But it makes me question why "the browser" is apparently still the inevitable platform of the future.
In order for a PWA to be normal and usable, it must be available offline, open in a window without browser chrome, have similar performance to a native application, be launchable via a shortcut on the host OS, and respond to the mouse and keybaord shortcuts the way you'd expect. I think I've just described... an Electron app?
It's cool that this kind of thing can run in a web browser. With no install hurdle, it's much easier to convince people to try it out, and it's cross platform. Beyond that I can't really think of any advantages to having it run in the browser.
If what's lacking is an easy way to try software, I can't help but imagine lots of ways this could be addressed that would be much more pleasant to use than loading PWAs. Right now I can't seriously see myself enjoying using a PWA for work.
I say this having recently finished several large design projects in Figma, which is apparently a gold standard success story for browser based apps. Despite the years of development and herculean engineering efforts, I can still feel the browser jank. I begrudgingly open the thing in chrome, as it completely chokes in Firefox. It still chokes on moderately sized canvases, moving things is slow and laggy compared to native apps, keyboard shortcuts sometimes don't work or keys get stuck in a weird pressed or unpressed state, loading is slow, elements pop-in over tens of seconds.
I know I'm an old man yelling at clouds at this point, I'm just disappointed that we seem to be going backwards in performance and usability of software.
> In order for a PWA to be normal and usable, it must be available offline, open in a window without browser chrome, have similar performance to a native application, be launchable via a shortcut on the host OS, and respond to the mouse and keybaord shortcuts the way you'd expect. I think I've just described... an Electron app?
You've just described a PWA. You can install them as a host OS shortcut, they run without browser chrome, should have performance equal to Electron.
Also if you really want extra bloat and faff, any PWA is trivial to turn into an Electron app.
Most of PWA criticism is based on misunderstanding PWA capabilities.
I'm reassured that my comparison to Electron was not far off. My point was, if a PWA is usable when all the above is implemented, what you end up with is very similar to a self-updating Electron app, just one that may or may not rely much more heavily on the network at runtime and might work offline (or not).
I admit I didn't realize creating a shortcut to a PWA was already supported as it's not pushed very hard. In Chromium it's buried under dots, "Cast, Save and Share" (which is a bizarre mashup of disparate functions), and finally "Install page as app".
The window that loads still has browser chrome, in the form of a back, refresh, and three dots button. As soon as you navigate somewhere, even within the same app, the url bar appears again, but you can't edit it. It seems that to be able to always hide this bar, you'd need a way to differentiate between "internal" links that should navigate within the page, and external links that should open in a browser.
I tried turning off my internet, and neither figma nor openDAW showed anything more than a blank page, which confirms my feeling of uncertainty around PWAs, namely, how do you know what will actually work offline. It feels fragile, like if I reset my browser or my clear my cache, my installed applications will disappear. I'm not sure I'm comfortable with the blurring of the lines between bookmarks and installed applications.
All this is of course addressable with a lot more infrastructure and work from browser and OS makers. To me it seems like a lot of development to end up with something that behaves a lot like Electron, with the added easteregg of being able to access applications in a browser, without intending that anybody actually do so.
> I tried turning off my internet, and neither figma nor openDAW showed anything more than a blank page, which confirms my feeling of uncertainty around PWAs, namely, how do you know what will actually work offline.
Figma nor OpenDAW don't seem to be configured as offline PWAs.
> All this is of course addressable with a lot more infrastructure and work from browser and OS makers.
The problems you encountered are mostly solved in PWA APIs already, at least for Chromium based browsers. There is some variation in some features between browsers and OSs (Safari and iOS are particularly bad).
> To me it seems like a lot of development to end up with something that behaves a lot like Electron, with the added easteregg of being able to access applications in a browser, without intending that anybody actually do so.
PWAs are e.g. easier to install, have smaller footprint, are more portable and are a lot more secure. Not sure what you mean by "accessing applications in a browser". PWAs can't access anything a normal website can't.
> why "the browser" is apparently still the inevitable platform of the future.
Because it doesn't require trust. The browser actually got the permission model (almost - but there are extensions) right. I can safely open this and not worry about security.
It is not defense of TikTok to point out that all platforms are practically identical: centralized, closed source networks that collect ridiculously detailed data about millions of people, and are in a position to process that data for all kinds of fun stuff like surveilling people's physical movement, creating recommendation algorithms, performing sentiment analysis, and using those tools to try to manipulate populations, influence public thought, and sway elections.
There is no behavior that TikTok exhibits that isn't equally applicable to every major social media.
I don't particularly care whether right now, the CCP happens to use that manipulative potential for its own ends more than the US does. It shouldn't be hard to take a principled stand against dystopian surveillance.
I don't understand people who correctly point out that TikTok is a "vector for influence of public opinion", but somehow think that's only a bad thing if it's controlled by a "geopolitical adversary".
Baby steps! Let's do both. Let's rid ourselves of a foreign adversaries influence, then let's ALSO get more competition and reduce big techs influence. Why not both?
> I don't understand people who correctly point out that TikTok is a "vector for influence of public opinion", but somehow think that's only a bad thing if it's controlled by a "geopolitical adversary".
By your implied logic, the US should allow Chinese police to function on US soil as the US already had US police on US soil.
As I read more and more comments like GP’s
In this thread, I keep getting this vision of a man in France in 1940 welcoming nazi tanks to his town because “well, we have a military too, how is this any different, they’re just a geopolitical adversary” and “well, we had a corrupt politician once last year so our government is just as bad”.
Can't remember the name (if it has one), but there's a linguistic phenomenon where words take on opposite meanings from the original. The "Peacemaker" was a missile, "literally" now means "figuratively", "awful" used to mean "awe-inspiring". And apparently "bare metal" now means "runs on an operating system".
OTOH, technical fields develop jargon specifically as a solution to the problems created by semantic drift in vernacular vocabulary.
Sure, languages evolve, but that doesn't mean "anything goes" -- to the contrary, novel mutations have to survive intense selection pressures in order to eventually become part of the standard language.
Where new ways of using existing terms create ambiguity and conflict with existing meanings, their survival chances aren't always great.
Although, there are multiple people in the talk section[0] arguing for the "'physical' machine" definition. Might have to get used to it, along with "crypto" and "algorithm".
The trendy uses are "algorithim" and "crypto" are specific cases of the general meanings of these terms -- there's no contradiction or ambiguity introduced here, so these uses are OK, although people presuming the narrower trendy meanings in broader contexts are wrong.
This use of "bare metal" does contradict the pre-existing meaning, so is not quite appropriate. What is valid is describing the OS itself as running on bare metal in contrast to running within a VM/container -- but an application running on top of that OS is not running on bare metal.
This isn't particularly egregious, though, since there are negligible cases of actually running applications on bare metal today: if you are talking about applications, the context can usually explain the intended meaning. But that wasn't always the case in the past (PC "booter" software used to be common), so this doesn't necessarily apply retrospectively, and may not be the case in the future, especially considering some of the interesting things companies like Oxide are working on.
I have no opinion on the sit-in, but it is always hilarious to see ex-Twitterites breathlessly bemoaning the sale, as if massive centralized ad-based networks feeding algorithmically presented "content" to the masses are somehow a boon to society as long as they are owned by a "good guy".
Any normal sidewalk would be wide enough that a single person could not conceivably block it, and wearing headphones while walking, especially noise canceling ones, is popular because US cities are largely unpleasant, deafeningly loud places full of fast-moving cars.
I used to agree with you, but a pure text web looks like Gemini, which I abandoned after a few days of getting lost in endless identical looking blogs.
There is no reason that websites shouldn't have room for some creative expression. For as long as writing has existed, images, fonts, spacing, embellishment, borders, and generally every imaginable axis has been used as additional expression, beyond the literal meaning of the text.
The body width is necessary because web browsers have long since abandoned any pretense of developing html for the average joe. It is normal to use web browsers maximized, so without limiting the body width the text is ridiculously long and uncomfortable to read.
As somewhat of a Zune fanatic, it always makes me happy to see a new Metro-inspired UI, but every one of these Zune-inspired projects falls short when compared to the actual Zune application, which imo is the absolutely pinnacle of music players. It presents your music library in a way that to me is aesthetically pleasing and entirely intuitive. The three column layout, with sorting options for each, is ideal. Filtering does not dump you into a new page. It's hard to describe what makes it so pleasant to use, but no application I've found yet comes close.
I encourage anyone with a local music collection to go download Zune and give it a try.
Dopamine is indeed a good software, but the version on non-Windows operating systems is based on Electron. I think that's really overkill for a music player.
IMHO Xbox 360, Zune, Windows 7-10 Mobile and Metro were very good UX implementations at their time, and bizarrely far superior to the touch mode Windows 11 offers today on tablets. WTF Microsoft, why are you regressing on all fronts?!
Zune never took off because it came too late and was going against a market already dominated by the iPod, and Metro was hated since Microsoft shoved it down the throats of desktop users with Windows 8, even though it was wonderful to use on tablets, except Windows tablets of the time sucked major ballsack since they were powered by anemic Intel Atom CPUs trying to run a full desktop OS compared to the ARM iPads running a mobile OS.
Microsoft’s decisions on windows will always be very funny. Push a gloriously designed tablet UI onto everybody, face backlash because nobody was buying windows tablets so it was tacky on desktop, come out with good windows tablets finally (or hybrid ones) but change the UI to a compromise that neither desktop or tablet users enjoy.
Zune never took off because Apple unveiled iPhone soon after.
When all the news and talks are about this iPod killer from Microsoft and then Apple themselves release a truly groundbreaking iPod killer themselves, you look foolish. Zune really got done in by marketing.
Definitely. But even if the iPhone were to be delayed there's no way Zune would have made a significant dent in iPod's market.
It was already the established player and the user base was already locked in with hundreds of $ in iTunes purchases. The usets weren't gonna throw that away no matter how much better Zune would have been.
iPod's market dominance wasn't in some UX magic that couldn't be replicated or braten by competitors, it was in the iTunes purchases that made it comfy to own specific songs and also locked users in.
I was sometimes involved in this project at MS. The music buying experience for Zune was awful, and I regret missed opportunities to be more forceful about telling the Windows Media execs that they were fucking up. Part of the problem was MS insistence on being fair to everyone and having an open system which led to 5000 shitty online stores where you could buy music. Instead of Apple that had exactly one where they owned and controlled the entire vertical.
This was the whole reason for creating Zune. Again, MS had licensed out their awful Windows Media products to 100 Chinese digital music player makers and got a fragmented market of trash, so MS jumped in to try and do it the right way themselves, but sadly too late.
I think the flat Metro UI evolved out of all the work we did coming out of the Windows Media division. I based all the flat design work I was doing for MS on British Sky TV's set top box graphics, simply because they were easiest for my developer brain to code and they also worked really well:
We chose a strict flat metro style on Microsoft Band because it was super easy to render with the tiny CPU we had! All squares and monochrome icons meant we could render at an absurd FPS.
What you’re describing is enlightening, thank you for sharing. The music purchasing workflow involved collaboration with the marketing people at Apple, which explains their successes.
What's funny is that Microsoft and its partners did all the groundwork to open up the labels and gave Apple an easy ride. I was working with Peter Gabriel and he could call any record label in the world and get us a meeting to show the demos. The startup I was working for would go in and it would spend a lot of energy proving to the labels that digital was the endgame and CDs were eventually going to die. This was a very hard job.
And then a couple of weeks after each demo we'd get a call from the labels and they would tip us off that Apple had come to show them something special that they were working on, and of course that meeting went a lot easier than ours...
(also every record label on Earth was using Macs not PCs, so Microsoft's software either had to be run on a laptop we took, or we had to show them Microsoft's absolutely awful Mac software. PG, bless him, gave me his personal iBook for testing (complete with all his passwords and AOL account), which he probably got from Jobs, but I was in a belligerent anti-Mac era so I shoved it in a closet with the button nailed down and stuck it on that "hold the button" game for two years)
No. I was involved a lot in these early rights meetings and the labels were adamant that the tracks had to have DRM. All the early demos that got the labels on board had DRM, but all the consumers hated it, everyone at the labels actually hated it too, but I suspect their legal teams were forcing the issue. It took a big personality like Jobs, with a market-leading product to finally kill it.
p.s. haven't Apple been sneaking DRM back onto their music lately? I know all the videos are DRM, right?
The world was also not ready for music subscriptions. That was the really compelling benefit of Zune for me, that we take for granted today with Spotify et al. The vast majority of people I knew, even early adopter types, just could not wrap their head around the idea. “What happens when you stop paying? You’re left with NOTHING!”
100%. These were the arguments we had with Microsoft, Nokia etc. We tested subscriptions a lot in 2000-2004, but you have to remember, this was DOWNLOAD subscriptions. So what would happen is, if you failed to pay you ended up with a player or hard drive filled with thousands of "MP3" files that you could no longer play because the DRM license had expired. The customers were livid as fuck about this scenario.
My argument was to have all-you-can-eat streaming subscriptions, but it was shot down, and at the time mobile data was pretty shitty (I built the original streaming service on a 9600bps GSM modem in 1999) so you could only stream at home.
Zune never took off just like HoloLens, Band, Kinect and Windows Phone didn't take off—not because they were in any way late, but because Microsoft is never "all-in" with any of these ventures.
The Xbox was somewhat of an anomaly owing to exclusives like Halo and Gears of War, but it's floundering, and in some countries, like Japan, it's just never taken off, period.
Apple Vision doesn't seem to be a sensation in terms of sales figures, despite the HoloLens beating Apple to the punch, and the stereotypical nonsense being "Apple is always late, but they always do it right". HoloLens just got killed along with Windows Mixed Reality.
Hell, even the Microsoft Band beat the Apple Watch to market.
The problem is that merely having a presence in a niche doesn't guarantee success as it once did. Now, you need to actually iterate, innovate, and satisfy the minimum expected threshold of solving real problems. This is also what Tim Cook's Apple struggles with these days.
What's the killer app for Apple Vision? What was it for HoloLens?
When digital technology was new and exciting, having anything would draw buyers. Now, we're spoiled for choice, things move fast, it isn't years between models, it's months to a year. Early adopters tend to get a bad experience too.
Wow, can I just say that my family and I loved our Bands? We had issues with the plastics breaking down on the first gen (had them replaced a few times due to them becoming unwearably uncomfortable due to it within the space of a couple months), but I'll be damned if it wasn't one of my favourite gadgets of all time in terms of tasteful design and innovative ideas.
Many a familial glass of wine was raised to whoever you might have been whilst discussing how much we loved them!
Band, Microsoft Band, oh how heartbreaking it is to read about the name. I once owned a Microsoft Band and was even excited about the second generation. Its design was captivating. But now it's all gone. My only remaining first-generation Band became unusable due to the notorious cracking issue. Now, I'm stuck with my Fitbit, which lacks design taste and inevitably has battery issues every three years. It's a product I have to patiently endure to keep using. Such a pity.
Apple controls the full driver stack while Microsoft does not. I've had to re-work HID interaction on Windows with WinForms and WPF to work around problematic touch interface drivers on Windows 7 and 10.
Microsoft is more towards licensing software to make money versus making a quality product top down.
Surface Pro with Windows 8 was pretty good, great performance on that device. I did a lot of sketching and some D&D world building with the pencil :). But the UX had a lot of frustrating spots where you still end up using Windows 95 UI because Microsoft didn’t care to update everything.
About a dozen years ago my employer wanted an intranet app for our mobile devices. Initially I was disappointed to have been assigned to create the Blackberry app because even though it was by far the most common phone in use by our employees, I could see the writing on the wall with Android and the iphone taking over. But I was also given free reign on the UI, which I borrowed very heavily from the Zune's UI. While it was well received, RIM really started falling apart around that time, which accelerated the replacing of the company issued Blackberries with BYOD. I was sad to see it go but also glad to be freed of the frustration of dealing with RIM and their often offline 'signing server'.
Yes, channels and groups are most likely what makes Telegram a threat where Signal isn't. That's an excellent argument for decentralized social media.
You're probably exasperated that others don't see what to you seems like an obvious truth. Rather than mocking the opposing argument, it's probably still worth rehashing yours when the topic comes up, even if it feels like banging the same drum with nobody listening.
This was entirely predictable and inevitable. I don't understand what Durov thought would happen nor why he rejects E2EE as a liberating technology.
Policy will never be the key to digital privacy, it must always be accompanied by cryptography. The status quo of allowing a third party read and store your messages forever, slurping up all the metadata along the way, is insane.
I think it is pretty obvious why Durov did not opt for universal E2EE. His main purpose of making Telegram was to make the chat app that is the most usable of all. E2EE comes with a cost on user experience which was for him too high.
Example: Signal can't handle more than one phone logged in, and if for some case you don't open the desktop app for more than 30 days, it logs you out there and you can never get these messages to the desktop.
Indeed this is. For some reason, all the implementations that I can recall suffer from some usability problems. I expect that if a solution that is acceptable for Durov is discovered, they will roll it out. Of course, my prediction might be wrong.
But it makes me question why "the browser" is apparently still the inevitable platform of the future.
In order for a PWA to be normal and usable, it must be available offline, open in a window without browser chrome, have similar performance to a native application, be launchable via a shortcut on the host OS, and respond to the mouse and keybaord shortcuts the way you'd expect. I think I've just described... an Electron app?
It's cool that this kind of thing can run in a web browser. With no install hurdle, it's much easier to convince people to try it out, and it's cross platform. Beyond that I can't really think of any advantages to having it run in the browser.
If what's lacking is an easy way to try software, I can't help but imagine lots of ways this could be addressed that would be much more pleasant to use than loading PWAs. Right now I can't seriously see myself enjoying using a PWA for work.
I say this having recently finished several large design projects in Figma, which is apparently a gold standard success story for browser based apps. Despite the years of development and herculean engineering efforts, I can still feel the browser jank. I begrudgingly open the thing in chrome, as it completely chokes in Firefox. It still chokes on moderately sized canvases, moving things is slow and laggy compared to native apps, keyboard shortcuts sometimes don't work or keys get stuck in a weird pressed or unpressed state, loading is slow, elements pop-in over tens of seconds.
I know I'm an old man yelling at clouds at this point, I'm just disappointed that we seem to be going backwards in performance and usability of software.