You know, we're going to see Apple announce this exact device next year at WWDC.
But in all seriousness, I wouldn't be surprised if that actually happened. It's the logical end of the present round of form-factor updates, until the phone can be made to be invisible itself.
Termux works, and I do use it, but it has weird issues with accessing parts of the filesystem and convincing tools to work right due to Android's limitations. I haven't looked at GNURoot. As to the 10 GB limit--that part is real and frustrating and stupidly artificial. I'm already using the transfer! I tried doing without the home connection, but I caved in after maybe 3 months. It does mean that any heavy duty network traffic for the laptop needs to be at home, work, or somewhere like a library, as you said.
There are real issues with the setup, but the only way I see them getting addressed is with individual effort. I find the services of many large corporations extremely convenient, and really don't want to void the warranty on the newest device I use (afterwards, who cares? I've got an original T-Mobile G1, rooted, still works just fine, if horribly out of date), but I would also prefer to see computing devices owned by the people holding them.
ARM is good for consolidating on a low-power many-core platform. Big iron, as mentioned elsewhere, still has real value. It just doesn't need to be in everything. Going forward, I'd really love to see ARM replaced with RISC-V, so that open hardware might actually be a useful thing.
Sure, that's a valid use case. The big thing I miss about current generation unrooted devices is the lack of ability to compile (to machine code) and execute directly on the hardware. That's the big showstopper, and the reason I still have a laptop and server dedicated to providing "backend" support for the phone's role as primary device/smart terminal.
The range of coding that can be done with an Android device, unrooted, right now, is in pretty rough shape. I might use QPython and QPython3 to do a few small things directly on the phone, but I run into limits pretty quick. If it's useful, that code is moved into a remote git repo and expanded with usage on a laptop, where I can more easily make use of various libraries and system resources. That's how the latest project got its start: I needed some answers on a project, and all I had handy was my phone. Later, it got a frontend for desktop usage.
But, it's rare that I don't have connectivity. I'm happy to pay for the "unlimited" data plan. Between my wife and I, we blow through 60+ GB of traffic a month on two phones alone (we both have laptops, but both use a smartphone as a primary device). And frequently, the phone *is* the WiFi for my laptop. A more typical scenario is editing directly in Vim over an SSH connection, or, less commonly, grabbing a copy of the code to to work on locally in DroidVim via SFTP/git.
As to tooling, I'm not an IDE person. I really like Vim, and that makes it very easy to be happy *and* productive on minimal systems--basically anywhere I can get a halfway decent terminal. (I'm a keyboard person all day long, so anything to avoid the mouse.) My concession to IDEs is NERDTree and a few bits of shell script. That also makes it trivial to dump my usual setup in a brand new machine, VM, headless, or otherwise, and have it be set up to do useful work.
A little off-topic, but tl;dr: We need low-level hardware access on mobile platforms to make them really viable without having to screw around trying to root phones we should own in the first place. (I've said I like ARM, but I'd also really like to see a halfway decent, $30 RISC-V board.) And... to really make them useful, they still need fast cellular radios. My primary ISP, from a usage standpoint, is my phone provider, not my home ISP.
This model isn't for everyone, but, I always remember wanting more out of my PDAs and calculators. I wanted a full-blown desktop replacement. We are nearly there, but low-level access is the big killer.
Nifty board! As a matter of fact, I just bought 2 OrangePi PC2's with the Allwinner H5 onboard (quad-core A53). 43$ shipped for the both of them, though less powerful than this board or my phone by far.
However, I want to try them out instead of the RK3999 chipset because there's indications of OpenBSD support in addition to Armbian and Lubuntu, and I think it'd be fun to play with some ideas I have at a bare metal level. If someone got the RK3999 running OpenBSD, I'd be much more interested, because there's more powerful boards available--like the one you showed.
I'm leaning toward this future. I've switched to using a Galaxy S9+ as my daily computing device, and the Thinkpad is reserved for longer coding sessions.
But, more and more, I just use a bluetooth keyboard from Omoton and use Termius to SSH in to my servers and the laptop from my phone. File Manager+ has SFTP support, among many features. I have browsers, VNC, DroidVIM (really an excellent port), etc.. Do I need a monitor? I cast the screen to a ChromeCast.
I write code primarily. If I need horsepower, I spin up a VM from my phone and use SFTP/Git to load up some code, and SSH in to administer it. The phone fits in my pocket while I'm running around between the machines at work, too. Nice bonus there, not lugging the laptop itself.
I personally look forward to this. I like the ARM ISA. I thought Torvalds was being short-sighted. For starters, it's a more popular platform by number of chips in the wild. These Intel and AMD CISC designs are all RISC under the hood now, anyway.
We're just doing away with the cruft of a legacy architecture that grew off track.
I found this comment hilariously incisive, as the OP called them out for not explaining it, and then promptly didn't explain it for those of us who might be curious.
Well... I shot my mouth off on Slashdot and came up wrong it would seem. I wasn't aware of the OpenDOS announcement.
I have to admit you have a compelling argument for Caldera being first to put out a license using that term. The mailing list entry, the license text itself, are all evocative of what Open Source became.
You're right, and I think it'd be interesting to hear ESR and Perens' take on them.
Regardless of how those technologies might have been acquired, you received them via the Apollo program. Your claim that you received nothing is provably wrong. QED.
You're crying over lost opportunity cost. Unless your research gives you a time machine, I suggest you focus on the present.
Nonetheless, interesting little post. I do feel like it confirms the original authorship of the term as ESR. It's claimed that Perens claims ESR invented it, regardless of the disagreement on timing.
"Raymond and I had met occassionally at the Hacker's Conference, a by-invitation-only gathering of creative and unconventional programmers. We had corresponded on various subjects via e-mail. He contacted me in February of 1997 with the idea for Open Source."
I agree with you here, but I don't have mod points.
A specific person probably used the term "Open Source" first, and that person has given a definition and description of intent, and control has been placed in the hands of the OSI.
As far as I can see, they've got a monopoly on the meaning of that term. I'm not really sure why drinkypoo tried to claim anyone was equating it with Free Software, a thing that the inventor of the term Open Source stated he was deliberately trying to be distinct from. Smells like burnt strawman.
Also sometimes you have to boilerplate some code for a while before you can coalesce it all into something simpler - like one of Blaise Pascal's letters that came with a forward, saying essentially "I'm sorry this letter is so long, I did not have the time to make it shorter".
I agree with you on this. I find that writing repetitive code or boilerplate for a short period allows me to more easily see what *should* be optimized, avoiding the temptation to prematurely optimize something (which might introduce complexity instead of the desired simplification).
Optimization doesn't just refer to runtime performance.
Definitely not dead or gravely wounded. Python may not be the new hotness, but it's absolutely a go-to language for new code for me. I have been, in point of fact, writing new Python code for my employer recently. It's still the tool I reach for when I need something working Now. I don't spend time writing code for small reasons either--if it isn't making us significantly more money than it costs to build, I don't build it.
Python makes us money, and that's the bottom line.
Now, I'm no picky eater. I've been coding 26 years, and I'll use the correct tool for the job any time I can, rather than "work like an asshole" [favorite saying of an old boss]. For me, that's often Python, but I won't hesistate to drop down to C/C++ or assembler if that's the appropriate tool--or work with Java or C# or Objective-C or Clojure or Ruby or Javascript or Erlang or Haskell or HTML or CSS (and once upon a time, BASIC, Pascal, et al...).
Why are people so stuck on one language or one way of thinking? I was listening to a software engineering podcast recently, and this guy with 8 years of experience is saying he thinks he knows it all now. Well, sorry to break it to you, 8 Year Master, but after 26+ years of this, I've realized that I will never learn it all (even though I keep trying).
Pay your bills first, keep your deadlines and promises, go home to your loved ones, and do things that expand your horizons.
Java, Python, C, Go, Rust: these things aren't important in the same way.
Same here. I've got a laptop with what I expect to be my last Windows install--and it only exists to run a Debian VM right now. Windows 7 is the end of the road for me.
I'm also kicking the Debian habit because it's become impossible to avoid systemd. Once that box dies, it's getting replaced with a Devuan box. I've invested some time recently in brushing up on OpenBSD, and that's my new solution for servers.
*nix wasn't broke. Didn't need fixing. Windows was tolerably useful, didn't need fixing, but they broke it anyway. I'm not unwilling to pay for software--I'm unwilling to pay for garbage I don't own. That's why Devuan and the OpenBSD Foundation made my list for donations over the holidays.
The Hidden Life of Trees by Peter Wohlleben Koevoet! by Jim Hooper Secrets and Lies by Marléne Burger & Chandré Gould The Push by Tommy Caldwell Five Past Midnight in Bhopal by Dominique Lapierre and Javier Moro Lab Girl by Hope Jahren A Novel and Efficient Synthesis of Cadaverine by SA Scoggin
Slashdot Mirror
Do what other municipalities did to coin miners. Raise the cost of high volume electricity to that consumer.
If you're going to slam the grid, you get to help improve it.
as the SoC gets highe
My systems on chip only smoke the magic smoke, not that reefer stuff.
You know, we're going to see Apple announce this exact device next year at WWDC.
But in all seriousness, I wouldn't be surprised if that actually happened. It's the logical end of the present round of form-factor updates, until the phone can be made to be invisible itself.
Termux works, and I do use it, but it has weird issues with accessing parts of the filesystem and convincing tools to work right due to Android's limitations. I haven't looked at GNURoot. As to the 10 GB limit--that part is real and frustrating and stupidly artificial. I'm already using the transfer! I tried doing without the home connection, but I caved in after maybe 3 months. It does mean that any heavy duty network traffic for the laptop needs to be at home, work, or somewhere like a library, as you said.
There are real issues with the setup, but the only way I see them getting addressed is with individual effort. I find the services of many large corporations extremely convenient, and really don't want to void the warranty on the newest device I use (afterwards, who cares? I've got an original T-Mobile G1, rooted, still works just fine, if horribly out of date), but I would also prefer to see computing devices owned by the people holding them.
ARM is good for consolidating on a low-power many-core platform. Big iron, as mentioned elsewhere, still has real value. It just doesn't need to be in everything. Going forward, I'd really love to see ARM replaced with RISC-V, so that open hardware might actually be a useful thing.
Sure, that's a valid use case. The big thing I miss about current generation unrooted devices is the lack of ability to compile (to machine code) and execute directly on the hardware. That's the big showstopper, and the reason I still have a laptop and server dedicated to providing "backend" support for the phone's role as primary device/smart terminal.
The range of coding that can be done with an Android device, unrooted, right now, is in pretty rough shape. I might use QPython and QPython3 to do a few small things directly on the phone, but I run into limits pretty quick. If it's useful, that code is moved into a remote git repo and expanded with usage on a laptop, where I can more easily make use of various libraries and system resources. That's how the latest project got its start: I needed some answers on a project, and all I had handy was my phone. Later, it got a frontend for desktop usage.
But, it's rare that I don't have connectivity. I'm happy to pay for the "unlimited" data plan. Between my wife and I, we blow through 60+ GB of traffic a month on two phones alone (we both have laptops, but both use a smartphone as a primary device). And frequently, the phone *is* the WiFi for my laptop. A more typical scenario is editing directly in Vim over an SSH connection, or, less commonly, grabbing a copy of the code to to work on locally in DroidVim via SFTP/git.
As to tooling, I'm not an IDE person. I really like Vim, and that makes it very easy to be happy *and* productive on minimal systems--basically anywhere I can get a halfway decent terminal. (I'm a keyboard person all day long, so anything to avoid the mouse.) My concession to IDEs is NERDTree and a few bits of shell script. That also makes it trivial to dump my usual setup in a brand new machine, VM, headless, or otherwise, and have it be set up to do useful work.
A little off-topic, but tl;dr: We need low-level hardware access on mobile platforms to make them really viable without having to screw around trying to root phones we should own in the first place. (I've said I like ARM, but I'd also really like to see a halfway decent, $30 RISC-V board.) And... to really make them useful, they still need fast cellular radios. My primary ISP, from a usage standpoint, is my phone provider, not my home ISP.
This model isn't for everyone, but, I always remember wanting more out of my PDAs and calculators. I wanted a full-blown desktop replacement. We are nearly there, but low-level access is the big killer.
Nifty board! As a matter of fact, I just bought 2 OrangePi PC2's with the Allwinner H5 onboard (quad-core A53). 43$ shipped for the both of them, though less powerful than this board or my phone by far.
However, I want to try them out instead of the RK3999 chipset because there's indications of OpenBSD support in addition to Armbian and Lubuntu, and I think it'd be fun to play with some ideas I have at a bare metal level. If someone got the RK3999 running OpenBSD, I'd be much more interested, because there's more powerful boards available--like the one you showed.
I'm leaning toward this future. I've switched to using a Galaxy S9+ as my daily computing device, and the Thinkpad is reserved for longer coding sessions.
But, more and more, I just use a bluetooth keyboard from Omoton and use Termius to SSH in to my servers and the laptop from my phone. File Manager+ has SFTP support, among many features. I have browsers, VNC, DroidVIM (really an excellent port), etc.. Do I need a monitor? I cast the screen to a ChromeCast.
I write code primarily. If I need horsepower, I spin up a VM from my phone and use SFTP/Git to load up some code, and SSH in to administer it. The phone fits in my pocket while I'm running around between the machines at work, too. Nice bonus there, not lugging the laptop itself.
I personally look forward to this. I like the ARM ISA. I thought Torvalds was being short-sighted. For starters, it's a more popular platform by number of chips in the wild. These Intel and AMD CISC designs are all RISC under the hood now, anyway.
We're just doing away with the cruft of a legacy architecture that grew off track.
I found this comment hilariously incisive, as the OP called them out for not explaining it, and then promptly didn't explain it for those of us who might be curious.
+1, Funny.
Zero tolerance for the win. This is also the root cause of many highly escalated fights in our schools.
Well... I shot my mouth off on Slashdot and came up wrong it would seem. I wasn't aware of the OpenDOS announcement.
I have to admit you have a compelling argument for Caldera being first to put out a license using that term. The mailing list entry, the license text itself, are all evocative of what Open Source became.
You're right, and I think it'd be interesting to hear ESR and Perens' take on them.
Well said, +1, Informative.
Regardless of how those technologies might have been acquired, you received them via the Apollo program. Your claim that you received nothing is provably wrong. QED.
You're crying over lost opportunity cost. Unless your research gives you a time machine, I suggest you focus on the present.
Are you that ignorant?
Didn't we just have an article yesterday on how space actually starts much further away if by space you mean "not Earth's atmosphere"?
Why oh why would you taunt us with the spectre of Clippy?
Nonetheless, interesting little post. I do feel like it confirms the original authorship of the term as ESR. It's claimed that Perens claims ESR invented it, regardless of the disagreement on timing.
"Raymond and I had met occassionally at the Hacker's Conference, a by-invitation-only gathering of creative and unconventional programmers. We had corresponded on various subjects via e-mail. He contacted me in February of 1997 with the idea for Open Source."
They're called homonyms. Open Source intel is not Open Source software.
Stupid bleeping...I'm not sure how I wrote "probably". A specific person, ESR, definitely used it first.
I agree with you here, but I don't have mod points.
A specific person probably used the term "Open Source" first, and that person has given a definition and description of intent, and control has been placed in the hands of the OSI.
As far as I can see, they've got a monopoly on the meaning of that term. I'm not really sure why drinkypoo tried to claim anyone was equating it with Free Software, a thing that the inventor of the term Open Source stated he was deliberately trying to be distinct from. Smells like burnt strawman.
Also sometimes you have to boilerplate some code for a while before you can coalesce it all into something simpler - like one of Blaise Pascal's letters that came with a forward, saying essentially "I'm sorry this letter is so long, I did not have the time to make it shorter".
I agree with you on this. I find that writing repetitive code or boilerplate for a short period allows me to more easily see what *should* be optimized, avoiding the temptation to prematurely optimize something (which might introduce complexity instead of the desired simplification).
Optimization doesn't just refer to runtime performance.
Definitely not dead or gravely wounded. Python may not be the new hotness, but it's absolutely a go-to language for new code for me. I have been, in point of fact, writing new Python code for my employer recently. It's still the tool I reach for when I need something working Now. I don't spend time writing code for small reasons either--if it isn't making us significantly more money than it costs to build, I don't build it.
Python makes us money, and that's the bottom line.
Now, I'm no picky eater. I've been coding 26 years, and I'll use the correct tool for the job any time I can, rather than "work like an asshole" [favorite saying of an old boss]. For me, that's often Python, but I won't hesistate to drop down to C/C++ or assembler if that's the appropriate tool--or work with Java or C# or Objective-C or Clojure or Ruby or Javascript or Erlang or Haskell or HTML or CSS (and once upon a time, BASIC, Pascal, et al...).
Why are people so stuck on one language or one way of thinking? I was listening to a software engineering podcast recently, and this guy with 8 years of experience is saying he thinks he knows it all now. Well, sorry to break it to you, 8 Year Master, but after 26+ years of this, I've realized that I will never learn it all (even though I keep trying).
Pay your bills first, keep your deadlines and promises, go home to your loved ones, and do things that expand your horizons.
Java, Python, C, Go, Rust: these things aren't important in the same way.
Same here. I've got a laptop with what I expect to be my last Windows install--and it only exists to run a Debian VM right now. Windows 7 is the end of the road for me.
I'm also kicking the Debian habit because it's become impossible to avoid systemd. Once that box dies, it's getting replaced with a Devuan box. I've invested some time recently in brushing up on OpenBSD, and that's my new solution for servers.
*nix wasn't broke. Didn't need fixing. Windows was tolerably useful, didn't need fixing, but they broke it anyway. I'm not unwilling to pay for software--I'm unwilling to pay for garbage I don't own. That's why Devuan and the OpenBSD Foundation made my list for donations over the holidays.
Yeah, this wasn't even subtle. It reads like a press release.
Here's a few recent books I've read:
The Hidden Life of Trees by Peter Wohlleben
Koevoet! by Jim Hooper
Secrets and Lies by Marléne Burger & Chandré Gould
The Push by Tommy Caldwell
Five Past Midnight in Bhopal by Dominique Lapierre and Javier Moro
Lab Girl by Hope Jahren
A Novel and Efficient Synthesis of Cadaverine by SA Scoggin