Heater
Posts: 14695
Joined: Tue Jul 17, 2012 3:02 pm

Re: Future of raspberry pi - software related

Tue Jan 14, 2020 10:00 pm

I think that progress bars and now a days spinners are not at all intended to inform you of progress in any meaningful way.

Rather, they are there to catch your attention and keep you amused, in the hope you don't stop paying attention and wonder off to do something else.

This is especially true of spinners on loading web pages. Eyeballs == money on the web.
Memory in C++ is a leaky abstraction .

Heater
Posts: 14695
Joined: Tue Jul 17, 2012 3:02 pm

Re: Future of raspberry pi - software related

Tue Jan 14, 2020 10:03 pm

TimG,

You mean like: https://github.com/WiringPi

Which seems to be getting updates as we speak.
Memory in C++ is a leaky abstraction .

jbudd
Posts: 1127
Joined: Mon Dec 16, 2013 10:23 am

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 1:04 am

Some of these things might already be in place. Sorry if I've missed them!

I wish Raspbian gave me more confidence that I could plug something (USB stick, hard disk, keyboard, monitor, ...) into a running Pi and it would work, and not risk killing the Pi. I have the impression, possibly wrongly, that everything has to be plugged in at boot.

I wish that when I plug in a hard drive it would by default mount somewhere less cryptic than /media/pi/verylongstringofhexadecialcharacters

I wish there was a forum dedicated to the Pi and Internet of Things.

I would like to see absolute beginners advised to go with Etcher rather than NOOBS.

Working out how to get something working usually involves trawling through YouTube, StackExchange & other search hits and the fourums; rarely the official documentation. If there was a Raspberry Wiki we could have an authoritative and updatable source of advice and recipes seperate from the forum.

Heater
Posts: 14695
Joined: Tue Jul 17, 2012 3:02 pm

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 1:24 am

jbudd,
I have the impression, possibly wrongly, that everything has to be plugged in at boot.
Certainly you can plug USB devices in an out at any time. That is the whole idea of USB. As Bill Gates demonstrates here: https://www.youtube.com/watch?v=IW7Rqwwth84
I would like to see absolute beginners advised to go with Etcher rather than NOOBS.
Absolutely.
Working out how to get something working usually involves trawling through...
Sadly that is not just a Pi problem in this modern world. I have to trawl the net to find out how to get almost any device to work as advertised.
Memory in C++ is a leaky abstraction .

trejan
Posts: 1202
Joined: Tue Jul 02, 2019 2:28 pm

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 1:37 am

jbudd wrote:
Wed Jan 15, 2020 1:04 am
I wish Raspbian gave me more confidence that I could plug something (USB stick, hard disk, keyboard, monitor, ...) into a running Pi and it would work, and not risk killing the Pi. I have the impression, possibly wrongly, that everything has to be plugged in at boot.
You don't need to plug things in only at boot. The main concern is with power hungry devices like hard disks which may try to draw more current than is available on the Pi USB ports. It won't damage the Pi but it won't work properly.
jbudd wrote:
Wed Jan 15, 2020 1:04 am
I wish that when I plug in a hard drive it would by default mount somewhere less cryptic than /media/pi/verylongstringofhexadecialcharacters
It already does this but you've not got your drive setup properly. You don't have a label for the filesystem on the drive so it has to use the UUID or serial number. If it is ext then use e2label. If it is FAT then use fatlabel.

Once you've set a label for each filesystem on the drive then you'll get a more friendly mountpoint name.
jbudd wrote:
Wed Jan 15, 2020 1:04 am
Working out how to get something working usually involves trawling through YouTube, StackExchange & other search hits and the fourums; rarely the official documentation. If there was a Raspberry Wiki we could have an authoritative and updatable source of advice and recipes seperate from the forum.
People tried to do that with the eLinux wiki but interest died off. It wasn't official but I don't see an authoritive Wiki being much better. Writing a forum/StackExchange post is far easier than writing a Wiki article about it.

A lot of people seem to favour video guides now for some reason. I still prefer written guides instead of having to trawl through a 30 minute video for the 2 minutes of actual useful information...

X-Gen
Posts: 87
Joined: Fri Dec 06, 2019 12:08 pm

Re: Future of raspberry pi

Wed Jan 15, 2020 1:41 am

jamesh wrote:
Mon Jan 13, 2020 9:31 pm
Imperf3kt wrote:
Mon Jan 13, 2020 9:18 pm
jamesh wrote:
Mon Jan 13, 2020 5:19 pm


We have some cracking stuff in the pipeline...lots of smiley faces in our future!
No more magic blue smoke?


Personally I'm quite content with the Pi4 hardware.
I'm personally hoping for software improvements over more raw power.
We have a HW team and a software team. We can do both!

But what areas would you like improvements in? We are already doing work on VLC and Chromium, 64 bit support, bug fixing, power improvements, USB boot and general all round bug fixing.
Funny, because most people's dreams are shattered, when they come to the forum, expressing their wishes on a future pi, and get barked down.
However, now that you've asked,
Personally I am looking for 2 types of devices,

One: a Pi zero alternative (that's about half size, with only few GPIO pins) and a quad core CPU with 2GB of RAM. Ram and ROM (emmc or ssd or something) can be soldered on the bottom, allowing it to receive cooling from a metal housing, while the CPU remains on the top; and wifi antennas are built into the board. Serving as a micro pc to do coding on.

Two: A server Pi, hosting a massive amount of cores on one board, for projects like Boinc.

I'm realistic in that I'll get neither.
But like anyone else, I just want to see tens of cores crunching through data, on a cheap board, with relatively stable software; as well as something that could take the edge off the Pi Zero W; in that it'd run the Raspbian OS a bit more smoothly, while being smaller in size.

For the Pi4, of course lower power consumption, even smaller size. And perhaps just only a few GPIO pins (like 3 power pins, and 5 data pins or so). A lot of people aren't using them at all.
If BT and Wifi are on the board, all I need is one (mini)HDMI, and 2 USB ports, Ethernet port, and audio jack (and power).
Last edited by X-Gen on Wed Jan 15, 2020 1:59 am, edited 1 time in total.

Heater
Posts: 14695
Joined: Tue Jul 17, 2012 3:02 pm

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 1:50 am

X-Gen,
I just want to see tens of cores crunching through data,
Have you ever tried to write a program?

Have you ever tried to get that program to reach a result quicker by using many cores?

It is not easy.

In the mean time recent Pi have four cores for you to practice on.
Memory in C++ is a leaky abstraction .

X-Gen
Posts: 87
Joined: Fri Dec 06, 2019 12:08 pm

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 1:57 am

Heater wrote:
Wed Jan 15, 2020 1:50 am
X-Gen,
I just want to see tens of cores crunching through data,
Have you ever tried to write a program?

Have you ever tried to get that program to reach a result quicker by using many cores?

It is not easy.

In the mean time recent Pi have four cores for you to practice on.
Boinc utilizes all cores just fine.
Most of AI workloads, crunching, folding, and imho the future of computing, is all parallel workloads.

I realize it's not everyone's cup of tea, but man, just showing htop, with half of the screen being CPU bars, lit up like a Christmas tree, is worth it!

X-Gen
Posts: 87
Joined: Fri Dec 06, 2019 12:08 pm

Re: Future of raspberry pi

Wed Jan 15, 2020 2:07 am

ejolson wrote:
Tue Jan 14, 2020 2:14 am
jamesh wrote:
Mon Jan 13, 2020 9:31 pm
Imperf3kt wrote:
Mon Jan 13, 2020 9:18 pm


No more magic blue smoke?


Personally I'm quite content with the Pi4 hardware.
I'm personally hoping for software improvements over more raw power.
We have a HW team and a software team. We can do both!

But what areas would you like improvements in? We are already doing work on VLC and Chromium, 64 bit support, bug fixing, power improvements, USB boot and general all round bug fixing.
I'm still waiting for the built-in sandwich maker that was mentioned in the post

https://www.raspberrypi.org/forums/view ... 5#p1467338

In the meantime, I would really like a CUDA-like programming language for the VC6 to allow young programmers an opportunity to learn how to use a GPU accelerator in their code.
As I recall, it was a sandwich toaster.
The Pi 4b is quite the toaster, but will need a few more overclocking voltage settings, before my sandwiches get crispy on that Pi.
While on it, can the Pi foundation create a 1000 Watt Pi?
I need to find a way to heat my livingroom, and a 1kw, 100 core Pi might be just what I was looking for!

X-Gen
Posts: 87
Joined: Fri Dec 06, 2019 12:08 pm

Re: Future of raspberry pi

Wed Jan 15, 2020 2:13 am

jamesh wrote:
Tue Jan 14, 2020 7:00 am
Remember, I said software work, not hardware. Hardware requests just get into a mess, but it would be interesting to know where people think the software needs work.

OpenCL/CUDA is not easy, and require external contractors as we don't have the internal expertise.
Jamesh,
Perhaps not Pi4B or future related,
But would it be possible to get an even smaller, trimmed down version of Raspbian for the Pi zero, 1, 2, and 3a?
The ones with low ram, low disk space (or ssd space), and slower processors can't really run Chromium, and might benefit from an entirely new desktop (one that's like ultra light).

I don't know if the idea is a bit far fetched.
My previous hardware suggestion was suggested in jest, but with this I'm just wondering if some tweaking can be done to make raspbian work better on the lower end Pis?
A working browser on low RAM should be one, but perhaps there's a way to get some gurus on the job that can pinch out every unnecessary byte of RAM out of the desktop.?

ankith26
Posts: 234
Joined: Mon Mar 25, 2019 11:08 am
Location: /home/pi/pythonprojects/test.py
Contact: Website

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 2:38 am

Just so you you know, this thread has turned to a thread of software discussion so mods won’t like your hardware question. Anyway, if you want a lite os, go for raspbian lite or raspbian desktop(without applications).
Although I never tested these on any lower gen pi so I am not sure how they work.
You have other OS too to choose from
I sat thinking for 5 minutes on what to put here. Finally I put something like this.
Check out my github page @ https://github.com/ankith26

ejolson
Posts: 4454
Joined: Tue Mar 18, 2014 11:47 am

Re: Future of raspberry pi

Wed Jan 15, 2020 2:48 am

X-Gen wrote:
Wed Jan 15, 2020 2:13 am
jamesh wrote:
Tue Jan 14, 2020 7:00 am
Remember, I said software work, not hardware. Hardware requests just get into a mess, but it would be interesting to know where people think the software needs work.

OpenCL/CUDA is not easy, and require external contractors as we don't have the internal expertise.
Jamesh,
Perhaps not Pi4B or future related,
But would it be possible to get an even smaller, trimmed down version of Raspbian for the Pi zero, 1, 2, and 3a?
The ones with low ram, low disk space (or ssd space), and slower processors can't really run Chromium, and might benefit from an entirely new desktop (one that's like ultra light).

I don't know if the idea is a bit far fetched.
My previous hardware suggestion was suggested in jest, but with this I'm just wondering if some tweaking can be done to make raspbian work better on the lower end Pis?
A working browser on low RAM should be one, but perhaps there's a way to get some gurus on the job that can pinch out every unnecessary byte of RAM out of the desktop.?
Another software consideration would be optimizing this forum and the main web pages to work with netsurf, lynx or perhaps dillo running on a Pi Zero.

Parallel is definitely important enough to be part of an introductory curriculum. It makes no sense to rewind the mistake of teaching word processing in place of programming all the way back to the 8-bit microcomputers of the past. Instead, it would be better to move forward with a curriculum adapted to practical results related to programming the multi-core machines available today.

ankith26
Posts: 234
Joined: Mon Mar 25, 2019 11:08 am
Location: /home/pi/pythonprojects/test.py
Contact: Website

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 2:51 am

But python, the main language for teaching beginners does not support parallel multithreaded tasking.
And I guess java is too much for teaching beginners.
I sat thinking for 5 minutes on what to put here. Finally I put something like this.
Check out my github page @ https://github.com/ankith26

ejolson
Posts: 4454
Joined: Tue Mar 18, 2014 11:47 am

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 3:06 am

ankith26 wrote:
Wed Jan 15, 2020 2:51 am
But python, the main language for teaching beginners does not support parallel multithreaded tasking.
And I guess java is too much for teaching beginners.
It looks like you've discovered another great software request: Develop a new programming language for beginners to learn how to write parallel code.

ankith26
Posts: 234
Joined: Mon Mar 25, 2019 11:08 am
Location: /home/pi/pythonprojects/test.py
Contact: Website

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 3:22 am

But that is not in the scope of raspberry pi development. Some third party software will have to work
I sat thinking for 5 minutes on what to put here. Finally I put something like this.
Check out my github page @ https://github.com/ankith26

User avatar
Imperf3kt
Posts: 3370
Joined: Tue Jun 20, 2017 12:16 am
Location: Australia

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 3:34 am

ankith26 wrote:
Wed Jan 15, 2020 2:42 am
And about NOOBS and etcher, is etcher considered the better option. What sort of advantages does it have over NOOBS. When installing stretch I used NOOBS and not etcher simply because it was easier.
You've got that backwards, etcher is the easier to use of the two.
NOOBS has several extra steps you need to do, etcher is simply select file, press write done.


Regarding USB - you can't actually plug anything in on a Pi Zero / zero W after it has booted, even a mouse will cause it to reboot.
https://www.raspberrypi.org/forums/view ... 8&t=237554
Last edited by Imperf3kt on Wed Jan 15, 2020 3:39 am, edited 1 time in total.
55:55:44:44:4C
52:4C:52:42:41

ejolson
Posts: 4454
Joined: Tue Mar 18, 2014 11:47 am

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 3:37 am

ankith26 wrote:
Wed Jan 15, 2020 3:22 am
But that is not in the scope of raspberry pi development. Some third party software will have to work
If Google can create Go, Apple can create Swift, Mozilla can create Rust, Sun can create Java and Microsoft can create C#, I see no reason the company which makes the Raspberry Pi could not also make a programming language. The main difficulty (as demonstrated by Go and Swift) is finding good name.

What do you think about calling it the
  • Universal
  • Parallel
  • Teaching
  • Object-oriented
  • Nano-language
or do you have a better idea?

ankith26
Posts: 234
Joined: Mon Mar 25, 2019 11:08 am
Location: /home/pi/pythonprojects/test.py
Contact: Website

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 3:50 am

But it is a very difficult task to create a programming language, even for a big foundation.
But upgrading on an existing one is not as difficult( but is still way more difficult than you imagine)
If I was given the task to create a programming language I would take python and build on it with a team.
In order to achieve true multithreading, you would have to completely revamp python.
I would name it PiMax or something (I am not good at naming stuff)
I sat thinking for 5 minutes on what to put here. Finally I put something like this.
Check out my github page @ https://github.com/ankith26

W. H. Heydt
Posts: 11653
Joined: Fri Mar 09, 2012 7:36 pm
Location: Vallejo, CA (US)

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 4:16 am

jbudd wrote:
Wed Jan 15, 2020 1:04 am
I wish that when I plug in a hard drive it would by default mount somewhere less cryptic than /media/pi/verylongstringofhexadecialcharacters
Give the drive a short, meaningful name. IIRC, gparted can do that.
I would like to see absolute beginners advised to go with Etcher rather than NOOBS.
Quite a few people do just that.

W. H. Heydt
Posts: 11653
Joined: Fri Mar 09, 2012 7:36 pm
Location: Vallejo, CA (US)

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 4:20 am

ejolson wrote:
Wed Jan 15, 2020 3:06 am
ankith26 wrote:
Wed Jan 15, 2020 2:51 am
But python, the main language for teaching beginners does not support parallel multithreaded tasking.
And I guess java is too much for teaching beginners.
It looks like you've discovered another great software request: Develop a new programming language for beginners to learn how to write parallel code.
Is that anything like developing a programming language that a "end user" can write in so that they don't need a programmer to do it for them? You know, like COBOL, RPG, Mark IV, SQL and a myriad of others?

ankith26
Posts: 234
Joined: Mon Mar 25, 2019 11:08 am
Location: /home/pi/pythonprojects/test.py
Contact: Website

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 5:14 am

Honestly, it’s getting a bit confusing here, all are talking about different things, let us all stick to a common topic and discuss on it
I sat thinking for 5 minutes on what to put here. Finally I put something like this.
Check out my github page @ https://github.com/ankith26

ejolson
Posts: 4454
Joined: Tue Mar 18, 2014 11:47 am

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 5:17 am

W. H. Heydt wrote:
Wed Jan 15, 2020 4:20 am
ejolson wrote:
Wed Jan 15, 2020 3:06 am
ankith26 wrote:
Wed Jan 15, 2020 2:51 am
But python, the main language for teaching beginners does not support parallel multithreaded tasking.
And I guess java is too much for teaching beginners.
It looks like you've discovered another great software request: Develop a new programming language for beginners to learn how to write parallel code.
Is that anything like developing a programming language that a "end user" can write in so that they don't need a programmer to do it for them? You know, like COBOL, RPG, Mark IV, SQL and a myriad of others?
I think the idea behind COBOL was not that a non-specialist could write the code, but that the end user could read and understand it after it was written. Measured by this more limited goal, COBOL was much more successful than people realize.

Teaching languages are different in nature than general-purpose production languages. They are much more flexible and can also be much simpler in design. While Python started as a teaching language, it is now as popular as Basic was in the 70's and 80's. Whether it will have the longevity of Basic or that of C is yet to be seen. The point, well taken, however, is that Python was never meant for teaching parallel programming.

It is my feeling that adding features to Python for parallel programming is likely to be difficult. An entirely different design could be simpler and allow the student and teacher to focus directly on concepts related to parallel algorithms and the efficient use of multi-core hardware.

As C++ demonstrates, adding features to a good language doesn't necessarily make it better or easier to use. On the other hand, languages like Oberon and Go demonstrate that simplification by removing features can result in a language easier to learn and fewer bugs.

The reason the request to develop a language to teach parallel programming belongs here is because it's software that fits with the educational mission of the Raspberry Pi Foundation.

ankith26
Posts: 234
Joined: Mon Mar 25, 2019 11:08 am
Location: /home/pi/pythonprojects/test.py
Contact: Website

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 5:26 am

In order for Python to achieve true multithreading, the GIL(which is everyone’s friend) has to be removed.
But GIL is important for the python interpreter to work. In order to solve this problem, python should be a compiled language(which is totally against its core beliefs) or it must run one some virtual machine using JIT compilers(like hove java does, it is quite successfully doing this).
So the real solution is to make a language with python-like syntax and java-like JIT compiler and virtual machine.
You get all the advantages of python and are freed of its downsides(speed, multithreading etc) this way
But python wants to remain an interpreted language so I don’t see the above solution working.
I would definitely love to work on such an open-source project, but I do not have so much technical know how on such topics, you see, I am only good at python itself.
I sat thinking for 5 minutes on what to put here. Finally I put something like this.
Check out my github page @ https://github.com/ankith26

W. H. Heydt
Posts: 11653
Joined: Fri Mar 09, 2012 7:36 pm
Location: Vallejo, CA (US)

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 5:29 am

ejolson wrote:
Wed Jan 15, 2020 5:17 am
W. H. Heydt wrote:
Wed Jan 15, 2020 4:20 am
ejolson wrote:
Wed Jan 15, 2020 3:06 am

It looks like you've discovered another great software request: Develop a new programming language for beginners to learn how to write parallel code.
Is that anything like developing a programming language that a "end user" can write in so that they don't need a programmer to do it for them? You know, like COBOL, RPG, Mark IV, SQL and a myriad of others?
I think the idea behind COBOL was not that a non-specialist could write the code, but that the end user could read and understand it after it was written. Measured by this more limited goal, COBOL was much more successful than people realize.
That's part of it, but it was intended that "non-programmers" would be able to write at least relatively simple programs. Part of the language design was to be English-like (that is, so that statements looked like natural language..."ADD AMOUNT TO AMOUNT-SUM.")
Teaching languages are different in nature than general-purpose production languages. They are much more flexible and can also be much simpler in design. While Python started as a teaching language, it is now as popular as Basic was in the 70's and 80's. Whether it will have the longevity of Basic or that of C is yet to be seen. The point, well taken, however, is that Python was never meant for teaching parallel programming.
Hardly the first language intended for instruction and then having people write production programs in it. But then, the first programming language I learned was FORTRAN. (The second was SPS, which very few--if any--people here have ever heard of.)
The reason the request to develop a language to teach parallel programming belongs here is because it's software that fits with the educational mission of the Raspberry Pi Foundation.
I wouldn't dream of arguing against that.

jamesh
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 25382
Joined: Sat Jul 30, 2011 7:41 pm

Re: Future of raspberry pi - software related

Wed Jan 15, 2020 6:20 am

When people are talking about parallel programming, do they mean threading or clustering?
Principal Software Engineer at Raspberry Pi (Trading) Ltd.
Contrary to popular belief, humorous signatures are allowed. Here's an example...
“I own the world’s worst thesaurus. Not only is it awful, it’s awful."

Return to “General discussion”