I agree with everything else you said, but if it makes you feel better, Pine isn’t making any money off of their phones, they only put the hardware together. The software is community-driven, so buying a pinephone isn’t compensating Pine for Purism’s efforts. It’s adding one more person to the group of people who are working to make mobile Linux better, either via developement, testing and bug reporting, or showing it off to everybody else.
4 Likes
There’s a link to a poll in:
8.4. How does mobile Linux have any chance against Android and iOS when other OSes failed?
I feel that supporting Purism is the fastest way to get mobile Linux to become a viable alternative to Android and iOS, because I think that Purism has the most feasible path to making it happen. 15 years ago, I probably would have bought the PinePhone, and supported its approach, but I no longer believe that approach will work if the goal is to reach non-technical users.
However, I actually believe that the PinePhone helps Phosh advance as a mobile DE, because it is bringing new users and developers to the platform. The Mobian and postmarketOS devs are helping to improve Phosh and Phosh has been packaged in so many different distros because of the PinePhone. Phosh needs both paid developers and lots of volunteers to succeed, and the PinePhone is helping to attract the latter. It is going to take a lot of volunteer labor to add libhandy and Kirigami classes to existing GTK and QT software, and the PinePhone is providing an army of volunteers to take on the task.
In other words, Phosh needs both types of phones to succeed–the Librem 5 to pay for development and PinePhone to pull more volunteers to the mobile interface.
As I explained in my essay, I think Phosh is the best approach to making a mobile Linux desktop environment and has the best chance of succeeding, so financing its development is very important. However, the existence of Plasma Mobile and Lomiri also helps Phosh to succeed in the long run, because most mobile apps developed for those platforms will also able to run on Phosh. Because Plasma Mobile is now adding support for GTK, mobile apps developed for Phosh will also be able to run in Plasma Mobile. Because the developers of Linux desktop applications can see mobile growing on all three platforms at the same time (Phosh, Plasma Mobile and Lomiri), they are more likely to take it seriously and work on making their own code adaptive so it runs on mobile. If all three of those interfaces can be made to run on the Librem 5, it means more sales of the phone and ultimately helps pay for more development of Phosh and its apps.
The overall effect of having multiple Linux phones (Librem 5, PinePhone, Volla Phone, Cosmo Communicator, Astro Slide, F(x)Tec Pro1, Xperia 10) on the market and multiple mobile DE’s is to create a wave that generates more buzz and attracts more users and developers to mobile Linux, so it ultimately helps us to have multiple phone models and multiple interfaces. Most of the app development targeting one device or one interface will end up running in Phosh on the Librem 5, so there is no reason for us to fear the diversity of Linux devices and interfaces.
7 Likes
Actually, in the link regarding the pinephone’s price increase (which I have since lost), they explain that the reason for the price increase is because non-technical-minded people keep buying pinephones and wanting help with the software, so Pine’s solution is to sell a “retail” version that costs more in order to support software support. Evidently the buzz and interest has already begun to expand.
3 Likes
Feel free to correct me if I am wrong. But my understanding of the open source software development culture is one of forced self sufficiency and requiring a high commitment to learning in the early stages as the early part of the learning curve is steep. To ask for help or technical support from others without giving something back (probably money if you’re a complete noobe) is like someone showing up at the Winter Olympics asking the professionals there if they can teach you to ski as their athletes gear up to compete. I would really like to take part in the software development of the L5 myself, but would probably have difficulties getting a “Hello World” program running with some tools/libraries and would need to learn a lot of protocols about contributing to the repositories first (with no one to teach me), just to join that game. So without a tutor and a higher commitment level myself, I watch and wait for the phone to arrive without burdening the primary builders of the phone-adapted Linux OS. I do okay as a semi-power user of the Linux desktop OS. Shouldn’t those who are too uncomfortable with a bash prompt just wait until everything they want is in the App store and the Pine phone or L5 is as much developed as Android? As it is, I think that Purism is ahead of Pine in phone development. The only difference is that Pine ships their equivalent of Aspen and Birch to everyone who orders one. I am glad that Purism didn’t ship me an earlier version of the L5 as I wait for Evergreen. I wasn’t qualified to do much with those earlier versions and would have been disappointed. A “non technical user” with an early pine phone is probably a very unhappy customer. Should a non-rechnical user even buy one without at least being comfortable with a desktop Linux or at least feeling personally compelled to really want to learn Linux before asking anyone for help? We do help each other in this forum. But I don’t know if a truely non technical person can be helped until after thebuilding of the L5 foundation is complete. My point is that: shouldn’t Pine just keep the price low and politely ask their non technical users to go away and then maybe come back later after they’ve learned something about Linux? Personally, I would offer to pay someone for support before mooching for it after showing no commitment to learning it myself.
For learning to program, I saw somewhere (forums maybe) that Purism devs use Gnome Builder, which is available on most Linux repos. Including PureOS… I’m playing around with it myself.
It isn’t as polished as Visual Studio, Xcode etc, but it works. You can install it from FlatHub, which is the way I did it. There are YouTube videos on it, some good, some not so great.
I am amazed at how far the Librem 5 is. I remember that Apple, super-rich Apple didn’t even get copy/paste working until 2010 (they launched the iPhone in 2007), and if I remember correctly a primitive form of multi-tasking (really just fast app switching) was also 2010.
The early versions of Android were pretty rough, to the point where when I asked the Android devs on my team whether they would recommend iOS or Android to someone technically challenged (thinking of a family member), and they said “Get them the iPhone for now”.
The Librem 5 might not be ready yet for that technically challenged family member, but it has come a really long way! As phones get into the hands of the community, I can see a lot of community members jumping in to get this thing polished and ready for the less tech-savvy consumer.
3 Likes
That’s an understandable understanding, and I happen to agree with it, but the vast majority of people don’t, either because it never occurs to them, they don’t care, they don’t know, or any number of reasons. But free software was never meant to be “free for some.”
Proliferation is, quite probably, the single biggest influence of free software’s success or failure in society. Noobs buying the phone means they’re getting more popular. More popular means there’s potential for market there, to expand and grow. There used to be no Linux phones. Now there are a couple (few?). Perhaps in a year there will be a handful. Maybe after that some large company will pour money into it. That on the surface looks all bad, but that means this wholenthing is getting that much more attention. People will start searching and find Purism. Many of them will buy their phones. And so the growth continues.
So “leeches” have their part, and it is actually a significant one. You can’t have success if nobody buys your phones, and if people outside the market want to get in on it, let them, because some of those will provide the valuable kickback you’re referring to but more than that, everyone deserves the opportunity to be free.
I’m their article they addressed this. They said they considered it, but felt it would be unfair to the noobs, and so instead opted to charge an extra $100 for noob support. A reasonable thing to do, and at $300, the pinephone is still worth the money, in my opinion.
So when we see the poors on the street , we can be like " pffft…pinephone " and we can flash that librem flex on ya’ll
@morgan, when you say “the masses” I get the impression you think of two camps:
Us, the free software lovers / elitists, and
Them, the dumb Facebook+Whatsapp smombies who can’t imagine a life without receiving likes for sharing the latest meme.
Of course, the real world ™ is much more nuanced, and Purism could grow its business several orders of magnitude before even touching that second group.
A lot of non-techy people (including my parents) in my surrounding would be perfectly happy with what the L5 has to offer as of today (minus some polishing maybe).
2 Likes
Generally I don’t say “the masses”, I was replying to somebody else who was using that term. I’m pretty sure I used “” whenever I used the term.
1 Like
Sadly, this is often true. Whenever I write software, I try to bring that sufficiency level lower, so that people can start easier. It’s paid off: we got many layouts from people who never even used Squeekboard (based on the bugs in the layouts), and even put this as a priority in the project principles.
To anyone reading this and hesitating, please don’t follow this advice. It’s more like coming to ask the athlete for advice while they are at the hotel. They may be busy, they may be tired, they may be flattered, they may be jerks, or they may see someone worth mentoring. There are a lot of different personalities you can meet. I have a different piece of advice: if you’re asking for a favor, don’t act entitled. For some, this is a hobby, for others, it’s a job, and they may have other priorities than helping, without it making them bad people.
10 Likes
I don’t think anyone is either right or wrong for not supporting a new person in code-writing efforts. It’s just that there is less forgiveness if the new person isn’t doing their part to learn and the learning curve is steep.
As an Engineer I was exposed to Linux more than twenty years ago. Others at work got me started and answered enough questions that I became self-sufficient as a Linux user after only a few months. It would probably have been much more difficult if I wasn’t exposed to it at work and had help there to get started. Not everyone has that opportunity. It is greatly dependant on whether or not you know someone who can help to get you started.
But the code writing is a whole new level. Even though I have some visual basic and #C programming experience, I don’t know what the open source tools are, where to find them, or anything about the libraries or even github. So I would need a whole lot of dedication and time and possibly a mentor to get started in collaborative open source programming. The person who starts a given project also has a big advantage. When you build the foundation, name everything, and set up the protocols yourself, everything is built around how you personally think. Everyone else adapts and learns whereas for you, everything in the project matches exactly how you think.
2 Likes
If one uses a lot of photo settings… Linux might actually be able to be a little better than Android, or at least better than my current OnePlus One running LineageOS with OpenCamera. I notice that in Linux with guvcview, I can get my cheap ELP webcam from 2014 to pick up the black-on-dark-grey letters of my HHKB better than on my 2014 phone, simply by turning the exposure way up. On my phone, I need a flash to get anything.
I guess the real test of that will come once I get my Librem 5 though, because I believe it actually has the exact same camera sensor as my OnePlus One! That could allow for a very close comparison. Hopefully some good mobile-optimised camera software exists by then with at least as much functionality as guvcview.
Why try to match Google Assistent? IMO, Google Assistent kind of sucks…
My dream for a voice assistant is more like the Linux command line, and combined with AR so… well, I think this excerpt this comment I made on Reddit should explain decently:
GUI design, particularly in its early days, made heavy use of metaphor: desktops, files, folders, recycle bins. It made a lot of sense, because computers were still trying to bootstrap themselves into legitimacy. The ease of implementation of metaphors was one of the huge advantages GUIs wielded over CLIs. Ironically, though, the CLI has embodied an accidental metaphor all along: it’s a conversation.
This makes me think: what makes voice interfaces so much worse than GUIs? Well, a main reason is that it simply can’t understand me with many things, with many things I might want to say even being seemingly impossible for the voice recognition to understand, no matter which accent I use or even if I use TTS, such as song names not in English. I guess the CLI has a similar problem, though, of names of files being in different character sets. What makes the CLI work is that I can just copy and paste any file I can’t write, usually with middle click. But… you can’t do that with a voice interface.
But what if you could? What about some hybrid interface, where there was voice and visual feedback, where I could say “blender /mnt MEGA Blends Mia… yeah, that one” and point to the file I want, or for characters I can’t type and don’t know how to pronounce all of them say “mpv Music Music2 Videos Hejme japanese characters… that one!”? What if I could say part of the file I want, then reach out and grab it to an application I want to send it to (one thing I can’t do right now in the CLI, if I want to send an image to Discord not easily able to be seen with the built-in file browser)? What if I could say “Waterfox… DuckDuck… yeah” (autocompletes to https://duckduckgo.com) then paste an error I found with my keyboard? Couldn’t a hybrid of CLI, GUI, and voice be so much more efficient, being able to say vaguely part of what I want and just get it like with the CLI, but instead quickly saying it with my voice?
Just out of curiosity, I tried to open Waterfox with Google now (a simpler task than dealing with files), just to see if we have anything close, and… no, we do not, not at all. I said “Waterfox” and it calculated a while then searched an unrelated term. I said “Open Waterfox” and it calculated a bit then said “Opening Waterfox” then opened it, not very snappily. It would be faster just to tap the app. I shouldn’t have to say complicated phrases like “open” instead of just the software I want to use, shouldn’t have to say “Use VLC to open Dragon by Two Steps From Hell” (and even that doesn’t work; it just doesn’t understand at first, then searches on the web) instead of “vlc Music Two Steps… yeah” <taps “Two Steps From Hell”, then taps on “Dragon”>. The current way it’s set up, it’s faster for my to play it via mpv on the CLI… with my on-screen keyboard. Current interfaces kind of suck, honestly. I think it could be cool to see a better hybrid that actually works.
Actually, even without dealing with voice nonsense, I would love an easy way for me to select a file in the CLI then drag it to a GUI application. I wonder if such a software exists yet? Just, a command then the file name, which would pop out a window I could drag from, causing the window to disappear but the dragged thing to stay. It could even quickly pop up a directory I could quickly drag a file from, or just drag the entire directory from the “.” directory. Perhaps I should work on that, and later, a more advanced voice/VR interface where it is easy to say things like I type them in the command line and seamlessly switch to selecting and dragging them whenever. That would be really nice and efficient, I think, and maybe then there could finally be a more efficient interface than the CLI? Assuming it could actually understand everything, of course…
It could even implement CLI commands with voice, since the first word is the software name. I could say “cat Elŝutoj slash Domain” click on some things then say “pipe tee ex tee two bin”, to write the equivilent of “cat ~/Elŝutoj/Domainlist/gen.xyz/2020-10-14/xyz_letters_3.txt|txt2bin”, which gives the output of that file but in binary (using https://github.com/happysmash27/bin2txt), or any other routine shell command I might want to use. If it used two many weird symbols, then it could be easy to fall back to a keyboard. I guess there could be some ambiguity problems (what if there is a file called “pipe”), but it is a nice dream I think.
What if we could have something like that instead? In my experience at least, current voice assistants are barely useful at all. Why try to merely meet that quality, instead of going beyond it? It wouldn’t really be able to be VR/AR on the Librem 5, but even on a flat interface… surely it would be nice to be able to say out a file path, then pick out which file one wants when it isn’t as convenient to say it, all within the same interface, instead of having to rely on something like Google Assistent which can’t find something at all most of the time? I don’t think we should aspire to match an interface that barely works for many tasks; rather we should aspire to surpass it, however that may be.
Edit: Just realised you said text-to-speach, not speach-to-text. …Is TTS even a problem on Linux? Espeak works fine.
1 Like
I think the big difference is the computational processing techniques. I have a 5MPix standalone camera from 2008, with a phone-sized sensor. Compared to the pictures I see posted online from modern phone cameras, there are some obvious differences: phones don’t have noise, they have smudges. What they produce looks like artistic impressions rather than (an attempt) to reflect reality.
I personally dig the noisy look. It looks terrible as JPEG, it compresses badly as PNG, but it allows me to guess the underlying sharpness through the noise. When I see a phone picture, the pixel level details are all erased, and I feel the 3000px width doesn’t bring anything: it would look just as good at 1/3 scale.
2 Likes
for me 1024x1024 with L5 using RAW format is dream for post-processing … the limitation is that 1024 res is not usable for PRINT (unless you’re printing for bugs - LOL)
if you’re doing indoor photography 1024x1024 for RAW is almost ideal. 2048x2048 if i had a 43 inch @ 4k monitor but i do NOT …
Well since it is linux, at least you’ll be able to use GIMP on your photo!
Of course fingerpainting a photo on a phone with GIMP may be a bit difficult. But it may lead to a new kind of art.
1 Like
TTS with eSpeak has terrible audio quality. Here is what is available in Linux:
I love to listen to books on my LineageOS phone, but I can’t bear to listen to words being badly pronounced and the stilted word flow of TTS in Linux. The only thing that I have found which doesn’t drive me batty is SVOX Pico, which is non-free according to Debian.