Sorry, but I STRONGLY disagree.

At present Aros is a half finished, half supported OS. The whole idea behind Aros was to find a home one cheap, readily available X86 hardware.

Problem is, without the resources available to the much larger and commited development teams available on more mainstream OS's, Aros is struggling to reach a level of usability that will make it a viable platform.

It would be great to be able to support the diverse array of available hardware, ala Windows or Linux, but this is simply not a realistic point of view.

Now, if, as stated, certain developers prefer to develop drivers - great, but would it not be better to have a COMPLETE driver for a single, cheap and available piece of hardware than to have 3 half working/finishes drivers?

Paolone made a good point about the pace at which hardware, especially x86 hardware, ages and is superceded. Problem is, he pointed out the logic in what I first stated! If only a single, or small set of COMPLETE drivers were worked on, there would be more resources available to update/change them to suit the newer replacement hardware as needed. If driver development resources are spread thinly over a heap of half completed hardware then everytime something new is released, how are the resources going to cover this?

A simple example - If ONLY a driver for a relatively new, cheap and easily obtainable gfx card, lets call it the GF1000 chipset for the sake of argument, was developed, then when the GF1010 was released it would receive a the attention required to complete a full driver in a reasonable amount of time.

On the other hand, if the same development resources were spread between the GF1000, MJ250, Acme35 etc etc etc, then not only do we end up with a bunch of half usable drivers, but by the time drivers are written for all of the NEWER versions of the chipsets (again, half usable drivers) they are all out of date.

Open source has plenty of merits. Lack of a viable business model, and resource management skills are NOT some of them. :(

As for the bounties - good one, I havent laughed so hard in ages. :) Strength of numbers and resource management comes into play again here. So, you are actually suggesting that having dozens of bounties on dozens of hardware is going to produce dozens of complete drivers and keep the Aros x86 platform up to date? :) Now, if there were a FEW heavily supported bounties on a FEW pieces of hardware there would be [a] - more development resources to go between the FEW drivers being developed, and [b] - there would be a much higher proportion of funds to be allotted to each driver bounty. Of course, this assumes that the same number of people contribute. Which model do you think has a the higher chance of success? :)

No one is saying to DROP existing Aros drivers. What I am saying is GOING FORWARD there needs to be some sort of direction or sanity to Aros development... 12 years on and there is still not a single fully supported laptop/netbook that I know of - How about we just continue on like this for another couple of decades? :)

As for forcing developers etc.... Wow, I dont remember saying ANYTHING like that. In case I didnt get my original point forward intelligibly, let me reiterate.

Aros is great, its fun etc etc. It must be, I have 3 machines running it. I would be happy to purchase a custom Aros hardware platform (at the right price), but Im not about to spend money on a pile of hardware that ONE GUY thinks is a good idea, has half finished drivers.... and two days later some OTHER GUY thinks he has a better idea for a pile of hardware with half finished drivers.

Give me ONE fully supported laptop/netbook and ONE fully supported desktop pc and I will take one of each. Until then I see no point in buying a non-officially supported "half system".

If, as I suspect, your point is that "variety is the spice of life" and that Aros should be as "hardware open" as possible - fine, I will continue to use hard supported hardware.... but why would anyone NEED to purchase an Aros specific pc if "everything and the kitchen sink" is half supported anyway?

Just use any old pc.

To show how strongly I, and I am sure others, feel about this...

I, abraXXIous, hereby state that I will pay $1000.00 AUD for a complete Aros system with supported gfx, sound, sata and wifi with WPA. A system like this can be built new in Australia for about $450.00.

There. So if we ever do get some completed drivers, one of you nice fellows can whack together a system and email me - I will purchase it immediately.

Quote:

NOPE. This is exactly in the opposite direction every open source project aims to go. While I do agree that having a common, standardised hardware platform would be good, developers must have the maximum freedom to decide how and where to spend their time, since nobody pays them to do that.

And I don't like single platforms, 'cos platforms in the x86 world get obsolete in no time. We risk - as any company creating a OS - to finish the job when the hardware has become practically useless. To do this right, we'd ask AMD or Intel to reveal their secret projects for 2012 or 2013, and start developing for them TODAY.

I'd never stop to point that with our current menpower AROS is already a little miracle per se, and we - users - should already be happy with what we have. Forcing developers to do what we want is wrong and, first of all, should be at least paid. In a nutshell: should we decide a commom platform? ok, good, let's do it, but then please open bounties or hire developers to support that.

And, then, try to convince people who already have dozens of spare HW parts, that they should forget about them and just buy another computer of your choice to run AROS.

this is the plan

Quote:

No, we don't NEED to sit down and decide anything, the current solution is the best, in that you've got projects like iMica and the rest working towards full AROS support with bounties, and everyone else gets as much hardware support as our limited development resources will allow.

The more people you allow to try out AROS without ponying up cash for new hardware the better, we don't want a situation like AmigaOS and MorphOS where hardware choice stunts the user base's growth.

If you want the fully supported AROS hardware, you get to encourage that through the bounties for iMica, AresOne, EeePC, etc... The only improvement to our driver situation I'd like to see is a compatibility layer for the BSD WiFi stack like Haiku has, in my opinion AROS' use of technologies like Gallium3D and AHI mean that the outlook for drivers on AROS isn't, on the whole, as bad as people often suggest.

I should also state that limiting development in one area of AROS does not automatically mean developer time is spent elsewhere in AROS. Developers have their own areas of interest, let them contribute whatever they wish.

Guys, this is great, BUT we have a real problem with Aros drivers, or rather, lack of them. We NEED, and I repeat, NEED to sit down and decide on the BEST hardware to use that is already supported and then STICK to this as the Aros hardware configuration.

This way, Aros development can proceed with little time wasted on extra drivers, and the drivers for supported hardware can be refined and improved. Also, this will help with sstability and compatibility in the future.

This goes for Aros supported laptops/netbooks.

Aros does not have the development resources available for us to all go in every hardware direction known to man. :)

I myself built an Imica, not because it was the most cost effective solution (but it was actually pretty cost effective) and not because it was easy to find the hardware (other mbs were available over the counter - I had to order the Imica MB), but because it was the first STANDARD suggested for running Aros.

Please pick a system/netbook and stick to it for all our sakes. Aros development resources should not be wasted developing an Ati AND Nvidia driver - pick ONE! sb128 or ac97? Pick ONE! Etc Etc. If there was just ONE MB FULLY supported with gfx, ethernet, sound, wifi etc I would buy it in an instant regardless of the cost.

Sorry for the rant, but I have followed the progress of Aros for a while now and believe if we are ever going to have a complete OS solution then we have to all start pulling in the same direction.

Little update:

the AresOne will be sold viahttp://www.vesalia.de/?lng=e - hope Aros gets even more attention this way. They are free to offer different configurations and have better prices for AmigaForever and other nice stuff then i could offer.

Nevertheless i will open my online-shop (with needful things for aros and AresOne) until my deadline.

If things run nicely in the next half year i will come up with
my concept of a new Amiga-like Computer. Thanks to the contact to AMD i got (for getting Aros to Cebit 2010 - please don't lose it of your mind because i could need some help in january) i have access to their developer pages. My plan is to start Aros Hardware-Project like the one for the Natami or the opensource PPC-board (p2p) based on a reference-system. I have some ideas and hope to get it working. Maybe some of you would like to work on this with me.

Before this i will try to get catweasel supported. This is needed for the hardware-project i have in mind.

Regards,

Pascal (very enthusiastic and dreaming about a better future ; ) )

Quote:

Good news :) I enjoy reading about AROS in AmigaFuture :)

Today i brought a test-pc to Mr. van Wahnem for a review in the next AmigaFuture. Hope he likes what he got ; ).

Is it possible to have a minimum Aros to put in a Bios-chip?
If yes how much space(memory) is needed?

Hi friends,

next weekend i will bring the first AresOne to the person who writes the review for the next issue (82) of AmigaFuture.

The box will cost a little more then previously planed.
This is because i will use other Hardware. For example i kicked the Sempron for the basic system an replaced it with
a Athlon64 X2 4850e AM2(2x 2,5GHZ).

Here are the now fixed specs:
AMD Athlon64 X2 4850e AM2 Tray 2x512kB
MSI K9N Neo-F V2 ATX
Arctic Alpine64 GT Sockel 754/939/AM2
1024MB DDR2 PC6400/800
JCP Midi Tower ATX "Dark Crystal", white or black, 500W
GeForce 7200GS 256MB DDR2
250GB, SATA II
DVD-RW (different vendors)
Realtek RL8139D Ethernet (i hope that hitchhikr at aros-exec can create a driver for the onboard ethernet)
SB Live 5.1
3,5Zoll Diskdrive

The price including A-live 09 will be 299?
After launch of the webshop i offer a "lifetime"-discount (5%) for people who register to it. A pre-order is not needed to get those 5%. This discount is valid for every product i offer and not limited for one order.

I will offer also other components like Sound-/Ethernet- and more-cards : ) which are supported by AROS.

Every time i sell "something" 25% go back to AROS (Bounties, Hosting and so on) When time comes we could decide together where the money goes. I would collect it for one quarter(of the year)otherwise we are talking perhaps about 12? only : ).

Paolone will get some money for every PC i sell since it will be added as free bonus to A-live as donation.

Also the author of Cinnamon Writer will get some money as donation.

Because of Gallium 3d i use now a geforce 7 instead a Ati x300(or above). If some AresOne`s are sold i will spend some money for a Gallium Ati driver.

I think the ones with a square background look far better. TBH the ones with a circle or shield look more like medals than logos.