In one of the recent threads about Microsoft switching their Edge browser to be based on Chromium, I thought I read that one of the arguments against Firefox was the lack of ARM support.
I will continue to hold that the Edge team/leadership at MSFT made a wrong choice betting on a web monoculture here, and I hope Firefox remains popular as an alternative.
It would have been jaw-droppingly shocking if Microsoft ever chose Firefox or a similar "free as in freedom" project as the basis of their own products because they are not goal-aligned. Whereas Webkit-based (which did fork from KHTML...remember Konquerer?) browsers have long since been serving as the platform for big-software webtech, Firefox and Mozilla largely remain oriented towards the interests of users. This means decisions that could help commercial entities take a far second to the needs of direct users with few exceptions (I'm looking at you EME and DRM).
Firefox has had ARM support for a long time. Heck, there's even MIPS support in the tree, though I don't know whether anyone's exercised that code recently.
I want a phone-sized Windows 10 ARM device that runs desktop-class software such as this version of Firefox when docked (and adapts to mobile capabilities when undocked, as a responsive web-app does). Although not Windows, Purism seems the most promising runner in anything like this race today, and I'm hoping more contestants enter the race.
In any event, thank you Mozilla for supporting this platform. I hope Microsoft puts some more serious weight behind it.
Unfortunately, Microsoft will not likely reenter this arena or support phones with Windows in any official capacity, the current direction from Satya appears to be to retreat from every competitive market and cede all standards to El Goog. Microsoft promotes Android, is building a browser based on Chromium, and is abandoning Cortana. Best hope for this would be a third party building an ultraportable themselves and writing their own phone support.
Oh, sure, Satya's strategy has done wonders for Microsoft's stock value, but Microsoft's constant retreats from various parts of their ecosystem sequentially harms the remaining parts of their ecosystem. The further they retreat, the less people have a reason to build on Microsoft's platforms.
And bear in mind, HP is a hardware company, they aren't worried about what platform is running on their hardware. Microsoft is a software company, which is a very different ball game.
The problem is they're playing in someone else's sandbox that way. Google is a competitor on SaaS and cloud infrastructure, and have the ability to directly integrate their services with Chrome and Android, whereas Microsoft does not.
Of course this is a hack, not a usable product, but I wouldn't be surprised if docking would be possible eventually, maybe someone even could hack the Windows 10 Mobile UI to run when undocked… (or design a new mobile UI)
Mobile was really based on an entirely different Windows system architecture than the "legacy" desktop people all know and love. In addition to that, making handset UX from scratch has shown to cause a lot of fun bad experiences as seen with the (much more mainstream-appealing) open Linux handset attempts.
It'd be a great use case to have such a thing for NT-on-ARM, but I don't see this as much more than quick hacks to waste some time, especially with the 855 and 8cx SoCs diverging again so that the brief peak of binary driver sharing being possible is pretty much gone again...
> In addition to that, making handset UX from scratch has shown to cause a lot of fun bad experiences as seen with the (much more mainstream-appealing) open Linux handset attempts.
At this point I just want X11 back. At least that worked.
X11 and Wayland works just fine on Allwinner A series & H series chips as the community has added nearly complete blobless support for everything from the HDMI phy to the Gigabit ethernet controller to the oddball WiFi chips some boards use.
Qualcomm and Mediatek are the ones with shitty out of tree Board Support Packages, preventing Android updates and blocking accelerated X11/Wayland from running.
I don't think you will see that soon or anytime. Microsoft doesn't have a clue about what a tablet experience should be. Sure, they have OneNote and some specific applications tablet aware but it is not well supported by "alien" and popular applications. For example, the copy and paste in Google Docs under Chrome doesn't work!
The bad tablet experience on Windows is entirely the fault of app developers, Chrome included. Chrome's HiDPI and touch support on non-mobile are an atrocity and it's entirely their fault. They've gotten a little better lately, at least...
In comparison if I use (classic) edge to browse and read on my tablet, it's totally fine. Firefox is relatively okay as well (similar issues though, because it just doesn't get much testing).
Having two monitors of different DPIs makes it especially obvious how bad most apps are at DPI. The good ones stand out to a dramatic extent. DPI is a difficult thing to implement, but Win9x has had tooling and support for it since Windows 10 and they improve it with every release. When I implement support in my apps it really doesn't take that much effort, just forethought.
For one example: Chrome's scrollbars and other widgets are utterly atrocious on touch (even on Android it's kind of bad). The standard Win32 scrollbars and other widgets can have all their sizes manually adjusted in the registry, and by default many of them will automatically adjust when you're in touch mode. I was easily able to make scrollbars bigger across literally everything except Chrome.
> The bad tablet experience on Windows is entirely the fault of app developers
I don't think so. Microsoft should make easier to develop for tablet mode if they want adoption, the DPI issues happen also in Microsoft applications! The Win32 APIs are a legacy mess and Microsoft has a long tradition in delivering lot of frameworks which ends in failure (WPF, Silverlight, etc).
BTW, Google Docs copy & paste in tablet mode doesn't work in Firefox either.
Everyone seems to be tripping over themselves in excitement about these qualcomm SOCs. Qualcomm has a history of being pretty hostile toward everyone, because of that these have a pretty low chance of being used in any kind of device that will be able to run current software for more than a few years.
Meanwhile we have devices based on freescale IMX chips that run GNU/Linux with X11 and whatever else you want (including firefox/iceweasel) right now, they probably just don't have the on board peripherals and connectivity you want.
Well… it's a complex situation, they officially contribute to Freedreno IIRC, but the firmware is still terrible. For some reason you're always booted under a hypervisor, even. There was kind of a hope that Windows requiring ACPI would improve the situation, but they seem to still use a lot of custom crap.
Most Qualcomm chips expect DDR3L/DDR4L, so desktop and most 2+ year old laptop ram won't work.
Additionally, the drivers for Snapdragon SOCs are a mess, and they haven't made an effort to get all their changes in the mainline kernel. For such a large company, Qualcomm has some pretty wretched development and business practices.
I'm looking forward to trying this out. I have a first get Windows on ARM device and the standby battery life is amazing. If the browsing was a bit more performant it would be a much better device. Youtube is quite sluggish with Edge and YouTube TV doesn't even support Edge (at the moment) so I've been using an x86 build of Firefox.
I will continue to hold that the Edge team/leadership at MSFT made a wrong choice betting on a web monoculture here, and I hope Firefox remains popular as an alternative.