The Icon Bar: MUG 2023 Show report

MUG 2023 Show report

Posted by Mark Stephens on 15:30, 1/7/2023 | Shows, News

The second MUG online meet-up took place today on Zoom. This is an online reboot of the old MUG Midsummer show.

The show consisted of 5 live talks (one after another) and a general chat area. All the presenters shared their screens and it was a very interactive event (I really enjoyed it). You could stay for the whole show or dip in and out. There were about 40 people attending most of the time.

The Show fills a real gap for the RISC OS world (not everyone can get to Bristol, Wakefield and London). So I hope it will become a regular fixture of the RISC OS calendar.

If you missed the event, the talks should appear in due course online. Here are my notes from the talks.

Amcog Games
First session was from Tony at Amcog Games.

We had a hands on demo (in StrongED) of RDSP playing samples and generating sounds. RDSP was first released in 2017. It takes lots of ideas from not just the BBC/Archimedes but also Commodore64 and Amiga as well. Uses fixed point to 14 binary places. It was written from scratch. RDSP includes full source code with a licence allowing you do do what you want with it.

There was also some nice interactive explanation on how sound work and wave formations.
Next was the Amcog Games Development kit which is used for all the games. It has lots of features to speed-up games development and do all 'the heavy lifting' for you. There are lots of samples showing how to use game features (sparky-sace study). You can buy the Games Development kit on !Store.

Next game is going to be escape to the Light and all games result in improvements to the Games library.

At the end there was plenty of time for questions.

Several games have been ported to other platforms but all written on RISC OS first.

In reply to what would you like so see in RISC OS, more 3D acceleration and more 3D texture routines.

Android attack on BBC micro was his favourite game as the first he played. Escape from Exeria is his favourite RISC OS game.

After the talk there were some videos of games running in one of the breakout rooms.

You can find out more on the Amcog website

There is also a youtube channel with currently 40 videos - please subscribe.

RISCOSbits
Andy started with an overview of FAST systems - bringing together Raspberry Pi 4 and SATA. It can support multiple drives, including optical drives. Uses ADFS rather than SCSI. Arguably fastest RISC OS machine currently available - Chris Hall has some benchmarks. About half the speed of RAMFS and limited by filecore not disc speed.

FAST has 4K resolution, is upgradeable, has hardware acceleration and multiple Linux options. It currently needs it own custom ROM (so do not install the default version).

Potential for dual screens in future. Cheaper to buy IO board than create an ITX board for RISC OS.

Recent May ROM update added 'big discs' (up to 2TB), CDFSsoftATAPI in ROM, HATs nearly working and OTA firmware updates.

Working on getting the wifi HAT working. Also working on HATX support, onboard Real time clock and Tan control, NVMe support, better eMMC support, supporting other boards, and a FAST mini.

Also looking to develop a dual boot system (Chimera) which could have dual/triple SSD.

ECOS will run Linux in a Window on RISC OS. Will also run Linux apps (ie Firefox) in a window.

MultiPi FAST will be a mulit-head FAST system. Could run other boards.

Future plans include, moving all Pi 4 systems to EMMC and adding access to linux apps in RISC OS

RISCOSbits website

Drag N Drop
Chris gave us a step by step tutorial on writing a MUG themed RISC OS application (written in BBC BASIC) using the WIMP. If you know BASIC, this was really easy to follow with lots of explanation on how to write a full desktop application.

The finished application and the slides and some other goodies are all free to download here.

Most of the code is based on the Applications tutorial book (special MUG show offer of 15 pounds). There will be a CD version of the 4 books released in 28th July. The new issue of the magazine will also be out in July.

Drag N Drop website

ROD/RComp
Andrew had been ill recently so talk more impromptu than originally planned. Talk done on a PineBook Pro.

First item was a new software announcement (still pre-release). Caching is expected on other OS and CacheFS helps when porting across software. Speeds up file loading even before file read into cache. Demo showed some impressive speed gains on Iris and viewing a PDF file.

Iris was shown next including the new landscape printing and Print preview. Still hoping for a 2023 general release. TIB will be reviewing the May release shortly...

PinBoard 2.02 is now available to download. Pinning folders from network drives more resilient. Andrew also showed how to set transparency on the Icon Bar.

TCP/IP stack has been out for a while. Work continues on wifi and network manager which Andrew was able to demo.

R-Comp website and ROD website

ROOL
Steve and Rob were representing ROOL.

The first was RISC OS 5.30. Still some platforms not ready for release - around 71% done. 9 ports in total. New user guide 96% done. Musical proof-reader needed for Maestro. Over 50 bugs squashed including 3 'CVE' security issues. Release includes 4 bounties. RC3 available to test, RC4 imminent. RC5 in August with hopeful September release for 5.30.

7 current bounties under way (USB, Git client, FS partitions, Toolbox, TCP/IP. Paint, PRM updates)

4 bounties still collecting.

Bounties good at dealing with small items but several big items will need a different approach. Big ideas ROOL think should be prioritised are:
1. Fresh visual theme (modern feel for new users)
2. Running RISC OS on new platforms (current approach does not scale). Idea to have a HAL hypervisor.
3. New content in PRMs
4. Multi-core support
5. VFP legacy support for older machines
6. Rewrite Window Manager and BASIC in C so easier to maintain.
7. C++ developer tools
8. Bluetooth stack

ROOL are figuring out how these might be made possible.

ROOL website

MUG 2023 Show report

nytrex (12:03 2/7/2023)
  arenaman (21:43 4/7/2023)
    Elesar (07:01 6/7/2023)
      arenaman (17:11 6/7/2023)
        Elesar (10:53 7/7/2023)
          arenaman (14:37 7/7/2023)

Alan Robertson

Message #125459, posted by nytrex at 12:03, 2/7/2023

Member
Posts: 117

I agree. It was a really good show.

My highlights of the show were:

The work gone into the RDSP sound engine looks (and sounds) impressive. Once a GUI front end is available, I think it will become a very nice tool for audio users and makers to play with. Reminds me a bit of 4klang / 64klang sound synthesizer available on the Amiga and PC.

The FAST systems look really nice. The significant speed increase for accessing drives over SATA brings huge benefits to RISC OS developers who compile code, and users whose software require lots of disc access (such as Iris and RISC OSM).

A brand new piece of disc caching utility titled CaccheFS from the stable of RISC OS Developments. It seems like it brings a huge increase in speed by automatically caching used files from disc into memory. I do hope this can become integrated into RISC OS by default in the future.

The Pinboard 2 saw a minor update that brings it up to 2.02. It provides lots of new features that really do make the desktop experience better.

Those were by fave bits.
Big thanks to all those who took the time or organize, present, host and of course all who attended the show. A great day out (in).

[ Log in to reply ]

Michael Stubbs

Message #125460, posted by arenaman at 21:43, 4/7/2023, in reply to message #125459

Member
Posts: 114

1. Fresh visual theme (modern feel for new users)

Personally, I'd say forget this until we have a 64bit, multi-threaded OS with modern coding tools and nothing of consequence requiring porting over. Also, please keep the option to use the Acorn icon set as we have now.

Post-Acorn, it's apparent that it's not been Sir Jony Ive working his magic on icons and logos. Hence I still use the Acorn icon set on RISC OS 5. Slightly dated (although the Acorn logo isn't) but it looks GOOD. I personally find RISC OS to be refreshingly understated and calm, and that's said as someone with a lot of admiration for the Mac OS GUI. Windows and Linux just can't seem to quite get it right, for some reason. Hence I think caution is required here. We already have something extremely good.

With Vantage being, let's say, a private project these days, we don't have software capable of rendering graphics to modern standards anyway so such a GUI refresh would be done using alternative platforms where you can use free, modern graphics tools.

2. Running RISC OS on new platforms (current approach does not scale). Idea to have a HAL hypervisor.

Would this keep RISC OS nice and compact and able to take advantage of the ARM architecture still? I don't understand this side of things.

3. New content in PRMs
4. Multi-core support
5. VFP legacy support for older machines
6. Rewrite Window Manager and BASIC in C so easier to maintain.
7. C++ developer tools
8. Bluetooth stack

The rest all sound good to me.

I'd add continued and stronger support for Python and the popular additional libraries, plus making it easier for open source software to be ported over (I'm thinking Inkscape and MuseScore, to name but two).

[ Log in to reply ]

Robert Sprowson

Message #125462, posted by Elesar at 07:01, 6/7/2023, in reply to message #125460

Member
Posts: 45

1. Fresh visual theme (modern feel for new users)

Personally, I'd say forget this until we have a 64bit, multi-threaded OS with modern coding tools and nothing of consequence requiring porting over.

Almost uniquely in this situation, a theme design doesn't need an army of ARM code monkeys to progress it as the other priorities you've mentioned do. A graphic designer could be thought of as a parallel activity who could deliver this for the benefit of existing RISC OS and a future 64b multi-core multi-threaded whizzy whatever.

Also, please keep the option to use the Acorn icon set as we have now.
Post-Acorn, it's apparent that it's not been Sir Jony Ive working his magic on icons and logos.

The idea is to add another theme to the existing selection, so it'd still be possible to choose one of the others or go back to the Acorn set if preferred. By their very nature visual themes are a personal choice but it's probably time for a new lick of paint to join the Morris4 (1992-1998), Ursula (1998-2002), and Iyonix (2003-) sets. We don't even make use of alpha transparency yet, even though the OS has supported it for years.

With Vantage being, let's say, a private project these days, we don't have software capable of rendering graphics to modern standards anyway so such a GUI refresh would be done using alternative platforms where you can use free, modern graphics tools.

You don't have to, and if hiring a graphics designer from the Mac world probably wouldn't want to, try to design the theme on RISC OS. I'd expect the master copies to be done in some vector format like SVG or Adobe Illustrator, then converted at the last step into sprites. That way, if in the future the Wimp can render vector icons we already have the bulk of the design done, but can make use of the output today as sprites.

2. Running RISC OS on new platforms (current approach does not scale). Idea to have a HAL hypervisor.

Would this keep RISC OS nice and compact and able to take advantage of the ARM architecture still? I don't understand this side of things.

The hypervisor presents the underlying system as a more generic one that RISC OS knows about. So RISC OS would be the same compact shape as now, and still written for or running on ARM, the bit the hypervisor is doing is taking away the hassle of forever chasing datasheets for the next latest and greatest platform we'd like to run on: do it once, run anywhere.

(Opinions are my own, not those of ROOL).

[ Log in to reply ]

Michael Stubbs

Message #125464, posted by arenaman at 17:11, 6/7/2023, in reply to message #125462

Member
Posts: 114

Also, please keep the option to use the Acorn icon set as we have now.
Post-Acorn, it's apparent that it's not been Sir Jony Ive working his magic on icons and logos.
The idea is to add another theme to the existing selection, so it'd still be possible to choose one of the others or go back to the Acorn set if preferred. By their very nature visual themes are a personal choice but it's probably time for a new lick of paint to join the Morris4 (1992-199, Ursula (1998-2002), and Iyonix (2003-) sets. We don't even make use of alpha transparency yet, even though the OS has supported it for years.

If it's to be any good at all, not the usual fare, then that will cost a pretty penny. That's my main point; the money is best spent elsewhere. Get the engine running properly before painting the car I guess I'm saying.

I suppose technically you could find a talented graphics artist with a knowledge of RISC OS, who is willing to do all this for free or almost-free, but it seems unlikely.

With Vantage being, let's say, a private project these days, we don't have software capable of rendering graphics to modern standards anyway so such a GUI refresh would be done using alternative platforms where you can use free, modern graphics tools.
You don't have to, and if hiring a graphics designer from the Mac world probably wouldn't want to, try to design the theme on RISC OS. I'd expect the master copies to be done in some vector format like SVG or Adobe Illustrator, then converted at the last step into sprites. That way, if in the future the Wimp can render vector icons we already have the bulk of the design done, but can make use of the output today as sprites.

I know we don't have to, but it doesn't sell the platform not to. However in the present situation, it is undeniable that the results will be far, far better if designed and rendered in Adobe or Serif software, unfortunately, so I wouldn't argue about doing it elsewhere if it were happening now. It's just a shame that there was the tech on RISC OS, and now there isn't.

If I were a programming wiz, I'd be working on porting a couple of things over that would solve the graphics problem. Sadly I'm not, though.

2. Running RISC OS on new platforms (current approach does not scale). Idea to have a HAL hypervisor.

Would this keep RISC OS nice and compact and able to take advantage of the ARM architecture still? I don't understand this side of things.
The hypervisor presents the underlying system as a more generic one that RISC OS knows about. So RISC OS would be the same compact shape as now, and still written for or running on ARM, the bit the hypervisor is doing is taking away the hassle of forever chasing datasheets for the next latest and greatest platform we'd like to run on: do it once, run anywhere.

(Opinions are my own, not those of ROOL).

Well, that sounds eminently sensible.

[ Log in to reply ]

Robert Sprowson

Message #125468, posted by Elesar at 10:53, 7/7/2023, in reply to message #125464

Member
Posts: 45

If it's to be any good at all, not the usual fare, then that will cost a pretty penny.

Oh absolutely - there's about 400 icons, so if a designer spent 2h on each one that's 800h = 100 working days = 6 months full time.

That's my main point; the money is best spent elsewhere.

Ah, but that presupposes that the pot of money is fixed size. In reality someone interested in supporting RISC OS may have no attachment to some of the more technical slanted goals. For example: why would I sponsor a Bluetooth implementation? I don't have any Bluetooth peripherals.

...whether drawing on RISC OS or Mac/Windows/Linux...

It's just a shame that there was the tech on RISC OS, and now there isn't.

True, though I'd expect the pool of talent of graphics designers will be bigger if the net is cast without the requirement to do the work on RISC OS. The designer can use their own (preferred) tools, just as if I hire a carpenter to make me a nice chair I'm not going to force them to use my blunt chisel that I use for cutting plasterboard.

[ Log in to reply ]

Michael Stubbs

Message #125469, posted by arenaman at 14:37, 7/7/2023, in reply to message #125468

Member
Posts: 114

If it's to be any good at all, not the usual fare, then that will cost a pretty penny.
Oh absolutely - there's about 400 icons, so if a designer spent 2h on each one that's 800h = 100 working days = 6 months full time.

That's my main point; the money is best spent elsewhere.
Ah, but that presupposes that the pot of money is fixed size. In reality someone interested in supporting RISC OS may have no attachment to some of the more technical slanted goals. For example: why would I sponsor a Bluetooth implementation? I don't have any Bluetooth peripherals.

Whether an individual or a group, I would be impressed if the money for six months of ace graphics work was ponied up. Never say never, I guess.

...whether drawing on RISC OS or Mac/Windows/Linux...

It's just a shame that there was the tech on RISC OS, and now there isn't.
True, though I'd expect the pool of talent of graphics designers will be bigger if the net is cast without the requirement to do the work on RISC OS. The designer can use their own (preferred) tools, just as if I hire a carpenter to make me a nice chair I'm not going to force them to use my blunt chisel that I use for cutting plasterboard.

It definitely would be.

I'd still like to see proper graphics tools on RISC OS again, though, personally. I reckon with ArtWorks seemingly not developed any more (except for porting to new hardware) and Vantage not even available, it's going to be a case of porting open source stuff over, if anyone wants to do it, that is.

[ Log in to reply ]