Live Report From Wakefield

Is anyone reporting live from the Wakefield Show this weekend?
Either on here, via Twitter, Facebook or IRC?

It would be good for us peeps that can’t make it.

A live webcam would be cool….

If I’m not mistaken, Wakefield has now been and gone. Any announcements of interest? I saw a new version of the DDE checked into CVS a week or so ago… was that announced by any chance?

The are some interesting pixs and videos (retro focus) at end of http://stardot.org.uk/forums/viewtopic.php?f=25&t=10736

Martin Würthner announced a new version of ArtWorks, 2.X4. Nearly ready. Has an eraser tool and I can’t quite remember what else.

A few announcements have appeared on comp.sys.acorn.announce but not here. We announced a new ImpEmail supporting CSV and a new RiscOSM.

Being on a stand, I have very little idea what anyone else was demonstrating!

Martin Würthner announced a new version of ArtWorks, 2.X4.

I think it was 2.X3 rather than 2.X4

I have uploaded some pictures from the show and notes from the theatre presentations (with the exception of Martin Wuerthner’s presentation as I was too tired to take notes of a fifth preentation by then). I think Vince has videoed the talks though.

It was an interesting show, Archive 24:1 was available as was volume 3 issue 7 of Drag ‘n’ Drop and there were some interesting talks.

See here.

Your website should get some sort of award for layout. Love it.

Nice report too.

Thanks Chris, wonderful! I missed out on the talks and missed the chance to speak to Sine Nomine due to time constraints – so really appreciated!

Thumbs up to the Drag n Drop stand. Tea and a fire extinguisher. That’s the essentials covered.

Any speed details on the Linux side lf the Titanium and how well it works as a dual RISCOS/Linux machine?

Hi,

I released a new game Xeroid which is available to buy from the PlingStore

Had a really good show. Looking forward to next year’s.

Tony.
www.amcog-games.co.uk

Any speed details on the Linux side

Linux is able to harness both CPU cores, any preferred benchmark to reflect that?
Note that in the test system at the only (non RAM) media is the micro SD card from which it booted, as other media like an SSD would depend on your choice of drive.

The are some interesting pixs and videos (retro focus) at end of http://stardot.org.uk/forums/viewtopic.php?f=25&t=10736

I have uploaded some pictures from the show and notes from the theatre presentations

Thanks for those; they were very interesting :)

Any speed details on the Linux side

Linux is able to harness both CPU cores, any preferred benchmark to reflect that?

I guess that most RISC OS users who are interested in running Linux on the Titanium (or any other machine) would be interested in benchmarks for things which can’t be done (or don’t work very well) under RISC OS:

• Browser benchmarks – http://browserbench.org/
• Audio/video encode/decode (preferably hardware accelerated)
• 3D benchmarks (hardware accelerated, obviously!)

Some I/O performance benchmarks might be nice too, to see how RISC OS’s USB/SATA/network stack holds up against Linux. For most other things I’d expect it to just be a case of “take the RISC OS performance and double it”.

Phoronix benchmarks.

It would be much easier to justify the cost of the Titanium if it was clearer if it could be used as a dual purpose machine (and it would then be really attractive to replace my existing Intel Linux box with something running ARM and RISCOS as well). What does A15 look like compared to Intel for running Debian?

I’d expect it to just be a case of “take the RISC OS performance and double it”.

Doesn’t that assume 100% threading efficiency? I tried the linux NOOBS distro on my Pi2 recently; in a day’s normal use processor usage maxed at 26%, i.e. equivalent to one out of 4 cores running flat out. And single-threading RISC OS has no overhead due to threading. Don’t get me wrong: I’m strongly in favour of multi-threading RISC OS, but I wonder just how much real-world performance boost there’ll be.

I think Vince has videoed the talks though.

I did. Some of Matthew’s talk may be missing, however – the SD card in the camera ran out of space during his slot. I’ve yet to look at it to see how much it got.

Also, I don’t think my current (cheap) laptop has the grunt to get the videos ready for Youtube, so I might have to bite the bullet and set up my ‘new’ PC (which is so new its still unboxed unopened – after getting on for a year since purchase).

(Edited the following day – see strike out. D’oh!)

What does A15 look like compared to Intel for running Debian?

ARM only woke up to the idea of using their processors for something other than a mobile phone relatively recently, so an Intel x86 is likely to be clocked faster than a contemporary age ARM, but ARM will likely win on efficiency (MIPS/watt).

The following benchmark attempts to measure the processing capability, sidestepping issues like which desktop was installed, since unlike RISC OS with its one choice Window Manager module there are all manner of ways to get a desktop – some of which are known to be rather heavyweight for ARM.

   BYTE UNIX Benchmarks (Version 5.1.3)

   System: arm: GNU/Linux
   OS: GNU/Linux -- 4.1.6-01293-g5148636
                    #1 SMP PREEMPT Mon Feb 8 18:58:14 GMT 2016
   Machine: armv7l (unknown)
   Language: en_US.utf8 (charmap="UTF-8", collate="UTF-8")
   CPU 0: ARMv7 Processor rev 2 (v7l) (0.0 bogomips)
   CPU 1: ARMv7 Processor rev 2 (v7l) (0.0 bogomips)

------------------------------------------------------------------------
Benchmark Run: Tue Apr 19 2016 20:43:15 - 21:11:52
2 CPUs in system; running 1 parallel copy of tests

System Benchmarks Index Values           BASELINE       RESULT    INDEX
Dhrystone 2 using register variables     116700.0    7848437.5    672.5
Double-Precision Whetstone                   55.0        683.7    124.3
Execl Throughput                             43.0       1419.5    330.1
File Copy 1024 bufsize 2000 maxblocks      3960.0     139268.5    351.7
File Copy 256 bufsize 500 maxblocks        1655.0      40404.2    244.1
File Copy 4096 bufsize 8000 maxblocks      5800.0     436352.4    752.3
Pipe Throughput                           12440.0     301859.5    242.7
Pipe-based Context Switching               4000.0      56467.6    141.2
Process Creation                            126.0       3052.5    242.3
Shell Scripts (1 concurrent)                 42.4       3107.7    733.0
Shell Scripts (8 concurrent)                  6.0        559.2    931.9
System Call Overhead                      15000.0     561698.3    374.5
                                                               ========
System Benchmarks Index Score                                     353.3

------------------------------------------------------------------------
Benchmark Run: Tue Apr 19 2016 21:11:52 - 21:40:49
2 CPUs in system; running 2 parallel copies of tests

System Benchmarks Index Values           BASELINE       RESULT    INDEX
Dhrystone 2 using register variables     116700.0   15707948.9   1346.0
Double-Precision Whetstone                   55.0       1367.3    248.6
Execl Throughput                             43.0       2715.0    631.4
File Copy 1024 bufsize 2000 maxblocks      3960.0     224862.6    567.8
File Copy 256 bufsize 500 maxblocks        1655.0      63101.5    381.3
File Copy 4096 bufsize 8000 maxblocks      5800.0     644011.8   1110.4
Pipe Throughput                           12440.0     596898.4    479.8
Pipe-based Context Switching               4000.0     110563.3    276.4
Process Creation                            126.0       4513.2    358.2
Shell Scripts (1 concurrent)                 42.4       4281.6   1009.8
Shell Scripts (8 concurrent)                  6.0        552.3    920.5
System Call Overhead                      15000.0    1072127.2    714.8
                                                               ========
System Benchmarks Index Score                                     584.9

= Atom D525 (1,8 GHz dual core)
https://memo-linux.com/byte-unixbench-tester-sous-linux-les-performances-de-votre-serveur-dedie-ou-vps/

Thanks for the figures.

I think Vince has videoed the talks though.

I did. Some of Matthew’s talk may be missing, however – the SD card in the camera ran out of space during his slot. I’ve yet to look at it to see how much it got.

The talks are now available to view on YouTube:

• Andrew Rawnsley
• Rob Sprowson
• Chris Evans
• Matthew Phillips
• Martin Würthner

The amount missing from the end of Matthew’s talk is only a few minutes. There’s a link under the video to Ruth Gunstone’s recording, which includes what I didn’t get.

(Having skipped to the end of Ruth’s video so I could establish what I missed, I failed to notice until now that it’s also missing a few minutes at the beginning! So that balances out, then. Or something.)