[ECOS] UPnP and Ecos

[ECOS] UPnP and Ecos

[jean-luc.philippe]

Hi all,

I'm evaluating the port of UPnP SDK from Intel on an embedding Linux like Ecos
probably for a ARM7 Samsung target.

Does anyone have an idea of the difficulty to port this ?

Are the different protocol available on Ecos ?
BOOTP, DHCP-Client, AUTO-IP, ARP for Addressing on IP.

Thank's a lot for help

Jean-Luc

Re: [ECOS] UPnP and Ecos

[Jonathan Larmour]

jean-luc.philippe@mgeups.com wrote:
> 
> Hi all,
> 
> I'm evaluating the port of UPnP SDK from Intel on an embedding Linux like Ecos
> probably for a ARM7 Samsung target.

(Note: eCos is not Linux, although many of the APIs are compatible). If
you're going to use existing code to start the port, make sure the licence
of that code permits it! It may have restrictions on use, and/or
restrictions on redistribution.
 
> Does anyone have an idea of the difficulty to port this ?
> 
> Are the different protocol available on Ecos ?
> BOOTP, DHCP-Client, AUTO-IP, ARP for Addressing on IP.

All those except AUTO-IP.

Jifl
-- 
Red Hat, Rustat House, Clifton Road, Cambridge, UK. Tel: +44 (1223) 271062
Maybe this world is another planet's Hell -Aldous Huxley || Opinions==mine

Re: [ECOS] UPnP and Ecos

[Lewin A.R.W. Edwards]

>I'm evaluating the port of UPnP SDK from Intel on an embedding Linux like Ecos
>probably for a ARM7 Samsung target.

It's quite complicated; eCos has MOST of the underlying requirements but 
the port isn't simple. If you can afford to do so, WAIT and work on 
polishing other parts of your product. Microsoft is working on a 
"standalone" UPnP implementation. They keep promising and re-promising and 
pushing the date further into the future, but it is coming eventually (we 
will not consider attempting to implement anything UPnP until this SDK 
addition is released).

Wait-and-see also lets you avoid implementing it if it all implodes before 
market-time. A couple of major players pulled out, and in other fields 
there are upcoming and/or recently-announced competing standards that will 
probably be better supported for some time. Unless you REALLY need a UPnP 
logo on your box in Q4'01 or Q1'02 (I hope you have more selling points 
than that.... :) a good strategy to pursue is enthusiastic talk only, since 
that's what everyone else is doing.

=== Lewin A.R.W. Edwards (Embedded Engineer)
Work: http://www.digi-frame.com/
Personal: http://www.zws.com/ and http://www.larwe.com/

"... a man who is endowed with real qualities of leadership will be tempted 
to refrain from taking part in political life; because [...] the situation 
does not call for a man who has a capacity for constructive statesmanship 
but rather for a man who is capable of bargaining for the favour of the 
majority. Thus the situation will appeal to small minds and will attract 
them accordingly."

Re: [ECOS] UPnP and Ecos

[Lewin A.R.W. Edwards]

> > BOOTP, DHCP-Client, AUTO-IP, ARP for Addressing on IP.
>
>All those except AUTO-IP.

AutoIP is simple and definitely implementable within eCos (though I think 
of little value in adding to the base OS). It essentially consists of 
picking an IP address and pinging it to see if someone else owns it. If no, 
then you squat on that address.

=== Lewin A.R.W. Edwards (Embedded Engineer)
Work: http://www.digi-frame.com/
Personal: http://www.zws.com/ and http://www.larwe.com/

"... a man who is endowed with real qualities of leadership will be tempted 
to refrain from taking part in political life; because [...] the situation 
does not call for a man who has a capacity for constructive statesmanship 
but rather for a man who is capable of bargaining for the favour of the 
majority. Thus the situation will appeal to small minds and will attract 
them accordingly."

Re: [ECOS] UPnP and Ecos

[Jonathan Larmour]

"Lewin A.R.W. Edwards" wrote:
> 
> >I'm evaluating the port of UPnP SDK from Intel on an embedding Linux like Ecos
> >probably for a ARM7 Samsung target.
> 
> It's quite complicated; eCos has MOST of the underlying requirements but
> the port isn't simple. If you can afford to do so, WAIT and work on
> polishing other parts of your product. Microsoft is working on a
> "standalone" UPnP implementation. They keep promising and re-promising and
> pushing the date further into the future, but it is coming eventually (we
> will not consider attempting to implement anything UPnP until this SDK
> addition is released).

Is this "coming eventually" like the Microsoft port of ActiveX to Linux?
:-)

And would it be free/open source/royalty free? 

Jifl
-- 
Red Hat, Rustat House, Clifton Road, Cambridge, UK. Tel: +44 (1223) 271062
Maybe this world is another planet's Hell -Aldous Huxley || Opinions==mine

Re: [ECOS] UPnP and Ecos

[Lewin A.R.W. Edwards]

> > polishing other parts of your product. Microsoft is working on a
> > "standalone" UPnP implementation. They keep promising and re-promising and
>
>Is this "coming eventually" like the Microsoft port of ActiveX to Linux?
>:-)

LOL. Let me put it like this: Every teleconference and in-person meeting of 
our working group within UPnP, if a MS guy is present I poke him for it, 
and I bitch and complain at every opportunity. I've also spoken a couple of 
times to the actual guy at MS who would be responsible for it (whether that 
means he'd be coding it personally or not, I don't know). It's always 
"coming soon". Well, our UPnP-supporting product is "coming soon" too.

Recently someone announced on one of the UPnP lists that they have 
developed a standalone UPnP implementation tested on WinME, so my thought 
is that MS would be best to acquire that and rebadge it (I haven't 
inspected the original poster's product, it may be that their definition of 
"standalone" is nothing close to mine. My definition is "compiles and runs 
on any system that has standard BSD socket calls").

>And would it be free/open source/royalty free?

I haven't seen any licensing for this part, I guess it would be covered 
under the same license as the current beta SDK (a product I bet never goes 
out of beta).

UPnP has lost a fair amount of momentum, and frankly most of the 
interaction models and proposals are utopian nonsense. I won't say it will 
never happen, but it is a certainty that today's average consumer doesn't 
want even a tiny fraction of UPnP's complexity. And the money for appliance 
sales is in average consumers, not gadget mavens.

=== Lewin A.R.W. Edwards (Embedded Engineer)
Work: http://www.digi-frame.com/
Personal: http://www.zws.com/ and http://www.larwe.com/

"... a man who is endowed with real qualities of leadership will be tempted 
to refrain from taking part in political life; because [...] the situation 
does not call for a man who has a capacity for constructive statesmanship 
but rather for a man who is capable of bargaining for the favour of the 
majority. Thus the situation will appeal to small minds and will attract 
them accordingly."

RE: [ECOS] UPnP and Ecos

[Dan Conti]

Don't you see the correlation between people taking a wait-and-see attitude
and UPnP losing momentum? It's a chicken and egg argument, you dont want to
develop UPnP support until microsoft has desktop support, but why would
microsoft want to provide desktop support for something that device
companies aren't committed to? Whether or not it is apparent now, UPnP is a
Good Thing, and will soon be a necessary thing.

-Dan

> -----Original Message-----
> From: ecos-discuss-owner@sources.redhat.com
> [ mailto:ecos-discuss-owner@sources.redhat.com]On Behalf Of Lewin A.R.W.
> Edwards
> Sent: Thursday, April 12, 2001 9:10 AM
> To: Jonathan Larmour
> Cc: jean-luc.philippe@mgeups.com; ecos-discuss@sources.redhat.com
> Subject: Re: [ECOS] UPnP and Ecos
>
>
>
> > > polishing other parts of your product. Microsoft is working on a
> > > "standalone" UPnP implementation. They keep promising and
> re-promising and
> >
> >Is this "coming eventually" like the Microsoft port of ActiveX to Linux?
> >:-)
>
> LOL. Let me put it like this: Every teleconference and in-person
> meeting of
> our working group within UPnP, if a MS guy is present I poke him for it,
> and I bitch and complain at every opportunity. I've also spoken a
> couple of
> times to the actual guy at MS who would be responsible for it
> (whether that
> means he'd be coding it personally or not, I don't know). It's always
> "coming soon". Well, our UPnP-supporting product is "coming soon" too.
>
> Recently someone announced on one of the UPnP lists that they have
> developed a standalone UPnP implementation tested on WinME, so my thought
> is that MS would be best to acquire that and rebadge it (I haven't
> inspected the original poster's product, it may be that their
> definition of
> "standalone" is nothing close to mine. My definition is "compiles
> and runs
> on any system that has standard BSD socket calls").
>
> >And would it be free/open source/royalty free?
>
> I haven't seen any licensing for this part, I guess it would be covered
> under the same license as the current beta SDK (a product I bet
> never goes
> out of beta).
>
> UPnP has lost a fair amount of momentum, and frankly most of the
> interaction models and proposals are utopian nonsense. I won't
> say it will
> never happen, but it is a certainty that today's average consumer doesn't
> want even a tiny fraction of UPnP's complexity. And the money for
> appliance
> sales is in average consumers, not gadget mavens.
>
> === Lewin A.R.W. Edwards (Embedded Engineer)
> Work: http://www.digi-frame.com/
> Personal: http://www.zws.com/ and http://www.larwe.com/
>
> "... a man who is endowed with real qualities of leadership will
> be tempted
> to refrain from taking part in political life; because [...] the
> situation
> does not call for a man who has a capacity for constructive statesmanship
> but rather for a man who is capable of bargaining for the favour of the
> majority. Thus the situation will appeal to small minds and will attract
> them accordingly."
>

RE: [ECOS] UPnP and Ecos

[Lewin A.R.W. Edwards]

>Don't you see the correlation between people taking a wait-and-see attitude
>and UPnP losing momentum?

First: Why should I care? _I_ am not trying to sell UPnP. Microsoft is. 
UPnP exists for political reasons, as I'm sure you're aware.

>companies aren't committed to? Whether or not it is apparent now, UPnP is a
>Good Thing, and will soon be a necessary thing.

Desktop support? I am talking about an SDK that does not rely on all sorts 
of miscellaneous OS support. I am talking about an embedded-friendly SDK 
that does not assume Windows CE on the appliance!

UPnP isn't even an "also-ran" at this point, it is a "maybe-ran" and it 
will not be necessary for any device that doesn't seek Windows logo 
licensing in the foreseeable future. It will be at least 2-3 years before 
there is a significant population of products containing UPnP.

For our particular product category, given all the _REAL_ feedback we have 
from _REAL_ users (not B.S. marketing studies from people paid to "prove" a 
particular result) UPnP is a distraction. It will never be a core feature. 
Users are not advanced enough to grasp, let alone use, the utopian ideal of 
UPnP. If MS wants to undertake a long re-education process, then good luck 
- call me back when it's over and people are actually willing to pay for 
the technology.

Further comments off-list please!

=== Lewin A.R.W. Edwards (Embedded Engineer)
Work: http://www.digi-frame.com/
Personal: http://www.zws.com/ and http://www.larwe.com/

"... a man who is endowed with real qualities of leadership will be tempted 
to refrain from taking part in political life; because [...] the situation 
does not call for a man who has a capacity for constructive statesmanship 
but rather for a man who is capable of bargaining for the favour of the 
majority. Thus the situation will appeal to small minds and will attract 
them accordingly."