Potentially bad news for those of us who prefer not to be assimilated into the Google hive mind: Mozilla seems to be on the rocks. Citing revenue problems, the maker of Firefox and other popular tools will be trimming 250 employees, about a quarter of its workforce, and shuttering its office in Taipei. CEO Mitchell Baker specifically mentioned that “development tools, internal tooling, and platform feature development” in the Firefox team would see reduced investment. Like a lot of companies do these days, she managed to blame COVID-19 for the company’s woes. That seems a little specious to us, but whatever the reason for the downturn in revenue, here’s hoping that Mozilla can keep Firefox alive.
Speaking of our evil overlords, looks like it another one of those “oopsies” moment for Google when it “accidentally” activated some of its smart speakers to listen into household events without the wake word. In this case, a user reported getting a text about a smoke alarm going off in their home. The alarm was not a surprise, since the user was cooking at the time, but the notification was, since they didn’t opt into that particular service. Google’s response was that an update pushed to the speaker accidentally activated that feature, a situation that they say has since been rectified. To be clear, this is an interesting feature and one of the more compelling use cases we’ve seen for a smart speaker, but it’s something we’d certainly want to sign off on before it’s activated. Yes, accidents happen, but these kinds of accidents seem to happen to Google an awful lot lately.
We’ve probably all had the experience over the last few months of being in public when the urge to cough hits. Masked or not, you struggle to fight back the tickle, lest someone hear you and think you’re infected. But now it’s possible for a computer to cough-shame you, thanks to a deep learning cough locater. The model was trained against recordings of people coughing and is coupled to an acoustic camera, which identifies the cougher with a bounding box and a contour image of the cough which looks for all the world like a virtual cloud of microbes. It’s genuinely interesting technology, sort of the public health version of ShotSpotter, but we doubt it’ll be of much practical use in public; if you want to find someone who has just coughed, someone acting like this will likely already be on the case.
Modern jet fighter technology is advancing rapidly, so much so that the forces they can apply during extreme maneuvers can quickly be lethal to pilots. Given that humans aren’t likely to evolve the ability to resist turning into a puddle of goo under high g-forces anytime soon, fighters of the future will likely incorporate AI of some sort. To prepare for that eventuality, the Defense Advanced Research Projects Agency (DARPA) is running some AI fighter competitions this week that really look interesting. Dubbed Alpha Dogfight Trials, the challenge starts with simulated dogfights between AI systems. The winner of those rounds will go up against a human pilot in the final match, which will be streamed live with commentary and multi-screen coverage. You need to register to get in on the action, and time is limited.
And finally, let these three words roll around in your head for a minute: robotic chameleon tongue. It’s actually nowhere near as disturbing as it sounds, since at its heart the “Snatcher” is actually just a beefed-up tape measure. Designed for remote retrieval tasks, the Snatcher can shoots it steel proboscis out almost a meter in just 600 milliseconds. Its designers envision uses for it on drones, but we can see it potentially being deployed on satellites too. It shouldn’t be too hard to build something like this at home, either.
I would not want to be poked in the eye with a high speede tape measure!
Mozilla’s history shades of Netscape.
” Given that humans aren’t likely to evolve the ability to resist turning into a puddle of goo under high g-forces anytime soon, fighters of the future will likely incorporate AI of some sort. ”
The movie Stealth coming to life.
Not to mention that one of the most expensive parts of a fighter jet is the equipment to keep the “meat fresh”.
One very long term problem in Firefox, going back to at least Netscape 3, is with submitting form data. You enter data into a site, click submit, send, OK, etc. And you wait, and wait, wait some more. Either it will sit there for hours until you give up and kill it, or it’ll just quit. Was the form submitted or not? Maybe it was, maybe it wasn’t. The only way to tell is if the site confirms it via some other method like sending you an email or refreshing the page shows what you sent or posted.
In the Time of Netscape and Early Firefox Period you could tell when it had failed because it would pop up an error message with the (false) claim “Document contains no data.” At some point the Mozilla people just removed that BS error message.
This problem persisted through several “complete rewrites” of the browser code. I assume how such a problem can persist even when starting a new version from scratch is whomever was responsible for the error simply repeated the same mistakes in the “all new” code, even if it wasn’t a copy and paste.
What definitely pointed to Netscape/Firefox as the problem was that in *any other browser* any page with data input *would submit instantly* while Netscape or Firefox was twiddling its digital thumbs then claiming your document contained nothing to send.
Firefox still has this problem in version 79. I did a correction for the Google Maps info on a local business. Firefox did get the data through, after being left to sit there for an hour or so without any confirmation it had worked. I’ve also had issues where it fails to make the submit button active so a data entry form doesn’t work at all. In both cases, in any other browser, it worked properly and instantly.
When a thing works for everyone else and yours doesn’t work – the problem is with your product, not everyone else’s. That was often their reason for not fixing the problem. “We’re doing it right, it’s all the websites where it fails that are wrong.”
Such programmer arrogance is really annoying, especially when it’s easy for pretty much anyone to demonstrate something they’ve screwed up. Insist that your broken code isn’t broken for long enough, you’ll lose users.
Another item with Firefox. Remember Firefox 4.0? It didn’t support scroll device input at all. Quite the annoyance to be able to easily scroll around in everything, then go to Firefox and be taken back to the old click, drag, click, drag over and over. The next update “fixed” that but only for mousewheels and only for vertical scrolling.
What I gathered from that is Mozilla made the bughouse nuts decision to *not use the operating system HID API*. There is no expletive reason what so expletive ever for a *program* to handle a hardware interface that is already being provided by the operating system. They seem to still be doing it because frequently with Firefox running (and only with Firefox running) at times my Logitech M570 will hardly respond, it jitters around, stops moving the cursor, the cursor will stop just short of what I want to click on etc.
This inter-connection with hardware which a web browser should not have also extends to storage access, especially optical discs. For a long time, on many different computers, Firefox has at times just slowed to a crawl when other apps are doing lots of data transfer, especially to or from an optical drive. Everything else running will be working fine. Firefox should also work fine, especially when the drive being heavily used isn’t C: where its cache is – which is the only sort of heavy disk use that should have any possibility of slowing down Firefox.
When things began to go downhill was with the Netscape version where they took away the option to have text only buttons on the toolbar. Users were stuck with a choice of big icons with text labels, or just big icons. If you wanted the good old Netscape UI then you had to use Internet Explorer for Macintosh. That was pretty much a straight up clone of Netscape, unlike IE for Windows, which the Netscape people foolishly went chasing after, doing unpopular crap like changing Location: to Address: and dropping the text only buttons (which IE Mac had until its end).
They quit listening to their users, especially when the users complained about changes for the worse and functions that frequently didn’t.
I must add that Firefox for Android comes in handy at times because it can force websites to deliver the desktop versions of sites and block Android from forcibly switching to the app version of a site. Want to watch a video on YouTube the website instead of the YouTube App? Firefox will make it so.
Why this is a good thing is when a search turns up something deep within a forum thread or some way drilled down page on a site, Firefox *will take you there*. Other browsers will forcibly switch to the mobile version of a site and since the site structure is usually totally different, you find yourself abandoned on the doorstep like a baby in a basket, without a clue where you are or how you got there. Good like using the site’s internal search to find the info you were after, because they universally suck.
If you are a webmaster setting up a site search, and especially if it’s a technology site, ensure that ALL the technical terms and product names used by the company *are in the database of words to be indexed*. There is nothing so irritating as to be looking right at a support forum post containing something like Funkinator 200 then you enter Funkinator 200 into the search box and get back “No matches found”, or you get “matches” only for the word Funkinator because “200” is too short of a word so it’s not in the database. “No, I need info on the model 200, not the 185, 2300, and 1234!”
i can recommend newpipe for android. opens the youtube links in its own client where you even can download the videos, play background, no login required,..
In mobile Chrome, you can watch YouTube without app by using incognito mode.
What the heck
Could you be more specific?
“And finally, let these three words roll around in your head for a minute: robotic chameleon tongue. … *snip* …It shouldn’t be too hard to build something like this at home, either.”
Should be possible to rig one with a party blower, https://en.wikipedia.org/wiki/Party_horn with a glob of tar or freshly heated hot glue on the tip or something, to fulfill all those prophecies of meeting with a sticky end.
Mitchell Baker, CEO of Mozilla, is a woman, not a “he”.
Thanks for catching that, fixed.
Hey don’t you go gendering people you fascist! (Just kidding, I ain’t no SJW)
“The alarm was not a surprise, since the user was cooking at the time”. Ouch.
Please not that Mitchell Baker is a she! How come you didn’t know that? It has been common knowledge ever since she got the job!
EDIT: “Please note ….” And ignore the report, I clicked the wrong button!