The Sailfish guys are actually the creators of libhybris.
If you run SailfishOS you have to first have Android flashed onto the phone. They use the same kernel, camera drivers, GPU drivers, etc as the original OEM including the prorprietary wireless BLOBs and the Android Radio-Interface-Layer ("RIL").
I've spoken to the Sailfish guys awhile back and I get why they did this -- 10+ years ago there was basically no choice but to use the Android port of drivers + the Linux kernel the vendor shipped because there was no other way to make these hardware pieces work, thanks to the silicon vendors.
The story of not needing BLOBs and things like a libhybris-shim has slowly improved, but not 100% . We can run Debian linux on the Qualcomm Snapdragon laptop devices (Thinkpad X13s, etc) but bits and pieces are still not there (audio, full power management, Bluetooth, etc).
To current Qualcomm's credit there are people inside who are pushing for everything mainline Linux, and minimizing proprietary pieces.
Libhybris is great for making a tech demo. I wouldn't base a product on such hack, anymore than sell Linux laptops with ndiswrapper...
If something doesn't work in the binary android drivers, the vendor won't help you (we support only android, sir). Nor can you fix the drivers yourself, because you don't have the sources or the knowledge how drivers work.
All the phones sold with Sailfish OS & all the officially supported Xperias used libhybris - same with most unofficial ports. While not ideal, it works reasonably well - saying that as someone who has been using a Sailfish OS device as primary phone since 2013.
"and yet, it works". To this day my Jolla is the best phone I had, always fluid and reactive compared to much more recent and powerful Android phones. I don't remember any driver issue with it.
Do you use Plasma on a phone or tablet? If phone, can you say which hardware platform and how "well" it works as an actual phone, making calls, texting, etc?
Are you saying that Plasma Mobile runs on any device that is supported by libhybris?
I was asking about Plasma Mobile ( https://plasma-mobile.org/get/ ) that the parent mentioned switching to and I wasn't aware that libhybris is a requirement.
If you want a plasma mobile device, your best option is to find devices that are supported by PostmarketOS. That is a distro (kind of like Ubuntu etc ...), which under the hood can use libhybris to talk to Android drivers etc..
I've used it for several years and the feedback from a user point of view is not positive. My sample size includes me and several members of my family who used Sony Xperia devices running SailfishOS for several years.
The Sailfish guys for some odd reason decide to invent their own "user interactions" where you click-slide ("one handed") to do certain opertaions. This makes the UI not only awkward, but NOT intuitive. You don't know what your options are until you perform this strange operation. I get why they did this, it was a way to potentially reduce swiping, etc but now that we have phones with big screens, you can actually put those options in one UI.
Further, basic things like composing a text and attaching a photo requires a round-trip to the photo app where you 'tag' the images you want ONE BY ONE rather than being able to do this inline from the SMS/MMS application. I think this has gotten better recently but for a long time it was SUPER awkward.
Two other perplexing points was how SLOW the UI felt for what should have been compiled Qt code and poor battery life on the older Xperia devices. Maybe they're using QML and it's not compiled?
The Sailfish guys have what I think is an ugly looking UI as well.
They've "dithered" certain parts of the UI so it really looks like old-school EGA/CGA graphics, even though the display is high-DPI and they have what's effectively a TUI style interface.
The only people I know who "LOVE" or claim "it's the best" UI are the same ones who LOVE Zune and Windows Phone UIs which are basically flat UI, almost monocolor nearly TUI type which is what you see pieces of in Win10 as well. Personally I dislike this UI and so do many people I know, there's a reason why UIs have icons and ideally text labels. TUIs have their place but so do GUIs.
If the Sailfish guys abandoned their weird UI ideas and frankly made it more like iOS or Android (I know, so boring, we have to re-invent the wheel just because...) it would actually be compelling.
On the very very plus side of Sailfish, as someone else pointed out, it's basically a GNU/Linux device that uses RPMs. I was able to install dnsmasq, set up DNS based adblock filtering, curate firewall rules and basically harden the device. You could SSH into the device via USB without adb stupidity and once I set it up, it stayed working until the VOLTE switch-over occured.
I think Ubuntu Touch has a better "UI" (I've also run this) but the Ubuntu guys have basically been ignoring VOLTE and since all major US carriers have switched over to VOLTE, your phone basically can't really make calls now on Ubuntu Touch (but that's OK, they've improved a bunch of other stuff! /sarcasm off).
Ubuntu Touch (not that you asked) is also a LOT slower than it should be and because the Ubuntu Touch guys are pursuing an 'Over the Air' update model, since the OS can basically be overwritten, applications aren't actually unpacked at install time but dynamically at run time. On a desktop this is OK but on a phone it leads to very slow app loading times.
I have high hopes for the current batch of Linux phone projects, Mobian, postmarketOS, etc but sadly I'm on Android until these are fully solidified.
I actually have the opposite opinion. I like their UI, with the drag-down feature to select options, etc etc. Though I liked the very early Sailfish version on my old Jolla phone even better (it was upgraded on the Jolla phone as well, but yeah, the old one felt better. But not everybody agreed with me on that).
The issues I have with my current Sony Xperia w/Sailfish is a) Nearly impossible to get a Japanese SIM card to function (as a second SIM), and when it worked it did so only sporadically. Basically useless. Same SIM worked fine in an Android phone and an iPhone (both brought over from Europe). And the camera.. it works, but it can take many seconds before the photo is actually taken, and all the features of the camera(s) aren't available. Not that I use it much for photos (though the ones I do take look good), but..
(And of course I also hate the long narrow super-slick easy-to-drop Xperia phones, but they're just like nearly every other phones these days, and the only option for an Android-enabled Sailfish phone now)
Funny, I used the Nokia N9 back in the days and the UI of (what was called Meego back the IIRC) was head and shoulders above everyone else. I believe the they were the first to have general gesture navigation so your comment about reinventing the UI is somewhat off the mark. Android implemented things after them, it's sort of like the argument that unix terminals should adopted ctrl-C for copy because it's the "standard".
I actually bought a Sony Xperia 10 and sail fish because I wanted the UI back so bad, but unfortunately I have some apps which didn't seem to work with android emulation (mainly banking...)
I am not saying the gestures in Android and iOS (app switching, etc) are actually the value add, but in fact things like toggles for options, or a "=" where the options are available to turn on/off. Sailfish forces gestures for things inside an application as well.
No doubt Meego innovated on ideas, but just because they came up with something doesn't make it "good" and just because Apple/Google copied it doesn't prove the validity of the idea.
To that point I would prefer we used more screen real estate (Android, iOS, whatever) and REDUCED the usage of gestures, it would end up being faster. It sometimes takes me multiple attempts to swipe from the bottom on a Android/iOS to get it to do something because I have a screen protector and/or case and the way I'm interacting the with the device is different than the developers who might have worked with a "nude" device.
The screen protector/case issue made UI navigation even worse on Sailfish devices because you had to use this gesture inside a program, not just to switch between applications.
Ubuntu Touch also has a swipe, but from the side where a screen protector is slightly less likely to affect it's ability to register the gesture.
Tbh N9 is still way ahead than today's Android experience imo.
It's also more consistent gestures experience than sailfish. Here you know the gestures are basically for "window/app management". Everything else - they look like regular Android apps.
Wow, this comment sounds like it's written by someone who's never interacted with "musicians" or thinks a musician is a person who goes on stage and plays the same song over and over. And further thinks that engineering or being an "athelete" or "musician" is a very narrowly scoped job.
Musicians, like any other "professional" have a broad range of functions from arrangers, composers to song writers, performing musicians, session musicians, touring musicians and so on.
A jazz musician (who might be performing live on stage) will likely not play the same song the same way twice. Is a jazz musician then not a musician because they aren't "repeating a set of movements over and over?"
If anything writing a CRUD type application IS something that could be automated because the patterns and the goals are largely the same but to take this example and apply to what atheletes do is pretty misguided.
Most atheletes are dynamically reacting to their environment or situation, taking into account the newest data and formulating a plan "on the fly" to meet their goals (scoring a goal, landing punches, etc).
My own credentials including multiple engineering degrees, experience designing equipment for "musicians" and "atheletes" so I don't think I'm talking out my ass here.
Anyone who wants a genuinely detailed treatment of this subject should read Allan McDonald's book "True, Lies O-rings" *. I happened to have finished this a few weeks ago and it goes on my list of all "engineers should read this".
This was really one of the most fascinating books I've read and likely the most definitive treatment of the subject by a subject matter expert. I kind of skimmed the blog article, the book explains in critical detail the issues with the original design and why the re-design (done after the disaster) was a much more robust approach.
In a nutshell the Shuttle SRB field-joint design was taken from a Titan missle design that was deemed to be "solid engineering" because none had blown up, but Allan mentions the SRB field-joint was flawed from the start and the joints suffered rotation and physically moved / flexed. (Later, it turns out a Titan missle exploded and the teardown showed the o-rings a primary point of failure).
Allan mentions it was the blowby past the o-rings that was consistently the issue and the engineers wanted to understand and address this problem for a long time.
What was striking to me, beyond the technical aspects of making these things work is the actual cover-up and attempt on NASA+Thyokol to blame McDonald and others for the resulting disaster. I knew of some parts of this, but you don't realize how messed up the situation was/is until you read the book.
Personally I'd ignore any negative reviews of the book, I think non-engineers, especially those who haven't worked in an Aerospace/Defense environment or in a big company might think Allan is arrogant or boasting, but he starts by providing the foundation for his statements before getting into the details which is a classic "engineer's engineer" way of thinking.
WordPerfect really was an outstanding word processor. Reveal codes (like many others here have pointed out) made "debuging" formatting issues relatively painless as was the "make it fit to a certain layout or size" feature. In an era when you didn't really have WYSIWYG they did an excellent job of enabling users to more or less get nice looking output without having to go to TeX.
I remember it took a LONG time before there was a Windows version of WordPerfect which I think took a lot of their momentum away. Combine that with Microsoft basically giving away Office or bundling Word+Excel they succeeded in eroding market share from Lotus / WordPerfect.
I think the Lotus Suite may have even pre-dated MSFT Office as a suite (not 100% certain) and as usual functionality was often superior or better implemented than MSFT's.
Credit should also go to WordPerfect for making a Linux version in the 2000's before Linux desktop was as mature as it is today. Sadly they didn't continue this effort.
I'm glad we have LibreOffice but it's frankly a clone of MSFT Office, the UI is very cluttered and it has the same "weirdisms" that Office has.
The thing that got me was I used to use the second to last version of WordStar. Which had paragraph and page styles that you could import, edit and apply to text. When it became apparent that I couldn't be using Word Star to share documents anymore I tried WordPerfect and it was so annoying. You just want to set the paragraph style and start typing not play with tags. Eventually I just used Word. But styles in word wasn't nearly as obvious and straight foreword.
First, thanks for this project and making your self accessible!
Will "plug-in" or "add-on" support be a first-party concept in Ladybird?
I ask that because in years past a few other browsers (Konqueror, Falkon, Dillo, etc) made it pretty far but lacking add-ons, useful capability such as 'NoScript' or 'uBlock' or even a tab manager made them non-starters.
NoScript for Dillo makes no sense as it doesn't support JS anyway.
uBlock... yeah, a little, but most annoyances will be blocked by the lack of JS support anyway.
The plugins for Dillo are only protocol plugins; there are no file format plugins and no other kinds of plugins. However, I mentioned they should implement file format plugins too; other people also wanted this, and it does seem to be wanted enough that they might do it. (Other plugins will be more complicated to consider how to support it)
I tried the bookmark bar and could find no difference. The bookmark manager for FF seems far more advanced, with Chrome going the simplicity, beauty, and lack of information route.
I would hope that plug-ins and add-ons can be written in C (although any extensions written in C should be only allowed if installed manually by the end user (e.g. by adding it to some configuration file); it should never install them automatically from a "app store" or similar). That is a feature I would use.
If by "C" you're asking for C compiled to WASM, then fine. But otherwise I would hope that WASN'T ever possible.
The endless security nightmare that was ActiveX and NPAPI should serve as more than enough reason why that shouldn't be a thing again.
"Installed manually not from app store" is even worse because then you're encouraging people to download random binaries from random websites and that's even worse
I do not mean C compiled to WASM. I mean native code (with dlopen).
My point is not for other people to make extensions that you must use. Rather, my point is in case the user wants to write their own extensions do things that have more permissions, without needing to recompile everything. It is specifically if the user does not want the extra security (because they intend to program it to do things beyond that provided by the browser's security context), and only for that case.
(However, there might be another alternative: Provide a .a file (in case do not intend to compile it by yourself, which might take some time and require several dependencies) and allow the end user to link that file together with their own .o files, instead of using .so files. The constructor functions can be used to tell the main program of the presence of these extensions.)
(Another alternative would be to provide a separate version that may permit this, e.g. "advanced version", that might also offer additional options and other features which are intended to only be used by advanced users, therefore making the user interface more confusing for users who do not read the documentation.)
C (and other programming languages) that is compiled to WASM could be installable from the app store, since then it is safe. Native code extensions must be installed manually.
I'm using a Dell S2716DG which is a 27" 1440p TN panel at 120 Hz. As far as I'm aware, there were only ever a few 1440p models that got 3D Vision (officially). I've tried running 3D Vision on an OLED ultrawide I have, and it works, but only on the bottom of the screen. I assume something to do with the refresh rate (144 Hz) or pixel response time (which I think isn't great with OLED).
I'm driving it with an A4500, on Linux (openSUSE mainly), 3-pin to the USB emitter, with glasses. The A4500 is somewhat gimped because only the 470.xx driver works. With the newer drivers X11 detects the display and emitter but displays both frames simultaneously. I think it might have something to do with the stereo declaration in the xorg.conf file being different with the newer drivers, but I'm still chasing down that lead.
A version control tool shouldn't take YEARS of usage to be productive, to me that's a red flag. Some engineering tools are just complex because what we're doing might need that flexability so the tools provide a myraid of options or workflows.
But writing software, tracking your changes, being able to back out your changes and getting your changes integrated into a code base should be as low activation energy as possible to be productive.
But again, git wasn't created for "you and me", it was created by one guy for his project, the fact "mere mortals" are using it is kind of our fault.
If you run SailfishOS you have to first have Android flashed onto the phone. They use the same kernel, camera drivers, GPU drivers, etc as the original OEM including the prorprietary wireless BLOBs and the Android Radio-Interface-Layer ("RIL").
I've spoken to the Sailfish guys awhile back and I get why they did this -- 10+ years ago there was basically no choice but to use the Android port of drivers + the Linux kernel the vendor shipped because there was no other way to make these hardware pieces work, thanks to the silicon vendors.
The story of not needing BLOBs and things like a libhybris-shim has slowly improved, but not 100% . We can run Debian linux on the Qualcomm Snapdragon laptop devices (Thinkpad X13s, etc) but bits and pieces are still not there (audio, full power management, Bluetooth, etc).
To current Qualcomm's credit there are people inside who are pushing for everything mainline Linux, and minimizing proprietary pieces.
Ubuntu Touch relies on libhybris as well.