, I would say at least a good chunk of the issue was on Google side
Most people believed it was on Google's side, possibly even on purpose to make people use Chrome vs. Firefox. It is no coincidence that the Firefox / Google Antitrust case kicked off two weeks ago and this magically started working better.
Read the issue I linked above. There seemed to be performance issues in Firefox, and the feature was turned off with a UA string test. At some point, spoofing the UA string made the feature work acceptably but broke other things. Then the feature was enabled, back to the versions where it worked on.
I tried Firefox with Google Meet today and found the performance to be, ehm, not to write mom about. Still, it's nice to have it and my computer is not really the latest and greatest these days (although Chrome works fine).
Yeah this was confusing since it does not reference functionality or ticket in the release notes page.
I found this ticket though: https://bugzilla.mozilla.org/show_bug.cgi?id=1703668
It seems like there were some cache improvements "Cache the World (the new a11y architecture)" and other performance improvements then a Firefox targetted roll out by Google following on from those improvements.
I guess Google Meet's team didn't want to offer users a choice that will slow down their web app so they didn't but since the performance got better they stopped blocking the feature to Firefox users with a minimum set of requirements?
But does Firefox typically have backports of security patches and others like performance improvements to some base versions?
Not related to the release notes, but I've recently discovered Chrome is unusable for my use case.
I have two M-processor based MacBooks in the house - one runs mullvad (WireGuard), and the other a company OpenVPN.
For some reason, both have major issues with Chromium-based software. I get ERR_NETWORK_CHANGED every couple of minutes and nothing can load. It kills XHR requests as well, so most web software becomes totally unusable. This also applies to Electron-based software like Discord and Bitwarden. Oddly, Slack seems unaffected so far but maybe it hides it / retries better.
I don't know what its problem is, but the issue does not affect WebKit or Gecko or any other software on the machine requiring networking. So I've dumped Chrome everywhere. I've spoken with several others that are affected by this but I have not found any bug reports mentioning this so I gave up.
I've been on Firefox for ages for personal use, but I can't even use Chrome at work anymore. I'm not sure how they expect people to use this browser in Enterprise with such a glaring issue.
Using the official wireguard client allegedly works, but I don't know how they handle ip assignment, which is the complicated part for commercial VPNs like that.
I'd speculate that the problem isn't related to the VPN connection, but to how the Mullvad client app handles the interface (especially if keepalives frequently fail and the interface keeps getting downed).
Even the official clean OpenVPN app has its own sense of up and down tunnels, which could cause similar problems. Wireguard, at least in its official linux implementation, is much lighter weight and if necessary reestablishes an active session on demand (whenever there's traffic) without any interface-level tweaks. That's probably invisible to applications, or at least much less visible even if there are low-level syscalls that can inspect the state of a wg tunnel interface.
Sadly, my case is the opposite. I use Firefox since the beginning so it is close to 20 years. I'm not going to use it as my main browser anymore as I'm working fulltime with PWAs and service workers and Firefox support of it (or lack of) is unjustifiable.
My new apps make use of modern HTML and JS APIs and I can't count on Firefox implementing them.
And Firefox solution is to ignore all developers that think this is a great platform compared to Play Store and App Store. Great way to lose market share.
I think you're justified in this case - this isn't about not supporting Chrome or the features it offers, just saying that it has some issues that make it not great as a browser for many end-users. This is one such case that hopefully has a resolution. But, as for any platform, I think it's fair to take advantage of additional things one platform gives you over the other. That's so long as you properly support the others still.
I use lots of modern APIs and CSS (on our own site) with no issues. I can only remember one site in the past month that had an issue in FF, and it was probably because of user agent checking. What API's are you referring to, and where did you come across them?
No, I'm currently on a monopolistic Canadian provider :)
I would wager IPv6 is likely on for most newer, provider-owned network equipment these days though. Others on different hardware and providers report similar.
Like I said though, I've been unable to identify a root cause, but Chromium does not like what's happening. At first I thought that my IP was being cycled by my VPN at these times and Chrome struggled to deal with it. I have no other leads, nor do I feel like I'm missing out XD
The local translation feature would finally be a good new Firefox feature again. A pity though the number of supported languages is so small and does not include at least Chinese still.
The actually had Russian in the beta add-on version. I wonder if there is still a way to add it back. It was the most useful language, but I agree about Chinese.
I think a huge chunk of web developers might be forced to switch to Firefox soon: Stripe’s API docs have become so big (and their asynchronous loading logic is so terrible) that Chrome regularly crashes for me jumping into section links in their api docs which made me curious to try it under Firefox. Happy report that slow DOM content loaded times aside, Firefox works great there!
I'm sure there are many people paid to spend a lot of time in meeting rooms, talking about UX and how users need the latest trend of single-page website with blackjack and hookers.
> Docs should be docs, not full blown apps
I actually started reading RFCs a few years ago and found that they are pretty easy to follow. And of course man pages are even more approachable. One just needs to get used to the fact that it is text-only, instead of a full sensory experience like modern websites.
I noticed something similar 10 years ago - large XML documents simply don't render in Chrome but Firefox had no issues. Surprised this is still the case that Chrome hasn't bothered to resolve.
It has been possible for quite some time. I remember a few years ago, Google Translate offered a offline translation option (for a specific language pair) on my Android that required a couple hundred of MBs of download.
It's possible with Neural Machine Translation models. Before being integrated into Firefox itself local translation was already available through the TranslateLocally add-on, see: https://browser.mt/
The translation is very nice. Hopefully they add more languages soon! Been waiting on a feature like this for some time. There are addons but they've been finicky for me.
Sounds like you were just updated from Firefox ESR 102 to 115. However, Firefox’s default theme hasn’t changed. It should still be “System theme - auto”, which selects a light or dark Firefox theme based on your OS’s dark mode setting.
FWIW, I think I just experienced a similar problem when updating Firefox Nightly from 119 to 120 (on macOS): I see light mode instead of my non-default theme, even though about:addons shows it is still selected.
I shared your problem report with a Firefox engineer. (I'm a TPM on the Firefox Android team.) If you email me your Firefox's about:support information (cpeterson at mozilla dot com), I will pass it along to the engineer.
Had you already been using ESR 102 or 115 on Win 7? Or were you recently switched from the Firefox Release channel to ESR 115 (when Mozilla EOL'd Win 7 support in Firefox versions > 115)?
Can anyone say how FDLIBM improves protection against fingerprinting? They sort of gloss over this completely but I bet it has a very interesting explanation.
191 comments
[ 4.5 ms ] story [ 165 ms ] threadI tried a French language page, and the translation to English seemed excellent. It was fun to watch the text change on the page as it was translated.
Google has no qualms about sucking up all your data.
Google also has no qualms about you using Firefox. They pay Mozilla $450 mil a year to insure it stays this way.
But if you need evidence that Google has qualms about you using other browsers, look at how poorly supported other browsers are for google services.
Goggle payments have done nothing but go up as FireFox market share has done nothing but go down --- to the point of insignificance, now around 3% world wide --- less than Opera.
Paying a huge sum of money for a miniscule market share doesn't provide any reasonable hedge --- as evidenced by the fact that Google is in the middle of an anti-trust lawsuit and FireFox has played virtually no role.
The DOJ has antitrust beef(s) with Google, but monopolizing the browser market is not one of those beefs. We can't read the minds of the DOJ, but it strongly suggests Google's support of Firefox is working within that context.
It does not, obviously, give them global blanket immunity against all conceivable antitrust shenanigans -- were you mistakenly believing this is what I was suggesting?
These obviously float up and down some over time but Firefox has been consistently under 3% globally for a while now.
See for yourself here: https://www.whatsmybrowser.org/
Different battleground, but could very well end up having a major impact on this.
- massive anti-trust lawsuit
- paying $450 million / year to Mozilla
- you using Chrome
- you using FireFox
It is very well possible that the preferred boxes to tick include you using Chrome and paying $450 million / year to Mozilla to recoup some of that money. Chrome is one of the more powerful planks in Google's control over the web, without Chrome they'd be fighting a continuous rearguard action, now they get to be the dictators. Think of the $450 million as a license to operate. Any users that end up using the product created using that $450 million are additional loss.
https://addons.mozilla.org/en-US/firefox/addon/traduzir-pagi...
This works exactly like the respective Chrome feature.
>this is coming so late
Because it isn't the same thing. This is done locally. (Not the linked extension but the Firefox built-in.)
GP wants a translation service built in to the browser. Even if they appreciate it being local-only, coming back with “no, here’s some largely irrelevant technical difference that makes this better!” misses the point.
This seems to be about the first time I see Mozilla knowing its own users better than a random HN user know Mozilla's users.
Mozilla/Firefox is meant for users who care about privacy, it's literally the first taglines you see when visiting the landing pages for Mozilla or Firefox.
So, with this newfound knowledge about who is using Firefox and who is using other browsers like Google Chrome, maybe now it makes sense why Firefox wanted local translations instead of sending them off somewhere else?
EDIT: someone who isn’t me should probably report the literal death threats in your bio.
(I mean no disrespect toward the experiences you're articulating, which have obviously been deep and impactful.)
https://news.ycombinator.com/newsguidelines.html
I'm excited to see a more open translation system that seems to perform pretty well and runs locally.
It definitely needs to add support for more languages, and based on other models I've seen recently... I have to wonder if building dedicated models for each pair of languages is still the best choice. I believe SeamlessM4T just uses a single model (available in different sizes), and I have definitely seen that Whisper only uses a single model for multiple language pairs as well (although it was only specifically trained to translate into English). Similarly, virtually all of the LLMs are multilingual. It seems like a single model is able to learn shared insights that apply across languages, reducing the amount of training data needed for each additional language (or increasing the accuracy with the same amount of data), but I admit that I could be wrong. This has just been my perception of how things are developing.
[0]: Some training info here, it looks like: https://github.com/browsermt/students/tree/master/train-stud...
Firefox translation is wrong: "Ursula von der Leyen elogia il compromesso sulle emissioni dei veicoli annacquati"
It is the compronise that is "watered-down" not the vehicles.
Google translation is better: "Ursula von der Leyen elogia il compromesso annacquato sulle emissioni dei veicoli"
Both Deepl and Papageo don't translate "watered-down" and it could make sense, since "watered-down" can be considered redundant.
I don't think that makes sense and isn't correct, a translation should be a translation without subjective copy editing. It may be redundant or not, but it's not the job of the translation tool/service/person to decide.
> Subimos al coche y condujimos a través de la isla
> We got in the car and drove across the island.
Making it explicit that you drove across the island :)
If I had to translate "We drove across the island", and make the car explicit without the first clause giving context, I'd say:
> Cruzamos la isla en el coche
> We crossed the island in the car
Just think about all the idioms that we use in English that would be totally baffling if translated literally.
These decisions naturally crop up all over the place during translation. It feels pointless to make blanket statements like you did about the removal of “watered down”. Without knowing that specific language intimately, I can’t say that it is wrong, and this is a very simple case.
Good quality translation will always be somewhat subjective.
It's a genuinely hard problem. Even the idea of "correct translation" is slippery.
> "Ursula von der Leyen elogia il compromesso sulle emissioni dei veicoli attenuato"
> Note: "Watered-down" is translated as "attenuato" in this context, which means "weakened" or "diluted." The term can be subjective, so ensure the tone and implication match the intent in the original context.
ChatGPT 3.5:
> "Ursula von der Leyen elogia il compromesso sulle emissioni dei veicoli ridotto"
Additionally, the auto-discovery seems to appear only once per domain? I tested it on Wikipedia, and after the first article, it never shows up again. But I guess this is ok, after all that dialog is Fu*ing huge. Couldn't they choose something more subtle? It's their own browser, they should have all the ability to make proper integrations for something like this.
This is overall a really great feature, which could have great use for people, but as it's now, it's more likely to be removed again in a while, because "nobody is using it", like so often in the past.
That it often doesn't show up at all is indeed a bit confusing. I can translate german to english or french to german, but english to anything doesn't seem to show up on english websites even thought it's an option when I'm on a non-english website.
Well, it's a beta, so I assume this will change.
And then? This would make it even worse, people will easily forget some random icon sitting somewhere.
You need to go to the settings, search for Translations, then click install all packs.
My only complaint about it really is that it has no Asian languages.
Finally! That was fairly frustrating.
The first result on Google (https://gist.github.com/olmstadfm/d6b4d37219e957d9cdcdb3ec88...) also seems to employ the :hover selector to accomplish hiding the tabs. I don't know how effective this particular script is, but it makes sense to use such a system.
Brilliant! Thank you!
https://github.com/piroor/treestyletab/wiki/Code-snippets-fo...
Tabs like History and Bookmarks won't hide but TST does.
https://github.com/mbnuqw/sidebery/wiki/Firefox-Styles-Snipp...
I've tried to comment things as I go. Hope you find it helpful!
- https://github.com/black7375/Firefox-UI-Fix/wiki/Show-Off-Yo...
And Tree Style Tabs still works very well today, the early bugs are all fixed now. In the meanwhile there appeared also some other vertical tabbars/trees, but none are really on TSTs level, so not sure how popular they are. So over all, there is little reason for Mozilla to bring a vertical tabbar on their own, especially as they are still missing support for a second sidebar.
Did TST improve on this front?
(Yes, it’s a horrible workflow, but I haven’t found one that works better for me, or even remotely as well)
Nice, I wonder if this will effect results here: https://www.amiunique.org/fingerprint
Update: Still unique and the results show 215 Fonts which are 0.0% unique so I'm not sure this is working for me.
However the fonts listed looked to be only system fonts so there's that.
*https://coveryourtracks.eff.org/
Sure I'm unique. I try it in private mode.
> Your browser fingerprint appears to be unique among the 180,370 tested in the past 45 days.
Close down the window, switch VPN endpoint, and retry, and once again I'm unique.
> Your browser fingerprint appears to be unique among the 180,381 tested in the past 45 days.
Looking at the values it seems they are identical. How do they know that the two tests are the same user
If I try outside of private mode I see differences in DNT (set to true in private mode, false in not private mode) and in Hardware Concurrency (4 in private mode, 2 in non-private mode). The rest of the entries are the same.
Nethertheless
"Our tests indicate that you have strong protection against Web tracking."
However it also says that 42% of users are on Firefox, which definitely suggests that this isn't a totally representative sample.
Hmm, according to this site, 27.87% of people use Linux and 42.47% use Firefox.
Those are great stats. Alas I don't believe them.
It showed me a 0.xx% (i.e. a number less than 1%) just due to the "no JS" alone. So not allowing JS is a strong identifier on its own.
That seems a lot easier to pull off than trying to blend in perfectly with the crowd. I wouldn't bet on my efforts to account for everything that might uniquely ID me in a never ending game of Whac-A-Mole with fingerprinters. They seem to find new tricks all the time.
I'm pretty sure my system fonts are a near-unique set. https://www.amiunique.org/fingerprint indicates that 0.0% of the net has the same set of system fonts.
OTOH, I know that installing user fonts (e.g.: non-system wide) is supported on fontconfig (Linux, BSD, etc), and I've done it in the past... but is this even common enough to make a difference?
Edit: my set of fonts is unique and I only have 6 fonts installed.
I've heard Linux support was down to like one guy who is doing half of the work in his free time out of goodness of his heart. Even if it's not true, sure feels like it.
Have you tried running it with a freshly created profile?
I'm guessing Manjaro (my immediate guess) or nVidia (much less likely). Which actually makes me really curious which distro it is, because the maintainer for the distro presumably should have caught it.
I would first try it out again with a new profile. Have you tried the flatpak instead of the distro version to iron out a packaging issue?
Linux support is down to you. It's down to all of us. Install rr (https://rr-project.org/) and capture a crash with it.
Then you can replay the crash, find out that it's actually crashing in your closed-source graphics driver, which will motivate you to switch to an open source driver and fix your issue.
Also, while you're at it, update your linux kernel and wayland. They've both had bugs that could manifest as random firefox crashes in the last several months.
And someone closing new reports of same bug as duplicate of one that is closed, despise it still happening...
There are a few old open bugs, the oldest Firefox bug seemingly being 9 years old, and is about pdf.js not automatically rendering hyperlinks as clickable. There are a few UI/usability bugs in the list, but most old bugs seem to be meta bugs linking several blocking issues, or features being completed part by part.
Firefox may lack engineering power to keep up with Chrome (especially after firing all of those Rust devs), but in terms of bugs I don't think they're doing all that badly. It's not like fixing "Consider to process DOMLinkAdded event using animation frame callback + setTimeout(, 0)" is going to help them gain any market share.
When I tried to look, the results maxed out at 10,000 and the bug tracker stopped being able to respond or sort by date. I wouldn't call 10,000+ "a few" it sounds like you're downplaying it for some reason.
Should releases include hundreds of closed bugs? No. But there could be a couple dozen. Mozilla has the resources to do that. Look at their release notes. A handful of security fixes (good) and otherwise nothing about bug fixes. These are not the release notes of an organization prioritizing software quality.
https://bugzilla.mozilla.org/buglist.cgi?f1=cf_status_firefo...
> Video Effects and background blur are now available to Firefox users on Google Meet! (Note: These effects have also been released retroactively to support Firefox versions back to Firefox 115.)
Now, does everyone know where the issue actually was? Giving that all of a sudden is retro-compatible, I would say at least a good chunk of the issue was on Google side
https://www.npr.org/2023/09/12/1198558372/doj-google-monopol...
https://blog.mozilla.org/en/products/firefox/google-meet-fir...
They worked with the Google team to resolve it.
Not related to the release notes, but I've recently discovered Chrome is unusable for my use case.
I have two M-processor based MacBooks in the house - one runs mullvad (WireGuard), and the other a company OpenVPN.
For some reason, both have major issues with Chromium-based software. I get ERR_NETWORK_CHANGED every couple of minutes and nothing can load. It kills XHR requests as well, so most web software becomes totally unusable. This also applies to Electron-based software like Discord and Bitwarden. Oddly, Slack seems unaffected so far but maybe it hides it / retries better.
I don't know what its problem is, but the issue does not affect WebKit or Gecko or any other software on the machine requiring networking. So I've dumped Chrome everywhere. I've spoken with several others that are affected by this but I have not found any bug reports mentioning this so I gave up.
I've been on Firefox for ages for personal use, but I can't even use Chrome at work anymore. I'm not sure how they expect people to use this browser in Enterprise with such a glaring issue.
Using the official wireguard client allegedly works, but I don't know how they handle ip assignment, which is the complicated part for commercial VPNs like that.
Changing locations sometimes helps.
Even the official clean OpenVPN app has its own sense of up and down tunnels, which could cause similar problems. Wireguard, at least in its official linux implementation, is much lighter weight and if necessary reestablishes an active session on demand (whenever there's traffic) without any interface-level tweaks. That's probably invisible to applications, or at least much less visible even if there are low-level syscalls that can inspect the state of a wg tunnel interface.
https://www.reddit.com/r/chrome/comments/14xqh92/chrome_help...
and this bug: https://bugs.chromium.org/p/chromium/issues/detail?id=147633...
My new apps make use of modern HTML and JS APIs and I can't count on Firefox implementing them.
I would wager IPv6 is likely on for most newer, provider-owned network equipment these days though. Others on different hardware and providers report similar.
Like I said though, I've been unable to identify a root cause, but Chromium does not like what's happening. At first I thought that my IP was being cycled by my VPN at these times and Chrome struggled to deal with it. I have no other leads, nor do I feel like I'm missing out XD
https://blog.mozilla.org/en/products/firefox/google-meet-fir...
Docs should be docs, not full blown apps
I'm sure there are many people paid to spend a lot of time in meeting rooms, talking about UX and how users need the latest trend of single-page website with blackjack and hookers.
> Docs should be docs, not full blown apps
I actually started reading RFCs a few years ago and found that they are pretty easy to follow. And of course man pages are even more approachable. One just needs to get used to the fact that it is text-only, instead of a full sensory experience like modern websites.
hello :teapot: -> bonjour :baguette:
HN unfortunately removes the emojis.
hello :maple leaf: -> bonjour :fleur-de-lis:
Wait how is it even possible without having to download gigabytes of datas ?
I'd also like to see FF support Scroll Driven Animations and View Transitions.
Available behind a feature flag
I just upgraded my ESR. Now the dark mode is the default. Worse, the light mode would not work.
Sigh... is there any quality manager left at Mozilla ?
People, please, don't change the UI without a very good reason. You look like idiots wanting fame.
Update: I hit bug https://bugzilla.mozilla.org/show_bug.cgi?id=1855151 in Nightly 120. The timing of two theme bugs after updating Firefox is suspicious, but this bug would not affect ESR 115.
Had you already been using ESR 102 or 115 on Win 7? Or were you recently switched from the Firefox Release channel to ESR 115 (when Mozilla EOL'd Win 7 support in Firefox versions > 115)?