"Don't buy that, wait for our version."

Hyperion never change.

Quote:

You mean the driver Costel has never finished because he thinks Aeon needs to provide him more technical information?
Yeah, THAT I wasn't mistaken on.
And Spectre660 already made sure I was aware of Rene's work.

The question is, how is Costel's work "better", if he can't gather the info he needs himself?
Even I have access to NXP technical support (and have had a relationship with Freescale and Motorola before them).
Its not hard to get technical info from the Soc vendor.

@Jim

It is not Costel's work.
He states that Hyperion are using an outside developer as well.
His assessment of the two drivers is based on how/if they handle getting the correct MAC address .
Of course if the X5000 production boards have a "Thing" with the MAC addresses then his basis for the assessment could be incorrect. No reason to go to war. At some point both driver versions may be available to X5000 users .


[ Edited by Spectre660 10.09.2018 - 16:11 ]

> What we can count on, is our developers listen to us, and we said a
> long time ago that the core used in the P1022 wasn't desirable.

I'm not sure they listened to us. It's absolutely plausible they came to the same conclusion completely on their own :-)

Quote:

Yeah, its not been an issue I really want to wade in on as the functionality can be provided by a card.
But now with issues using some NIC cards, the availability of a driver for the on-board component is very useful.

I should avoid making provocative statements.

We have the driver we need for our community.

Quote:

OK...my hubris was on display there.
A bit like assuming a connection because the Amiga community and my own endeavors started with the same processor family (or cpu vendor).


[ Edited by Jim 11.09.2018 - 00:40 ]

> You mean the driver Costel has never finished because he thinks Aeon
> needs to provide him more technical information?

I can't remember such statement by Costel. Do you have a link?

> how is Costel's work "better", if he can't gather the info he needs himself?

It surely can't be better if it isn't finished and he decided to stop any work on it.

Quote:

My post #113 this thread. I believe the contents of those 2 posts spawned the resulting dialog on AW as well.

Source

#6

>>> You mean the driver Costel has never finished because he thinks Aeon
>>> needs to provide him more technical information?

>> I can't remember such statement by Costel. Do you have a link?

> My post #113 this thread.

I followed the links therein but can't find any statement by Costel to the effect that he didn't finish his Ethernet driver(s) "because he thinks Aeon needs to provide him more technical information". Maybe you can quote the part?

Quote:


Quote:

The link above to cgutjahr's post certainly applies to this, but I trust your translation over my google translation to english above.

paragraph beginning with "The tabor"

#6

>>>>> You mean the driver Costel has never finished because he thinks Aeon
>>>>> needs to provide him more technical information?

>>>> I can't remember such statement by Costel. Do you have a link?

>>> My post #113 this thread.

>> I followed the links therein but can't find any statement by Costel to the effect
>> that he didn't finish his Ethernet driver(s) "because he thinks Aeon needs to
>> provide him more technical information". Maybe you can quote the part?

Quote:

Nothing indicates that this interrupt issue is referring to any Ethernet driver, and even if it was, this interrupt issue can't be the reason Costel didn't finish his own P1022 Ethernet driver as Costel's driver has been abandoned by him and isn't supposed to be part of OS4 anyway, but the driver by another author will be:

Quote:

So if any Ethernet driver for Tabor is affected by said interrupt issue and needs further information by A-Eon, it cannot be Costel's driver.

So my take on this sorry affair hasn't changed since a year and a half ago. What I'm guessing happened, beginning quite some time ago:

1) 2006: Most OS4 developers leave after the court settlement and Hyperion's debts mount and its dirty laundry is first aired
2) Much later: A-Eon want uboot and OS4 ported to X1000
3) Hyperion either don't want to or can't do this (i.e. have lost the developers who worked on relevant parts or even own them)
4) A-Eon bodge OS4 onto X1000 without Hyperion's help, hence terrible loading times
5) X1000 flops harder than a sealion falling off the Empire State Building, mainly due to costing more than a small 2nd hand family car
6) Cheaper X5000 comes along and A-Eon find themselves in same situation
7) Network drivers require information from uboot (at least MAC address, apparently) so A-Eon can't continue
8) Hyperion pretend they can continue but can't either, and their own driver halts and/or they hold it back so they can (barely) justify a paid update to OS4 as soon as they can start trading again
9) A-Eon won't wait so contact Olsen1 and Olsen2 to do working drivers by using a different method to extract MAC address
10) Four copies of the same driver, and everyone is having great fun.

Am I close?

@KennyR

Way off.
X1000 uses CFE not U-Boot .
X1000 sold all machines produced .
X5000 U-Boot Maintainer still on board.
2 versions of 2 Drivers .

Quote:

I think you meant 2009?
settlement

#6

Sorry for my part in the obfuscation of Hyperion's issue over "interupts".

It appears that the issue is one of deriving MAC addresses.
And I shouldn't have drawn Costel into it.

No point in driving the OS4 fanatics onto a warpath.

Mark figured out a solution for the X5000 under MorphOS and shared it.
Not something that should cause controversy.

> 1) 2006: [...] court settlement

See comment #138.

> 2) [...] A-Eon want uboot [...] ported to X1000

See comment #137.

> 4) A-Eon bodge OS4 onto X1000 without Hyperion's help

Implausible. Doesn't make sense.

> 5) X1000 flops

See comment #137. I'm sure A-Eon would and could have sold some more X1000, but the availability/pricing of the PA6T really got too bad.

> 7) Network drivers require information from uboot (at least MAC address,
> apparently) so A-Eon can't continue

Implausible, as U-Boot for the X5000 is maintained by A-Eon (see also comment #137).

> 8) Hyperion pretend they can continue

Costel says that Hyperion is ready and it's A-Eon's turn now (interrupt issue, maybe board redesign if parts aren't available anymore, board production).

> [...] as soon as they can start trading again

http://www.amiga-news.de/en/news/AN-2018-08-00011-EN.html

> 9) A-Eon [...] contact Olsen1 and Olsen2 to do working drivers by using
> a different method to extract MAC address

The MAC address issue is specific to the X5000, so only 1 Olsen required for this.

> 10) Four copies of the same driver

See comment #137.

> Am I close?

2/10 without objection from my side :-)

> Sorry for my part in the obfuscation of Hyperion's issue over "interupts".
> It appears that the issue is one of deriving MAC addresses.
> And I shouldn't have drawn Costel into it.

As far as I understand, the interrupt issue is specific to A1222/Tabor (but not necessarily to any Ethernet driver for A1222/Tabor), whereas the MAC address issue is specific to X5000/Cyrus. And none of these issues is specific to any of Costel's drivers, which are abandoned anyway.

Quote:

Bingo, unrelated issues, Costel is irrelevant, all just likely to stir up OS4 fanatics like a hornet's nest.
So, that quote starts starts with "sorry".

And what doesn't get enough attention is that another piece of software started under MorphOS is released for use by the OS4 base.

Quote:

At least it is being used ethically, A-Eonkit paying Bigfoot. Unlike in previous times where it is just outright pirated and claimed as original (see fake MUI4).

Quote:

They weren't built until they were ordered, so of course they did. They couldn't have not done so. It's more like there were 70 to 280 (depending on who you ask) ever made that made them a flop.

Quote:

Appears there are four.

1) Costel's version (not working)
2) Version Hyperion paid for from someone else which Costel is now talking about
3) A-Eon's first version
4) A-Eon's second version (Bigfoot)

> another piece of software started under MorphOS is released
> for use by the OS4 base.

...while it's still unreleased for the MorphOS base ;-)

Quote:

This does not exist.

In August 2015, A-EON Technology Ltd contracted Hyperion/Mincea to complete the X5000 network driver. Approaching three years later, we had not received any drivers. Due to the delay with this development and our eagerness to deliver a solution for the customers, an interim solution was found. Mark was commissioned to write the driver for OS4 and MorphOS, which he promptly did and presented us with an alpha version then a subsequent beta version. We released the beta version to our internal Amiga Developer beta test team who validated it ready for public release.

Technically we have paid for another driver from Hyperion which I presume is still being developed, maybe by another developer now on behalf of Hyperion? The most important thing is the end-users (our customers) have a working, high quality driver and do not have to wait any longer.



[ Edited by amigakit 11.09.2018 - 17:57 ]

> They weren't built until they were ordered

I can't remember anyone telling he ordered at AmigaKit and had to wait months for his X1000 until the number of orders matched batch size.

> there were 70 to 280 (depending on who you ask) ever made

Trevor said 1st batch alone was 250 boards, 100 of which went to beta testers. And there were several further batches of unknown (probably smaller) size. Maybe you only have clueless people to ask :-P

>> 2 versions of 2 Drivers

> Appears there are four.
> 1) Costel's version (not working)
> 2) Version Hyperion paid for from someone else which Costel is now talking about
> 3) A-Eon's first version
> 4) A-Eon's second version (Bigfoot)

What's "A-Eon's first version"? I can't remember I read anything on that.
Including Costel's abandoned set of abandoned drivers it's 3 sets of 2 drivers each. And btw, the A-Eon set of drivers is from bigfoot (X5000/Cyrus) and rwo (A1222/Tabor).

>> 2) Version Hyperion paid for from someone else [...]

> Technically we have paid for another driver from Hyperion
> which I presume is still being developed

As I'm sure you read, Costel says Hyperion's commissioned drivers are ready.

Quote:

Another?

Quote:

Source

That reads as "2" more. What am I missing here?

#6

Yeah, I'm not sure I'm clear on this either.

Is the driver Aeon is waiting on, the driver they originally contract with Mincea for, or the driver which Mincea now claims is ready (that would appear to have been created by a subcontractor), or for the purpose of fulfilling the contract does the later subsume/replace the earlier?

Assuming that when Mincea's driver development was discontinued, a sub was hired to fulfill the contract, this is the clearest understanding I can come to from this jumbled description of events.

That the latest driver (again, supposedly ready) is a substitute for the earlier project.

And that regardless of Hyperion's position, Aeon wanted to get a driver in the hands of their beta testers NOW and paid for a second (which could, in a weird way, be considered the third).