@olegil

Quick, someone port the driver for this: http://www.display-solution.com/en/products/chips/elite_silicon_E2868_E3868.html so we can have some extra USB ports if we ever have too many ethernets

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@olegil

"...so we can have some extra USB ports if we ever have too many ethernets"

Nice find.


After quick peek to our (at least in one design) tablet HW block diagram .... we use UART_0 to connect Z3745 to wifi chip.

3G/LTE module uses miniPCIe slot, but only via USB2.0.


The following is almost related to topic, :
About the reasons why CBM went down:
http://www.bigbookofamigahardware.com/bboah/product.aspx?id=23
AA3000+ "Revision 0 of this system first booted successfully in February of 1991."
Then: "the new management had kind of gone on the warpath against all current projects, trying to ensure that nothing we had been working on would see the light of day"

The rest is history.
To me it seems the management tried to kill Amiga and in the process they killed CBM!

I hope we today are more open minded than how CBM management was.


((Another thing... I would not mind if we continue from what R&D guys were doing around 1991 etc...
we would never catch up mainstream (not realistic anyways?) but we would just/anyway have fun niche in our parallell world.))

...anyone has any info of when this block diagram has been drawn?
http://www.bigbookofamigahardware.com/bboah/media/download_photos/0bc3926f-1ab8-4a5d-9c47-5cb11770a140.jpg
AA+ computer with PowerPC and PCI ... it must be drawn after CBM had died?

Last edited by KimmoK on 18-May-2015 at 07:49 AM.
Last edited by KimmoK on 18-May-2015 at 07:42 AM.
Last edited by KimmoK on 18-May-2015 at 07:28 AM.
Last edited by KimmoK on 18-May-2015 at 07:28 AM.
Last edited by KimmoK on 18-May-2015 at 07:22 AM.
Last edited by KimmoK on 18-May-2015 at 07:22 AM.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@KimmoK

Awww, not this AA+ thing again. There are PCIe FPGA eval boards out there, just get to work

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.

@KimmoK

Quote:

...anyone has any info of when this block diagram has been drawn?

For all its relevance and insight it could have been drawn yesterday. By you.

Seriously, there is nothing interesting in that simple scribble. It doesn't even explain how the accessory port interfaces with the CPU to MC path.

_________________
All the way, with 68k

@KimmoK

Wait, it just hit me.
If we can somehow reverse the polarity on the phase transducers before the ion flow enters the compensation matrix, then maybe, just maybe, we can modulate the positron frequency to interleave the accessory port signal with the CPU's memory access.

Of course, we'll need to send out a synchronisation pulse so that the AA+ can intercept the DRAM access from the CPU, injecting custom data bits into the signal so it works with the MC interpolation filter, thus creating an advanced SIMD array.

_________________
All the way, with 68k

@agami

>By you.

It was from this page:
http://www.bigbookofamigahardware.com/bboah/product.aspx?id=23

I believe it might be some David Haynie's memo that has A3000+ written on top of it.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@KimmoK

I don't doubt that it is from Dave Haynie's notebook.
What I was implying is that it holds no true insight into anything. It could just as well be drawn up by someone without all that electronics engineering knowledge.

_________________
All the way, with 68k

@Damocles

When you say: "Notebook? That's just so 90's..."

you meant 1790, 1890 or 1990? Please be more specific!


See ya! / A la prochaine!

M.

@Hillbillylitre

Quote:

I reckon arteffect, imagefx, photogenics, tornado3d, adpro, real3d, dpaint and all that need to be emulated on an Amiga computer too when running OS4.

As some of these (from my mind ArtEffect, Photogenics, ImageFx) work even on AmigaOne (Petunia), I don´t think real Amiga/OS4 will have problems with these. DPaint works under OS4 (on Amiga), but there are problems with some menus (eg. colour picker). There is easy rule - if it works on Amiga with RTG hardware, it will highly probably work on AmigaOS4 or MorphOS.

Quote:

So my conclusion is that Amiga OS4 is completely useless on an Amiga computer.

It seems you have suitable hardware, why not try yourself?

@Megamania

It's one of those previous centuries from the last millennium. Can't be arsed to look which one it was, I rather pay attention to the current and future trends of computing and operating systems.

_________________
Dammy

@agami

Quote:

But to create a leading OS, AmigaOS has better underpinnings to achieve this in a more elegant way than the other operating systems.

I think this is absurdity up on stilts, but, of course, you have a right to your opinion.

Quote:

They are only major hurdles because of resource constraints. But if you had the millions required to hire the right team, they would solve those issues much quicker than implementing the forward thinking stuff.

Throwing millions of {currency of your choice} at software development projects has never produced anything much better than mediocre. OS X might be the best it can do. You usually end up with something like Microsoft Windows. Historically the good stuff has been produced by a few people with good ideas and a lot of passion, and not much of a budget.

_________________
"Unix is supposed to fix that." -- Jay Miner

@agami

Quote:

agami wrote: @megol Quote: AFAIK there are no research that points to this being true. There are a lot of systems that are considered user friendly _but_ user friendly doesn't mean fully intuitive! Just as one have to learn driving a car or using the TV remote one have to learn the user interface. The world is replete with examples of this. In order for AmigaOS to become a leading OS it would not place demands on the user to learn new ways of doing things, i.e. The OS will adapt to the user. That is the essence of user friendliness and to pull that off one needs lots of complicated routines.

To communicate one have to learn a language. Thinking it is possible to do it without learning a language is just a failure to understand the topic. Having an adaptive user interface doesn't change that and is a well studied area (IMHO more research is still needed).

Look into: Smalltalk and Squeak, GEOS (adaptive user interface) and OpenDoc (user configurable applications from components).

Quote:

Quote: Really? Rich APIs? Just use some of those that are already available. I never implied that all of them would need to be developed from scratch. But in order for AmigaOS to be a leading OS it would need these. UX stands for User Experience. Quote: Do you work in PR? I do have a degree in marketing. Not sure why you're asking.

Because you aren't making technical sense and like to combine words into a mess. Typical PR.

Quote:

Quote: And here you show yourself completely clueless. AOS 4 isn't better suited than many other existing solutions, in fact it is extremely ill-suited in any measurable category. The only thing it got is nostalgia - and that is mostly make believe as all important subsystems are rewritten/heavily changed. First, I never talk about things I'm not fully clued into. Second, whilst many parts of the OS have been rewritten, they haven't been heavily changed. Third, think of it as first semester in acting class. The first thing the instructors will do is have you un-learn anything you think you may have learned about acting. And once stripped of allusions the instructor can start to teach acting. The other operating systems are too far down the wrong path. If you ask people who have worked on an OS for the past 10-15 years if they would have done things differently if they knew what they knew now, the answer will overwhelmingly be yes. But the AmigaOS is exactly at the point where you would start to "teach it acting". That is only one of the reasons it is more suitable.

They are? Look at Multics, KeyKOS, EROS, Microsoft Singularity, Nemesis, K42, AS/400, Go! (obscure component based system that can be used as an OS), Oberon (and its relative Bluebottle OS) or even QNX.

What do you think you can add to just this small selection of innovative operating system designs?

AOS is a traditional file-based, monolithic unprotected system optimized for small systems. There are no advantages using it as a starting point.

Quote:

Quote: You just don't get it. "Minor hurdles"? I guess you routinely solve the halting problem too? It's my specialty.

Not sure you understand that it is a unsolvable problem - just like adding those features to AOS and still being able to call it AOS. On technical grounds at least.

Quote:

In order for AmigaOS to be a leading OS it requires leading features. In most cases this would mean features the other operating systems don't posses. They haven't even started to think of what those are as they have no pressure to do so. It's like Windows Mobile OS; It was moving along slowly and then BOOM! Apple comes out with the iPhone and a completely new smartphone UX. And it took Google, Blackberry (RIM), and Microsoft several years just to catch up. That new UX is harder to do than implement something that has been implemented all over the place. So yeah, in the scheme of things those are minor hurdles.

So you think that everything having to be removed is an advantage compared with traditional systems. Why not instead base your system on things that work and already can do most things you want?
And why do you think your "UX" requires an innovative core system?

@damocles

Quote:

damocles wrote: @Megamania It's one of those previous centuries from the last millennium. Can't be arsed to look which one it was, I rather pay attention to the current and future trends of computing and operating systems.

Like what? Most things I see aren't innovative nor generally useful.
We are striving towards shallowness and magic thinking in both computer interfaces and cultural attitudes IMNSHO. Perhaps we'll see the inevitable bounceback soon or perhaps it will accelerate towards a singularity of ignorance. ;)

@olegil

Quote:

olegil wrote: @cdimauro I must be getting old, as I don't want to run multiple DVIs, GbEs, USB2s and at least 2 serial ports off of a USB3. I would actually much rather prefer to have some of this crap directly on the laptop.

De gustibus. I prefer to have as less as cables possible.
Quote:

HOWEVER (and this is fairly important) with for example the T10x2 series, you choose what feature you want to bring out to pins.

That's not about me, but the direction that the market is getting.

Consider that even desktop platforms soon will go for "cable-less". Intel will present Skylake in a few months, where this is one key concept.
Quote:

So while it CAN have 5 ports of ethernet, it can also have 2 ports of SATA and 4 ports of PCIe. It cannot, however have all of these at the same time. There's 2 RGMII and 8 serdeseseses, where 4 PCIe + 1 SATA + 2 ethernet (or alternatively 3+2+2) is doable at the same time. One PCIe would be x4, the rest x1.

Which is just sufficient for a desktop, but too much for a laptop, where usually the GPU is integrated on chip and only a SATA is needed, with some USB port for external connectivity.

What happens to such unused ports? Is it possible to completely power-gate them, so they draw near zero power?
Quote:

And do you have any links to SoCs with built-in wifi?

Currently some big ARM vendors provide them: Qualcomm and MediaTek for sure. Samsung, Apple and nVidia use external chip currently. I don't know about minor players.

Intel came late, but she's working on it.

Quote:

olegil wrote: @olegil Quick, someone port the driver for this: http://www.display-solution.com/en/products/chips/elite_silicon_E2868_E3868.html so we can have some extra USB ports if we ever have too many ethernets

But that'll draw more power.

@megol

Quote:

Like what? Most things I see aren't innovative nor generally useful.

Today, it's about packing in a lot of punch in a relatively low cost mobile devices. Tomorrow, it's about VR that makes today's attempts of VR look primitive. Might be wrong, but I doubt it.

_________________
Dammy

@cdimauro

Quote:

Which is just sufficient for a desktop, but too much for a laptop, where usually the GPU is integrated on chip and only a SATA is needed, with some USB port for external connectivity. What happens to such unused ports? Is it possible to completely power-gate them, so they draw near zero power?

The T10x2 SOC don't have audio or video integrated so the configuration of the serdes lanes would be as follows:

GPU 4 lanes
2X SATA 2 Lanes
Ethernet 1 lane
Audio 1 lane

That's all 8 lanes used and no unused ports that need to switched off.

@Rob

>The T10x2 SOC don't have audio or video integrated

T10x2 and T10x4 have simple 2D GFX integrated (like P1022 had).
T10x2 and T10x4 does not have i2s that could be directly used for audio (unlike P1022)

>so the configuration of the serdes lanes would be as follows:
...
>Ethernet 1 lane

I might have misunderstood, from RM, but it might be that Ethernet does not block any serdes.

>Audio 1 lane

Using PCIe v2.0 lane for audio is a overkill. But yes, that is one option.
If RadeonHD would be used as GPU, it's integrated audio should be used.

>That's all 8 lanes used and no unused ports that need to switched off.

There is a lot of (other) SoC ports to take care of, to avoid unnecessary current leak in mobile solution.

_________________
- KimmoK
// For freedom, for honor, for AMIGA
//
// Thing that I should find more time for: CC64 - 64bit Community Computer?

@megol

Quote:

To communicate one have to learn a language. Thinking it is possible to do it without learning a language is just a failure to understand the topic. Having an adaptive user interface doesn't change that and is a well studied area (IMHO more research is still needed).

By the time most people reach adulthood they already know multiple languages so why should they need to learn another?
A shift in computing and particularly in UX started about 5 years ago where less and less people need to know how to operate functions on a device, as the device would be able to apply skeuomorphic concepts to meet the users existing knowledge.
Example: Turning the page of an e-book on a touch device is the same as turning a page in a physical book.

Quote:

Look into: Smalltalk and Squeak, GEOS (adaptive user interface) and OpenDoc (user configurable applications from components).

1. The topic is how to make AmigaOS a leading operating system, not how to make some other obscure OS into a leading operating system.
2. Whilst each of those operating systems you mentioned are doing cool things in certain areas, it is my view that the AmigaOS sits well at the crossroads of everything I would like to accomplish.
3. I care more about AmigaOS than any other OS.

Quote:

Because you aren't making technical sense...

I am deliberately avoiding getting too technical in this public forum; I'm not about to give away more than 10 years of R&D. Following the lead from @bison I used high level descriptions for core features.
Personally, I am plenty technical. I used to be a coder/developer in the '80s and '90s, I have managed development teams from 2001 to 2007, and now I am part owner of two development practices in India.

Quote:

Not sure you understand that it is a unsolvable problem - just like adding those features to AOS and still being able to call it AOS.

No problem is unsolvable.
In this scenario where I am turning AmigaOS into a leading operating system, it is still called AmigaOS by the virtue that it is based on AmigaOS code, it supports the execution of classic 68k Amiga applications and games, and allows the nostalgic few to configure the UI to mimic the old Workbench.

Quote:

So you think that everything having to be removed is an advantage compared with traditional systems. Why not instead base your system on things that work and already can do most things you want? And why do you think your "UX" requires an innovative core system?

Because I disagree with how most things have been implemented.
Because the things that "work" already don't do the things I want them to, and to extend them in that direction would result in an inelegant solution.
Because I am also working on a new hardware platform that the more "mature" operating systems are not designed for.

Last edited by agami on 19-May-2015 at 04:11 AM.
Last edited by agami on 19-May-2015 at 04:04 AM.
Last edited by agami on 19-May-2015 at 04:02 AM.
Last edited by agami on 19-May-2015 at 04:01 AM.

_________________
All the way, with 68k

@bison

Quote:

I think this is absurdity up on stilts, but, of course, you have a right to your opinion.

Just like none of the currently available flat screen technologies have captured the display performance profile of the classic CRT, none of the mainstream operating systems with leading features have been able to capture some of the basic performance aspects of AmigaOS.

Yes, they have integrated (well or poorly) many new features over the past 20 years which AmigaOS does not currently posses. But in order to be a leading OS you don't have to repeat everything these others have done.

Another example is the African nations which never developed a copper telephony infrastructure and are now implementing largely a glass infrastructure for their network backbone. They are skipping the copper step most developed nations built out over the last 50-60 years.

Quote:

Throwing millions of {currency of your choice} at software development projects has never produced anything much better than mediocre.

Not entirely true.
Yes, small passionate teams have repeatedly done amazing things. But that's not doing it for us in AmigaOS. So either we don't have enough of them or they're not passionate enough.
Bottom line, to implement some of the more modern features AmigaOS lacks, it's required to hire some talent. They are not going to work for free. Passionate developers often work for below-market rates, but they also often get stock options as an incentive. If I wanted to get Carl Sassenrath to be a part time advisor on the project, he will say yes a lot easier if he is compensated.
It's not about throwing $Millions at it and hoping for the best, if I want some of these things sooner rather than much later, I need to hire some more people.
By the time you pay 10 developers $100k + benefits per year each, you are looking at $2M+ for the next 24 months. Add license fees, 10 x X1000's and you're looking at another $1M (just kidding ).
I will take a small dedicated team of top-notch talent over a large generic and mediocre team any day. But even the small teams require funds.

Last edited by agami on 19-May-2015 at 04:55 AM.
Last edited by agami on 19-May-2015 at 04:54 AM.
Last edited by agami on 19-May-2015 at 04:52 AM.
Last edited by agami on 19-May-2015 at 04:52 AM.
Last edited by agami on 19-May-2015 at 04:51 AM.

_________________
All the way, with 68k

@cdimauro

Dang, accidentally closed the window while researching for my reply.

Anyway, personally I don't want an Intel GPU, I want an Intel GPU with an AMD GPU in my next laptop (then at some point I'll retire this one and use the components in the waifu's laptop which has a slower CPU but a dedicated GPU. Unfortunately that's a Acer which requires the equivalent of "brain surgery through the urethra" for CPU replacements).

On a totally different note, I'm looking at one 4700hq + R9 M265X + DVD vs one 4510u + R7 M260 and the former is cheaper than the latter. Same manufacturer, other components are virtually identical. Can anyone shed some light into why newer models with half the performance are more expensive?

Edit: Oh, and if sites like GPUBoss would kindly include Intel HD graphics in their database so it would be frelling possible to frelling compare two frelling laptops I would be a happy camper. Thankfully notebookcheck acknowledges Intel as a GPU manufacturer

Last edited by olegil on 19-May-2015 at 09:58 AM.

_________________
This weeks pet peeve:
Using "voltage" instead of "potential", which leads to inventing new words like "amperage" instead of "current" (I, measured in A) or possible "charge" (amperehours, Ah or Coulomb, C). Sometimes I don't even know what people mean.