SCIOPTA ?

SCIOPTA is a very new kernel based on (a carbon-copy) OSE Epsilon, see

Download an evaluation copy and enjoy! OSE Epsilon is proven since 1986 and worth the money.

Good luck!

/Filip

Dear Filip,

I was struggling long if I ever should reply, but now I do:

First: I do work for Sciopta Systems, but tried to keep my postings private though I often could have made more advertisement.

Do you still work for ENEA ?

Right. But do you know anything about Epsilon ? So far I don't think:

1) Though the OSE API is quiete old, any Epsilon kernel is written from scratch. 2) The Coldfire kernel is from 2001 and AFAIK it does not yet support the 5282 ! => See data sheet.

Totally wrong. Sciopta was developed by the team (and I am the main designer of the kernel) that did the maintenance for the OSE Epsilon products from 1998 till 2003. Sciopta was designed from scratch, but of course I used some principles that were appealing to me:

pre-emptiv kernel : Is this an OSE invention ? No ! priority based scheduling: Is this an OSE invention ? No ! fixed buffer-size memory pools: Is this an OSE invention ? No ! direct message passing: Is this an OSE invention ? No ! timer processes: Is this an OSE invention ? No ! assembly written: Is this an OSE invention ? No ! transparent IPC over CPU boundaries: Is this an OSE invention ? No !

But these are the only 7 similarities between Sciopta and OSE Epsilon (the only OSE kernel I know, and I know it by heart, I even gave support to customers 1/2 year after leaving OSE ! ). But these can be found all or parts of them in most operating systems.

No Epsilon code went into Sciopta ! (Maybe you post some greek-code to prove me wrong :-)

I wonder, how you could compare Sciopta with Epsilon, without even knowing Sciopta. AFAIK you did not see any line of code of Sciopta. Do you even know any source line of Epsilon ? I guess you did load our Technical Reference from

?

See above comment. I don't know to which kernel you are referring, but I myself wrote one OSE Epsilon kernel (the V850) and improved the ARM kernel. A colleague wrote the Coldfire kernel. And only because the API is quiete old ( not necessarily good) does not make a new port stable automatically. But actually, the team that wrote Sciopta, did write stable Epsilon kernels. Now what does this mean ?

One last comment: We never wanted to compete with OSE (Epsilon,Treat or Delta) since we target a different price-level. But it is nice to see that OSE fears us :-)

--
42Bastian
Do not email to bastian42@yahoo.com, it's a spam-only account :-)
Use @epost.de instead !

Yes I do and I agree it is sometimes a thin line between advertisment and just wanting to provide information.

I was referring to the architecture of the kernel more than the actual implementation.

Ok, I was surprised that the reply for my statment that it was a carbon-copy came so late. Good that you clear some things up. You must however agree on that reading the description of Sciopta on your home page makes people with an OSE background recognize themselves and wonder. Maybe you should try to rewrite some parts of the description and invent a SCIOPTA terminology.

Ha ha, I will not... I do not think you had anything to win from that. It is better to re-write.

No, I have not seen a single line Epsilon nor Sciopta and I have just downloaded the Brochure since the Technical Reference was by mail only. As I mentioned above I more refer to the architecture, i.e. Scheduling, Memory managment (pools), Message based, Error management (centralized) etc. No, it is true OSE did not invent these concepts but why construct another kernel so similar?

This was info from my side (with a little bit of advertisment terminology)...

Well, welcome to the jungle...

/Filip

NOTE: The above mentioned opinions do not necessary reflect the opionions of my employer!