An Open Letter to Dave Haynie
  • Priest of the Order of the Butterfly
    Priest of the Order of the Butterfly
    boot_wb
    Posts: 874 from 2007/4/9
    From: Kingston upon ...
    Thanks Dave,

    A sane voice amongst the screaming banshees of Moobunny.

    Happy 50th btw.
    www.hullchimneyservices.co.uk

    UI: Powerbook 5,6 (1.67GHz, 128MB VRam): OS3.1, OSX 10.5.8
    HTPC: Mac Mini G4 (1,5GHz, 64MB VRam): OS3.1 (ZVNC)
    Audiophile: Efika 5200b (SB Audigy): OS3.1 (VNC + Virtual Monitor)

    Windows free since 2011!
  • »02.06.11 - 09:27
    Profile Visit Website
  • Just looking around
    humantarget
    Posts: 10 from 2011/6/2
    Andreas_Wolf,
    Quote:

    Commodore -> Escom (1995):
    dubious


    Don't think so. Yeah, some of the stuff Hembach did was dubious... but keep in mind, he was only Trustee of Commodore GmbH. He initially sold the Commodore trademark to ESCOM, even though he didn't have the authority:
    http://articles.philly.com/1995-03-09/business/25700886_1_commodore-s-amiga-commodore-international-sale

    The final sale was for all of Commodore's intellectual properties, and certain inventories. Hembach had nothing to do with it.
    http://articles.philly.com/1995-04-22/business/25686725_1_amiga-commodore-products-commodore-international

    In retrospect, it's a shame Dell didn't win... at least they didn't kill themselves the way ESCOM did. Maybe something Amiga would have survived, maybe not.
  • »02.06.11 - 09:28
    Profile
  • MorphOS Developer
    jacadcaps
    Posts: 2974 from 2003/3/5
    From: Canada
    Quote:

    I mean, what happened to Intuition, or whatever the equivalent in MorphOS ought to be?


    It works pretty much the same as original.

    Quote:

    AmigaOS fundamentally doesn't lock up when overloaded, doesn't stutter in audio even when overloaded, doesn't get a jittery mouse, etc.


    Can't agree with you on that. If a graphics operation takes a long time under a layers lock and is being performed on the input.device's context, this is what you'll get as well. Imagine doing a large, complex cpu draw leading to loads of accesses to chip ram - ouch.

    Quote:

    I might expect those web browser windows to redraw slowly if the system were overloaded in CPU, but Intuition should always run at a higher priority, along with the whole UI event chain.


    This is totally unrelated to intuition, but simply a limitation of our current graphics subsystem (CGX/Layers/Radeon driver). What you're seeing is the effect of CPU swapping bitmaps from system memory to graphics memory and back to system memory on each draw, because the system is running at a high-res screen on a machine with mere 32MB of graphics memory. Surely something we do need to address in future releases, but not in the area of intuition like you imagined.
  • »02.06.11 - 09:32
    Profile Visit Website
  • Priest of the Order of the Butterfly
    Priest of the Order of the Butterfly
    boot_wb
    Posts: 874 from 2007/4/9
    From: Kingston upon ...
    Quote:

    There is no possible good answer for why it's so slow, compared to something like MacOS.


    Erm, mac-mini here running MacOS and MorphOS dual-boot. MorphOS kicks MacOS arse on every single task as far as I can see - asynchronicity, redrawing, video rendering... not sure on audio, but I betcha reggae wins. Lower ram usage... video graphics usage not sure about, but hey - triple buffering and multiple screens takes its toll.

    There are some glitches - target windows frozen when drag/drop copying between directories (which can be changed in ambient config files), screen-blanker/dpms activation causes momentary stuttering on MPlayer, hardware limitations of the mac-mini ATA bus (cd drive locks up ata hdd) - but far less irritating without that damn MacOS colour wheel...

    When was the last time you played around with/demoed MorphOS?
    www.hullchimneyservices.co.uk

    UI: Powerbook 5,6 (1.67GHz, 128MB VRam): OS3.1, OSX 10.5.8
    HTPC: Mac Mini G4 (1,5GHz, 64MB VRam): OS3.1 (ZVNC)
    Audiophile: Efika 5200b (SB Audigy): OS3.1 (VNC + Virtual Monitor)

    Windows free since 2011!
  • »02.06.11 - 09:36
    Profile Visit Website
  • MorphOS Developer
    Piru
    Posts: 576 from 2003/2/24
    From: finland, the l...
    humantarget,
    Quote:

    Can anyone explain this: http://www.youtube.com/watch?v=NvmWIqqRR-g

    IIRC this particular slowness was due to running out of VRAM and the resulting swapping. Think of a windows system running out of physical memory and you get the idea (except that in this case the VRAM is being swapped to system memory. It isn't as slow as HDD but still slow). OS X does have better VRAM management. Obviously this is one area we're attempting to improve. This can be worked around somewhat by reducing the eye candy and in extreme cases (16 or 32MB VRAM) reducing the resolution and or screen depth.

    Assuming you don't run into this situation MorphOS quite easily runs circles around OS X in most areas. I am fairly certain there are numerous users who can back this up.
  • »02.06.11 - 09:38
    Profile
  • Just looking around
    humantarget
    Posts: 10 from 2011/6/2
    itix,
    Quote:

    Dave Haynie being PCB layout guy at some ex-toy company doesnt make him any special.


    PCB layout guy? Get your facts straight, Sherlock. I did do PC layout on the Nomadio Sensor and React. And everything other hardware task as well, and some software just for fun. And yeah, they were toys to you, I guess... much as the Amiga was a toy in its day to those who didn't know what they were talking about.

    Just to set the record straight, the Sensor was the first all digital R/C controller. It offered bidirectional communications, first of its kind for R/C racing. Four of the top five professional R/C racers used our radio. Yeah, I didn't know there were professionals either... but if you can drive an R/C car well enough, you can make well in excess of $100,000 a year doing that thing.

    The main reason was low latency.. a trained musician can detect about a 1ms difference in tempo... that's the goal that MIDI set out to meet, way back when. A typical R/C controller for serious hobby vehicles (eg, you'll be spending several hundred dollars US on the car, more for servos, etc) usually had a latency of about 20ms. We did it in about 5ms, typical. When you're racing a 1:10 scale car at an actual 60-70mph, this apparently makes a difference (I was no RCer).

    A version of this same technology was commissioned though a US government funded project to deliver a super low cost robot for bomb hunting and destruction. We put about 3,000 of these systems into Iraq, a few hundred more went to Special Forces... no idea what they did with them. This was a $5,000 robot system that could carry up to 10lbs of C4 explosive, with a camera system to let a soldier drive up to about 2,000 feet away. It was mainly used to identify and blow up IEDs along the roadside in Iraq. Without this, they would have sent some private... some of whom would have been killed. Not a thing that's touching a few million lives like the Amiga, but then again, the people who didn't get killed because they had a robot to send instead are presumably pretty happy they had this "toy" available.

    There are certainly other robots... our current radio, developed with funding from Homeland Security, is being used on the PackBot (http://en.wikipedia.org/wiki/PackBot) and other larger robots. A PackBot costs about $250,000 or so. In typical government behavior, the cheap robots were pretty successful, so they're not making any more. There's more being done with these expensive robots now, but only so much, because they're too expensive to just put in the back of every HumVee... they're for EOD squads only. And their default radio isn't good enough.. thus the testing with our new one. And yeah, I designed the custom hardware parts of that (does modified 802.11n, MIMO, on 5MHz and 10MHz channels, at any frequency between 400MHz and 2000MHz, with a secondary radio at 2200-2600MHz or 4500-5800MHz). It runs a proprietary mesh network... each radio is an end-point and a router. I had to teach myself hardcore RF design to build this... nothing I needed before Nomadio. I did the architecture, design, software, FPGA, and yeah, the PCB.
  • »02.06.11 - 09:49
    Profile
  • Priest of the Order of the Butterfly
    Priest of the Order of the Butterfly
    boot_wb
    Posts: 874 from 2007/4/9
    From: Kingston upon ...
    Quote:

    What you're seeing is the effect of CPU swapping bitmaps from system memory to graphics memory and back to system memory on each draw, because the system is running at a high-res screen on a machine with mere 32MB of graphics memory.


    64MB Vram according to the video.

    Surprised they managed to run out of display memory so quickly, even at 1920x1200 res 32-bit triple buffered. I have never run into anything like that using multiple 32-bit triple-buffered screens at 1366x720.
    www.hullchimneyservices.co.uk

    UI: Powerbook 5,6 (1.67GHz, 128MB VRam): OS3.1, OSX 10.5.8
    HTPC: Mac Mini G4 (1,5GHz, 64MB VRam): OS3.1 (ZVNC)
    Audiophile: Efika 5200b (SB Audigy): OS3.1 (VNC + Virtual Monitor)

    Windows free since 2011!
  • »02.06.11 - 09:54
    Profile Visit Website
  • Just looking around
    humantarget
    Posts: 10 from 2011/6/2
    jacadcaps,
    Quote:

    Can't agree with you on that. If a graphics operation takes a long time under a layers lock and is being performed on the input.device's context, this is what you'll get as well. Imagine doing a large, complex cpu draw leading to loads of accesses to chip ram - ouch.


    The problem is certainly that it's being done on the input device's context, rather than in the application proper. That at least explains why the UI is bogged down... input devices would be running at the same priority as Intuition... which is why you'd never want to do something that expensive this way.

    Doesn't explain the audio glitching... that should be at a higher priority. Audio needs to be realtime, web browser overhead, not so much. I guess you have to know the code... maybe there's a good reason. It's just fairly shocking to see a MacOS display outperform... well... much of anything.
  • »02.06.11 - 09:58
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Zylesea
    Posts: 2053 from 2003/6/4
    humantarget,
    Quote:

    I don't know much about MorphOS, honestly. I know it was sold as a replacement for AmigaOS. I'm convinced it's not, but not in a good way.


    Hi Dave and welcome to the Morphzone. Tone may be a bit rough here sometimes, but usually nobody gets really biten. One suggestion though: If you don't know much about MorphOS then try it out yourself! Grab a compatible computer (Efika, Pegasos 1/2, Mac mini G4, emac G4 1.25/1.42 GHz or any Powermac G4 || complete list at http://www.morphos.de/hardware.html) go to www.morphos.de, download the free iso of MorphOS 2.7, burn that to cd and boot the computer off with that cd and make your own picture. It is really easy.
    For a starter - how MorphOS looks from the Amiga user POV - you may read this very brief introduction: http://via.i-networx.de/wim.htm
    --
    http://via.bckrs.de

    Whenever you're sad just remember the world is 4.543 billion years old and you somehow managed to exist at the same time as David Bowie.
    ...and Matthias , my friend - RIP
  • »02.06.11 - 10:04
    Profile Visit Website
  • MorphOS Developer
    jacadcaps
    Posts: 2974 from 2003/3/5
    From: Canada
    Quote:

    It's just fairly shocking to see a MacOS display outperform... well... much of anything.


    Well, the youtube video shows the currently worst possible scenario. Normally the whole thing works a lot faster than on OSX. For instance, you actually can play h264 720p on a Mac Mini 1.5GHz on MorphOS while it's too slow on OSX on the same machine.

    [ Edited by jacadcaps 02.06.2011 - 11:06 ]
  • »02.06.11 - 10:06
    Profile Visit Website
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Zylesea
    Posts: 2053 from 2003/6/4
    These massive slow downs are with an enhanced didplay screen, then the swapping has major impact and even with 64 MiB it happens quite soon, but on a non enhanced screen even with 20 opened (OWB) windows all remains nicely fluent. Maybe it would be nice if enhanced display screen automatically switches to non enhanced diplay mode when VRAM gets too low.

    Edit: And if play around with stuff like taht don't forget that owb may start with a saved session which includes not only tabs, but also windows. I.e. if you used OWB with more than plenty windows last time, don't forget to switch off enhenced display *before* you start up owb next time (cough)...

    [ Editiert durch Zylesea 02.06.2011 - 11:29 ]
    --
    http://via.bckrs.de

    Whenever you're sad just remember the world is 4.543 billion years old and you somehow managed to exist at the same time as David Bowie.
    ...and Matthias , my friend - RIP
  • »02.06.11 - 10:21
    Profile Visit Website
  • MorphOS Developer
    jacadcaps
    Posts: 2974 from 2003/3/5
    From: Canada
    @Zylesea

    The real solution is to use the AGP to do the bitmap swapping instead of the CPU. The screens cannot be switched once opened - I do however disable multibuffering in low memory conditions.
  • »02.06.11 - 10:30
    Profile Visit Website
  • Moderator
    Kronos
    Posts: 2239 from 2003/2/24
    @boot_wb
    1366x720 = 983,520 pixel
    1920x1200 = 2,304,000 pixel

    Running out of 64MB is dead easy at that res (plenty 1st hand experience).
  • »02.06.11 - 10:44
    Profile
  • MorphOS Developer
    Henes
    Posts: 507 from 2003/6/14
    Quote:


    Quote:


    I might expect those web browser windows to redraw slowly if the system were overloaded in CPU, but Intuition should always run at a higher priority, along with the whole UI event chain.


    This is totally unrelated to intuition, but simply a limitation of our current graphics subsystem (CGX/Layers/Radeon driver). What you're seeing is the effect of CPU swapping bitmaps from system memory to graphics memory and back to system memory on each draw, because the system is running at a high-res screen on a machine with mere 32MB of graphics memory. Surely something we do need to address in future releases, but not in the area of intuition like you imagined.


    Maybe it should also be noted this is an issue the old AmigaOS could not get into... because once it runs out of video ram (well, chip ram), any new bitmap/window/screen allocation will just fail. Even if all the memory is used to store non displayed elements. You have plenty of fast ram and yet you run out of memory...
    Where MorphOS (and CGX/P96 in the old times too) manages to re-use those, even keeping hardware acceleration as long as possible.
  • »02.06.11 - 13:19
    Profile Visit Website
  • MorphOS Developer
    Krashan
    Posts: 1107 from 2003/6/11
    From: Białystok...
    humantarget,
    Quote:

    In fact, I would have loved to see the same demo, on the Mac under MacOS running UAE. Betcha it also embarrasses MorphOS.

    You have to find some real WWW browser for AmigaOS 3.x first. Or you want to compare OWB with IBrowse 2.4? ;-)
  • »02.06.11 - 13:50
    Profile Visit Website
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12079 from 2003/5/22
    From: Germany
    > Phase V was trying really hard to get involved with AT

    Yes, and it seems this attempt even resulted in an AT press release mentioning a co-operation:

    "Thanks to a close co-operation between Amiga Technologies and Phase V, a German turbo board manufacturer, a full range of Power PC boards will also be available for the A1200, A3000 and A4000 series."
    http://www.cucug.org/amiga/aminews/1995/at951111.html
    http://www.amigareport.com/ar320/news10.html
    http://www.amigahistory.co.uk/ppcrelease.html

    > there were definite plans to build [...] a PowerPC port.

    Considering AT in 1995 announced a Power Amiga for 1997 this only seems logical.

    > MorphOS was fragmenting an already weak community by offering an Amiga clone.

    As I see it, MorphOS was offering the only serious (AROS, albeit in existence, wasn't to be considered serious back then) way forward in terms of "AmigaOS" as we knew it. As you said yourself, Amiga Inc. tossed AmigaOS out for Tao's Intent/Elate. The fragmentation happened only later when OS4 was announced in March/April 2001 and more specifically end of 2001 when Hyperion restarted the OS4 project after Amiga Inc's own OS4 project as well as the negotiations to make MorphOS the new OS4 failed.
  • »02.06.11 - 14:23
    Profile
  • MorphOS Developer
    itix
    Posts: 1516 from 2003/2/24
    From: Finland
    Quote:


    I mean, what happened to Intuition, or whatever the equivalent in MorphOS ought to be?



    Dunno but is there anyone using Intuition directly anymore?

    Quote:


    AmigaOS fundamentally doesn't lock up when overloaded, doesn't stutter in audio even when overloaded, doesn't get a jittery mouse, etc.



    It does. Just use Amiga 600 with an internal IDE :-) Anyway, MP3 players on AmigaOS (and MorphOS) can stutter if some higher priority task is using too much CPU time. Obviously because then there isnt enough CPU time for decoding task. You can workaround it by raising priority for decoder but other tasks can still play around with Forbid()/Disable().

    That AmiNetRadio in that video is streaming so it could be just that network buffer is running out. I didnt look at that video carefully enough to check whether it does so.
    1 + 1 = 3 with very large values of 1
  • »02.06.11 - 14:30
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12079 from 2003/5/22
    From: Germany
    > I was the one who actually outed Merlancia as a complete fraud

    Only to those who didn't know that before already ;-)

    > Hyperion took longer to deliver AmigaOS 4.x than it took to write
    > AmigaOS from scratch, plus 1.2 and at least 1.3.

    Hyperion took 2.5 years to (pre-)release a first version of OS4. AmigaOS 1.3 was released in 1988, so I don't think your assessment reflects reality.
  • »02.06.11 - 14:56
    Profile
  • MorphOS Developer
    itix
    Posts: 1516 from 2003/2/24
    From: Finland
    Quote:


    And MorphOS was fragmenting an already weak community by offering an Amiga clone.



    Is it our fault?

    And then -- we are not living in communism where you are offered only one option, one opinion and one community. Users are free to make their decision and use whatever they want to use.

    Particularly I grew tired to old "buy this and that to support true Amiga" sentiment.

    Quote:


    But hey, anything you can cling to I guess -- the AmigaOS and all spinoffs have been largely irrelevant for a good decade now, at least on the scale of computer industry things.



    Amiga really died in 1994 and never came back. It is nice hobby platform, however. Well, at least sometimes :-)

    Quote:


    If you're wondering why I stopped trying to re-create the Amiga and just went on to other projects, that's pretty much it. I did not see any point in another false promise to the Amiga community.



    Once you stop worrying about Amiga only then you can start enjoying again. Ten years ago I had this strange idea Amiga should be revived, restored back to the mainstream. It is silly. Just use it when it is fun and stop using it when it is not fun anymore. No more worrying whether one should buy this and that to keep Amiga companies alive. We didnt worry about Amiga in 1990 -- why should we worry about it now?

    So dont take any promises, dont believe anything, just let it flow.


    [ Edited by itix 02.06.2011 - 17:10 ]
    1 + 1 = 3 with very large values of 1
  • »02.06.11 - 15:00
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12079 from 2003/5/22
    From: Germany
    > some of the stuff Hembach did was dubious... but keep in mind, he was only
    > Trustee of Commodore GmbH. He initially sold the Commodore trademark
    > to ESCOM, even though he didn't have the authority:
    > http://articles.philly.com/1995-03-09/business/25700886_1_commodore-s-amiga-commodore-international-sale
    > The final sale was for all of Commodore's intellectual properties, and certain
    > inventories. Hembach had nothing to do with it.
    > http://articles.philly.com/1995-04-22/business/25686725_1_amiga-commodore-products-commodore-international

    Thanks for the links. Very interesting reads.
  • »02.06.11 - 15:04
    Profile
  • Priest of the Order of the Butterfly
    Priest of the Order of the Butterfly
    Crumb
    Posts: 730 from 2003/2/24
    From: aGaS & CUAZ Al...
    @humantarget
    Quote:

    And MorphOS was fragmenting an already weak community by offering an Amiga clone.


    Amiga Inc ditched AmigaOS3.x for that Elate thingie. OS4 wasn't even planned. There was no future for AmigaOS users but MorphOS.

    Quote:

    I mean, ok, MorphOS moved a bit faster, but Amiga, Inc. and Hyperion took longer to deliver AmigaOS 4.x than it took to write AmigaOS from scratch,


    And MorphOS would have moved even faster if some people hadn't attacked it for no reason.
  • »02.06.11 - 15:05
    Profile Visit Website
  • Caterpillar
    Caterpillar
    HenryCase
    Posts: 39 from 2008/1/2
    Quote:

    As I see it, MorphOS was offering the only serious (AROS, albeit in existence, wasn't to be considered serious back then) way forward in terms of "AmigaOS" as we knew it.


    Of course missed opportunities are easier to see with the benefit of hindsight, but do think what you've highlighted shows we missed out on a truly great opportunity.

    Just imagine if AROS had been taken seriously back then. There was no logical reason it shouldn't have been. Imagine an open-source AmigaOS-like operating system with polish beyond that found in MorphOS ('beyond' as development responsibilities would have been shared between MorphOS and AROS devs).

    Most Amiga fans accept MorphOS is the most polished of the NG Amiga systems, but combining it with the open-source nature of AROS (and with the extra large developer pool) would have made it unstoppable. I doubt Hyperion would have even attempted to compete.

    Of course this is just describing an alternate reality, and we have to live in our own reality, but perhaps we should learn lessons from our past stubbornness.
  • »02.06.11 - 16:42
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    amigadave
    Posts: 2794 from 2006/3/21
    From: Northern Calif...
    Quote:

    I have no first-hand knowledge of ANY of the Phase V code going into MorphOS. The copyright issues certainly still apply -- if anyone studied, say, the Amiga Exec source code, then went and built their own clone, that would still violate copyrights. But I don't know that to be true either, and Ralph says it's not, so I'm more than happy to take him at his word.


    Maybe you can explain why you felt in necessary to bring up this topic again within the last few months by posting your comments regarding the possibility of MorphOS2.x infringing on any Amiga copyrights? What was your motive, or reason(s)?
    MorphOS - The best Next Gen Amiga choice.
  • »02.06.11 - 18:03
    Profile
  • Jim
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Jim
    Posts: 4977 from 2009/1/28
    From: Delaware, USA
    Has Dave always suffered from schizophrenia or is it a recent event?
    "Never attribute to malice what can more readily explained by incompetence"
  • »03.06.11 - 01:56
    Profile