Hacker News new | past | comments | ask | show | jobs | submit login
Firefox OS 2.5 Developer Preview, an Experimental Android App (hacks.mozilla.org)
154 points by thallian on Nov 12, 2015 | hide | past | favorite | 44 comments



And yet my FirefoxOS phone is still stuck on 1.3, because Alcatel apparently could not be arsed to support their "flagship" FxOS phone for a mere six months.

Can we get an open-source mobile OS that is not fragmented to the point of uselessness?


Your very comment provides a hint that fragmentation depends on the state of OS ecosystem, not the OS per se.


His comment provides a hint that fragmentation happens because device ownership does not provide an adequate means to update or access the device's firmware.

For lots of devices out there (sans Google Nexus-devices and other honest exceptions), the device firmware is locked down and you need to find and exploit vulnerabilities in the firmware to gain access and fully assert device-owners. While hobbyist hackers often are able to produce updated firmware (look at OpenWRT and Android if you need proof), actually gaining access to upload that firmware typically requires more expert knowledge. And thus most devices remain locked down for their full lifetime.

As more devices become smart and have the capability to work against you given hostile firmware (TVs, fridges, etc), this becomes more and more important.

We should legally be allowed to actually own the devices we buy, but if the firmware is locked down and out of reach, we can never truly own it anyway.

The devices can be (some already are, sans spying TVs) programmed against us. How can we know? How can make sure they don't?

Not to go all Stallmansque, but I find this deeply worrying and very unethical.

We need to require open access to firmware everywhere, for everything. Asking for all original firmware to be open is clearly unrealistic, but at least give us access to provide our own should we want to.


All of this is true, but if people keep buying devices with their firmware locked and that aren't being kept up to date, then nothing will ever change, unless government regulations are introduced and that can create other problems.

I've had very positive experiences with my Nexus devices, though not all models are the same. Acquaintances have had good experiences with other, more open models, like OnePlus One. Yet I'm seeing even software developers buy phones that are tightly locked and that aren't being kept up to date properly, software developers who should have known better by the nature of their work. And then they bitch and moan about it.

This notion of buying things based on ethical and health-related considerations has become more prevalent in recent years in other markets (e.g. food, clothes), yet in the software industry after the great run we've had in early 2000 in regards to open standards and open-source, it seems that we are going back, willingly giving up one freedom after another.

But and I think this is important - I rather prefer the current ecosystem of fragmented Android phones, because the very essence of open-source is the right to fork. And if anything, Android is not fork-able enough due to Play and the Play Services. If you want a tightly controlled OS on a tightly controlled device, iPhone/iOS is that way. And I own both, I'm glad that both options exist and I hope Firefox OS will become a good alternative as well.


The problem is that being "open" is one deciding factor in 1000. I can't go through a checklist and say: "I want $screen_size, $battery_life, and make sure it's open as well!"

I have to first search which phones are open, then narrow down available size, battery, etc.


This might be new to you, but PC users that run linux have been dealing with this reality since 1996.

my first PC had a winmodem and an OPTi OPL-3SA sound card. traded the winmodem for a rockwell 33.6Kbps, and opted for a Yamaha sound card (because paying for OpenSound drivers was stupid, and the free trial gave you 30 minutes of working sound drivers).


It's not new to me, but yeah, your point stands. It's the same thing as linux was ~20 years ago. Hopefully the situation will improve like linux has


While I would also want everything free or open source, firmware is not necessarily like other free or open source projects that can just assume that they are running within a standard environment. It has a tendency to be much more specific to the hardware, so unless that hardware conforms to some standard to reduce the need to have to have many different firmware projects, there may not be the resources available to maintain all of those projects.

rms is not wrong, imo, but it would take much more effort than just putting the source out on the net somewhere. Since you have to get that source reviewed, compiled, and somehow put into the fridge or dishwasher now you need the fridge or dishwasher listening for firmware updates from the manufacturer and user and that has additional cost, probably beyond what would be needed for a repairman to update it, since it would have to be more straightforward to do for a regular user, or automatic. And some users may not want automatic updates, so you'd have to make that optional somehow. All of that adds to cost, so companies have little incentive to do all of this when weighing against the risk of the appliance's firmware getting misused and the cost of that.


Yes, devices are becoming more hostile to the use, I agree with most of your points. However, I don't think that there is a pure technical solution (open firmware updates etc.) as it depends on technical expertise of the users. Instead, we should support non profits and b-corps which can be trusted by users.


No.

Scare users about reflashing without having expertise in the field. Warn about not having any support, yet still provide some means to support them if they don't heed the warning and fail, so they don't brick the devices - keep a tiny emergency recovery-only bootloader that would allow to reflash system in any state.

But please don't insult their intelligence and don't deny their rights to control their own devices. Sure thing, there are plenty of stupid persons who'd wreck everything and then some. Heck, I hope I'm not too stupid myself, but when I first time flashed a router I screwed things up because of a typo. Let users end with the brick (they were warned, right?), recover it - or pay someone to do so - and learn.

The only real issue that's in conflict with this is theft prevention. A legit use should be able to run any software on their device, but some don't want to give that opportunity to anyone but its legitimate owner. The technological part for this is here, but that's not a technical issue, but a purely legal one.

But please, don't just shift the trust from vendor to some non-profit. Let the device owners manage their trust by themselves and have the ultimate decision on such matters. And for all the sanity still left in the world, let end-users still be the owners.


Your attack is naively misdirected, I fully support the right of device owners to have a full control of their gadgets including firmware, hardware, software, you name it. But as most users (and we already know this) don't care about these freedoms, for-profit companies don't have an incentive to guarantee them. Especially if it weakens their competitive advantage. Hence my call to support non-profits, like Mozilla.


Sorry if that sounded like an attack. No offense was meant.

I guess I got your point wrong. Now I see. Yes, that makes sense. Even though I don't like, say, Mozilla (I have my reasons), with a popular after-market firmware they may fight for a way for everyone to flash anything - not just their firmware. Just hope this won't end up like it happened with UEFI Secure Boot, where everyone's who's unlucky to have firmware that doesn't support custom keys are stuck with Ubuntu-signed GRUB.


No offense taken :)

You may not like Mozilla and that's ok. But you still should support them as Mozilla's interests are mostly aligned with the interests of the users.


About Mozilla - I have an impression that they put their own interests (product popularity) somewhat above user interests.

Yet, as long as everything doesn't end up with some Mozilla-specific partnerships or highly proprietary (in a sense of uniqueness, not source code availability) solutions that are hard to use for anything but FirefoxOS, I think you're right.

Add/Edit: although - in reference to your original comment up the thread - I'd wont say "instead". I think it's what felt wrong to me and got me confused with what you had suggested.


In my opinion, popularity of Mozilla products in itself is in the best interests of most users, directly and through positive externalities. Minor transgressions like Pocket integration are a petty reason not to support Mozilla and their mission as a whole.

Edit: sure, my wording could be more clear.


Did you ever like mozilla, or were these reasons more recent in nature?

I personally liked Mozilla up to the SJW/Eich situation. Since then, every move they've made has been counter to the tenets of free software that JWZ and team instilled. Now, I merely use it because everything else is less-free.

I'm hoping Servo will end up being to Firefox what Phoenix/Firebird/Firefox was to Netscape.


<rant>

They used to be good when they started. I only care about what they do with and to the technology. Whoever is on their team doesn't concern me.

My current issues with Mozilla/Firefox:

- Caving in with adding DRM support, which contradicted their principles because, duh, otherwise users would hop to Chrome to watch Netflix.

- Dropping relatively sane sync-1.1 in favor of sync-1.5. It's as proprietary Firefox-specific solution as it only could be, with internals being a terrible mess I can't call anything but over-engineered. And after more than an year it's still insecure. And they purportedly removed UI for self-hosted sync.

- Promoting Persona/BrowserID, contributing to the very wrong concept that someone's very identity is their account with a third party (email or domain address, as leased by email provider or domain registry). Good thing they put this on hold.

- Simplifying UI where it's unwanted. Say, just recently the "green lock" UI was crippled to the extent it doesn't provide any useful information anymore. To heck with "secure connection", I can see this in the address bar already.

- Bundling more and more stuff into browser, like it's a new Netscape Navigator.

- Gradually showing more and more "hints" and "useful tips" all around, without means to opt out once and for good. Like that annoying popup at the bottom of the "top sites" page on Android (the only UI to disable is about:config).

- Starting to add non-opt-in analytics/tracking anti-features (Adjust on Android release builds).

I feel that those things are not in favor of end-user personal freedoms, but were introduced just because marketing thought they'd look appealing to the audience - who doesn't care about the mission and other stuff that Mozilla likes to mention - or necessary to keep market share. It's just my personal perception, though. I'm by no means objective here.

</rant>


While I agree with most of your criticism, I actually find them visualizing new features a good thing.

Right now the Firefox dev team is preparing to axe Panorama/Tab-groups[1], one of those superior Firefox-only features, because supposedly only 1% of the users are using it.

That sounds reasonable, until you ask "Has the feature been marketed at all?". When the answer is no, even 1% seems like high adaptation and axing suddenly sounds less like the right answer.

Why didn't they market this feature more? It's absolute genius, and allows an efficient keyboard-only driven browsing experience with almost countless tabs, and it's correspondingly bonkers to be removing it.

[1] https://support.mozilla.org/en-US/kb/tab-groups-removal


> Since then, every move they've made has been counter to the tenets of free software

I object and would like to provide this link to back it up: https://github.com/mozilla


Note: on HN, taking Brenden's side in the Eich/SJW catastrophizing will cost you karma.



I have the Onetouch Fire E, which is a completely different device.

In fact, Alcatel released four FxOS phones, of which even the community only supports one.

You see my problem?


You're lucky! Mine is 1.1.


Comments like this are why I think Google should merge Android into Chrome OS and not the other way around. It seems that most manufactures have concluded that offering long upgrade periods for their devices is against there best interest. As a result I only recommend iOS and Nexus smartphones.


On My nexus 7 it's really slow. Staring any application takes ages compared to times that ParanoidAndroid does it. Start-up time with Foxfood is noticeably longer than PA, actually I got bored waiting. 8 minutes and counting... It's not possible to stop it and go back to android launcher.

There is an updater built-in. UI looks better than my Android, there is no theme changer, so it's a no-go for me.

It provides access to applications from Android, but icons are HUGE. On N7 I have 3 icons per row (you can change it to 4, but it's still ugly), so they're really ugly and pixelated. Settings are pretty nice, it looks like something between iOS and standalone Android 5.

It didn't fully replace my android UI, as the bottom bar with back, menu, recent is still visible, but pressing back doesn't do anything, instead back button is placed in top left. Not every application has the button, contact settings doesn't have, so I had to restart FirefoxOS to go back to main menu. Also I still see Android heads-up notifications from VPN reconnecting. Doubled.

There are a few unnecessary UI-fireworks, which I would like to disable, but can't see any settings about UI.

I think battery consumption is big as well, in several minutes of testing I lost 11%, which normally I use 6%-10% in 1.5h.

It's good, needs a lot of polishing, and it's getting somewhere.

I listed only disadvantages here, but it looks nice, their own UI is clear and consistent, better than android, everything what comes from FxOS runs very smoothly. It's 7/10 for a developer preview package.

When those things improve in the future, I will be happy to replace Launcher3 with Foxfood.


I'm really hoping that Firefox OS will save us from a future overtaken by walled gardens, save the mobile web that apple and google have less and less incentives to support. But unfortunately so far adoption has been nominal from mayor manufacturers, I understand this "trojan horse" approach even if it seems to defeat the purpose. The world needs a mainstream open mobile device, fingers crossed.


Could it be pushed to F-Droid? At of people using FD as main source of apps would like to try it.


I don't think so. Similar to many Linux distributions F-Droid apps should be build from source. There are some issues about the branding (logo, name etc.) which leads to things like Iceweasel (Debian) or Fennec (F-Droid). See: https://f-droid.org/repository/browse/?fdid=org.mozilla.fire...


Beside this, Firefox is not really welcomed (and a drop is planned eventually) on F-Droid because it's really complex to build, depends on some closed source libraries (Google Play Services, for example) and has lots of tracking in it (everything is at least opt-out as far as I can tell, but the guys behind F-Droid aren't really keen on any kind of tracking it seems).

It's possible to build it without those problems (see Fennec on F-Droid in fact), but it takes a lot of time for someone to create a recipe, and it's prone to break on every upgrade (and that's why the last version of Fennec on F-Droid is based on Firefox 40).

I don't know if/how many of those problems apply on Firefox OS (I guess at least the complex build environment applies), but that's likely why you'll not find it on F-Droid anytime soon. Unless someone decides to pour lots and lots of energy into this.

More info/sources:

https://f-droid.org/wiki/page/org.mozilla.firefox

https://f-droid.org/wiki/page/org.mozilla.fennec_fdroid

https://gitlab.com/fdroid/fdroiddata/blob/master/metadata/or...


Yea, I know all of that.

>It's possible to build it without those problems (see Fennec on F-Droid in fact)

Fennec is stuck on 40 because of difficulties with building it, and even when one person managed to build it, .apk didn't start on others developers phone. A few people tried upgrading it to 41 and all failed, I guess fennec is kill now. I talked about it a few times on their IRC with people responsible for fennec.


And if you wonder why Ubuntu has a real Firefox branding, they have a deal with Mozilla.


Boot To Gecko (B2G) is the underlying open source project. "Firefox OS" is the official Mozilla-branded product.

It should not be a problem basing it on B2G.


They could set up their own fdroid repo anyone could add to fdroid. Decentralization over an open protocol is a good thing. Guardian Project hosts their own and it's bundled, but there are a few private repos in the wild you can add today. It's similar to Ubuntu's third party sources or ppas. If big open companies started supporting it, maybe it would gain some steam. This would sidestep some of fdroid's official repo's more stringent requirements.

https://f-droid.org/wiki/page/Setup_an_FDroid_App_Repo


Is that necessary, though? The application is already available in this very article, without involving the Play Store if that's what you're trying it avoid.


Does it autoupdate itself?


Yes it does, from the nightly channel.


Honestly: I don't know, haven't tried it so far. But Mozilla's Nightly (Aurora?) does on Android, so it should be possible.


For the love of all things good, why didn't Mozilla adopt the webOS UI instead of the terrible, unscalable iOS UX? Sure, webOS couldn't survive the awful hardware designs and last-generation JITs, but the UI was the best mobile UI/UI to date.

Even at this point, I believe there's a lot of potential that could be gained if Mozilla worked with LG. Mozilla has the web experience and LG has the patents and hardware ability.

I'd go for a $350-$400 LG/Mozilla phone with nexus 5 specs running a new version of webOS in a heartbeat.


It's extremely hard to find a Firefox OS device with acceptable (as in "> 512 MB RAM") specs in Europe.

I looked for months and I ended up buying a Moto G 2014 and installed CyanogenMod (which took me ~3hs because newb).

I couldn't be happier today and I sincerely hope FirefoxOS offers its OS on newer hardware if it ever wants to stand a chance.



Glad to see Mozilla announcing FxOS release versions, at last (previously they didn't).

Should I be able to upgrade any FxOS device to 2.5 now?


I am REALLY tempted to buy a Nexus 5 just to put FirefoxOS on it. Kinda looking for a more reasonable phone though, also not sure how well it's supported.


I've been running Firefox OS on Android for a couple of months now on my Nexus 5 with CyanogenMod and everything works alright. As long as you have KitKat or above things should go smoothly.


https://firefoxos.mozilla.community/device/LG/Nexus%205 it is "Mozilla maintained" so as good as it gets =)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: