All of lore.kernel.org
 help / color / mirror / Atom feed
* netperf in meta-networking
@ 2012-11-30 16:18 Mark Hatle
  2012-11-30 16:34 ` Joe MacDonald
  0 siblings, 1 reply; 6+ messages in thread
From: Mark Hatle @ 2012-11-30 16:18 UTC (permalink / raw)
  To: openembedded-devel; +Cc: Joe MacDonald

I was helping someone with building netperf from meta-networking last night, and 
stumbled on what I think is a bug.

SUMMARY = "A networking benchmarking tool"
DESCRIPTION = "Network performance benchmark including tests for TCP, UDP, 
sockets, ATM and more."
SECTION = "console/network"
HOMEPAGE = "http://www.netperf.org/"
LICENSE = "netperf"
LICENSE_FLAGS = "commercial"

In the above, the LICENSE_FLAGS are set to 'commercial'.  I think this is 
incorrect.  It should be set to 'non-commercial'.

There is a subtle difference between them which is why I think it's a bug...

commercial -- there are some commercial requirements necessary to use this 
recipe in a commercial device... you are responsible for understanding them and 
doing whatever is necessary... (for non-commercial devices you can likely use it...)

non-commercial -- this item is restricted to non-commercial users only.  As in 
the case of netperf, the license says it's only for non-commercial use.

So anyway, my suggestion is to simply change the value of the flag.

--Mark



^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: netperf in meta-networking
  2012-11-30 16:18 netperf in meta-networking Mark Hatle
@ 2012-11-30 16:34 ` Joe MacDonald
  2012-11-30 16:54   ` Joe MacDonald
  2012-11-30 16:55   ` Paul Eggleton
  0 siblings, 2 replies; 6+ messages in thread
From: Joe MacDonald @ 2012-11-30 16:34 UTC (permalink / raw)
  To: Mark Hatle; +Cc: openembedded-devel

[-- Attachment #1: Type: text/plain, Size: 1933 bytes --]

[netperf in meta-networking] On 12.11.30 (Fri 10:18) Mark Hatle wrote:

> I was helping someone with building netperf from meta-networking
> last night, and stumbled on what I think is a bug.
> 
> SUMMARY = "A networking benchmarking tool"
> DESCRIPTION = "Network performance benchmark including tests for
> TCP, UDP, sockets, ATM and more."
> SECTION = "console/network"
> HOMEPAGE = "http://www.netperf.org/"
> LICENSE = "netperf"
> LICENSE_FLAGS = "commercial"
> 
> In the above, the LICENSE_FLAGS are set to 'commercial'.  I think
> this is incorrect.  It should be set to 'non-commercial'.
> 
> There is a subtle difference between them which is why I think it's a bug...
> 
> commercial -- there are some commercial requirements necessary to
> use this recipe in a commercial device... you are responsible for
> understanding them and doing whatever is necessary... (for
> non-commercial devices you can likely use it...)
> 
> non-commercial -- this item is restricted to non-commercial users
> only.  As in the case of netperf, the license says it's only for
> non-commercial use.
> 
> So anyway, my suggestion is to simply change the value of the flag.

Ugh.  I had a quick look around and the first thing I found was this:

http://www.netperf.org/svn/netperf4/trunk/include/netperf.h

So I had a look at the netperf source tree fetched by bitbake during a
build.  The COPYING file says this:

  The enclosed software and documentation includes copyrighted works
  of Hewlett-Packard Co. For as long as you comply with the following
  limitations, you are hereby authorized to (i) use, reproduce, and
  modify the software and documentation, and to (ii) distribute the
  software and documentation, including modifications, for
  non-commercial purposes only.

So I'm inclined to agree, this sounds like the definition of
non-commercial above to me.

-- 
-Joe MacDonald.
:wq

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 205 bytes --]

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: netperf in meta-networking
  2012-11-30 16:34 ` Joe MacDonald
@ 2012-11-30 16:54   ` Joe MacDonald
  2012-11-30 16:55   ` Paul Eggleton
  1 sibling, 0 replies; 6+ messages in thread
From: Joe MacDonald @ 2012-11-30 16:54 UTC (permalink / raw)
  To: Mark Hatle; +Cc: openembedded-devel

[-- Attachment #1: Type: text/plain, Size: 406 bytes --]

[Re: [oe] netperf in meta-networking] On 12.11.30 (Fri 11:34) Joe MacDonald wrote:

> [netperf in meta-networking] On 12.11.30 (Fri 10:18) Mark Hatle wrote:
> 
> > So anyway, my suggestion is to simply change the value of the flag.
[...]
> So I'm inclined to agree, this sounds like the definition of
> non-commercial above to me.

Updated.  Thanks for the heads-up.

-- 
-Joe MacDonald.
:wq

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 205 bytes --]

^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: netperf in meta-networking
  2012-11-30 16:34 ` Joe MacDonald
  2012-11-30 16:54   ` Joe MacDonald
@ 2012-11-30 16:55   ` Paul Eggleton
  2012-11-30 17:11     ` Mark Hatle
  2012-11-30 17:56     ` Joe MacDonald
  1 sibling, 2 replies; 6+ messages in thread
From: Paul Eggleton @ 2012-11-30 16:55 UTC (permalink / raw)
  To: Joe MacDonald; +Cc: openembedded-devel

On Friday 30 November 2012 11:34:11 Joe MacDonald wrote:
> [netperf in meta-networking] On 12.11.30 (Fri 10:18) Mark Hatle wrote:
> > I was helping someone with building netperf from meta-networking
> > last night, and stumbled on what I think is a bug.
> > 
> > SUMMARY = "A networking benchmarking tool"
> > DESCRIPTION = "Network performance benchmark including tests for
> > TCP, UDP, sockets, ATM and more."
> > SECTION = "console/network"
> > HOMEPAGE = "http://www.netperf.org/"
> > LICENSE = "netperf"
> > LICENSE_FLAGS = "commercial"
> > 
> > In the above, the LICENSE_FLAGS are set to 'commercial'.  I think
> > this is incorrect.  It should be set to 'non-commercial'.
> > 
> > There is a subtle difference between them which is why I think it's a
> > bug...
> > 
> > commercial -- there are some commercial requirements necessary to
> > use this recipe in a commercial device... you are responsible for
> > understanding them and doing whatever is necessary... (for
> > non-commercial devices you can likely use it...)
> > 
> > non-commercial -- this item is restricted to non-commercial users
> > only.  As in the case of netperf, the license says it's only for
> > non-commercial use.
> > 
> > So anyway, my suggestion is to simply change the value of the flag.
> 
> Ugh.  I had a quick look around and the first thing I found was this:
> 
> http://www.netperf.org/svn/netperf4/trunk/include/netperf.h

This is netperf4 though which I gather is distinct from the version of netperf 
we are building.

> So I had a look at the netperf source tree fetched by bitbake during a
> build.  The COPYING file says this:
> 
>   The enclosed software and documentation includes copyrighted works
>   of Hewlett-Packard Co. For as long as you comply with the following
>   limitations, you are hereby authorized to (i) use, reproduce, and
>   modify the software and documentation, and to (ii) distribute the
>   software and documentation, including modifications, for
>   non-commercial purposes only.
> 
> So I'm inclined to agree, this sounds like the definition of
> non-commercial above to me.

It's a subtlety - in either case you can't just go ahead and use it for 
commercial use. I wouldn't object to changing it to "non-commercial" though if 
it makes more sense.

Would it be practical for us to move to netperf4 which is GPLv2+ licensed? I'm 
not sure if it is a complete replacement for netperf 2.x and we'd need to 
confirm that first, but at least the license is more reasonable...

Cheers,
Paul

-- 

Paul Eggleton
Intel Open Source Technology Centre



^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: netperf in meta-networking
  2012-11-30 16:55   ` Paul Eggleton
@ 2012-11-30 17:11     ` Mark Hatle
  2012-11-30 17:56     ` Joe MacDonald
  1 sibling, 0 replies; 6+ messages in thread
From: Mark Hatle @ 2012-11-30 17:11 UTC (permalink / raw)
  To: Paul Eggleton; +Cc: openembedded-devel, Joe MacDonald

On 11/30/12 10:55 AM, Paul Eggleton wrote:
> On Friday 30 November 2012 11:34:11 Joe MacDonald wrote:
>> [netperf in meta-networking] On 12.11.30 (Fri 10:18) Mark Hatle wrote:
>>> I was helping someone with building netperf from meta-networking
>>> last night, and stumbled on what I think is a bug.
>>>
>>> SUMMARY = "A networking benchmarking tool"
>>> DESCRIPTION = "Network performance benchmark including tests for
>>> TCP, UDP, sockets, ATM and more."
>>> SECTION = "console/network"
>>> HOMEPAGE = "http://www.netperf.org/"
>>> LICENSE = "netperf"
>>> LICENSE_FLAGS = "commercial"
>>>
>>> In the above, the LICENSE_FLAGS are set to 'commercial'.  I think
>>> this is incorrect.  It should be set to 'non-commercial'.
>>>
>>> There is a subtle difference between them which is why I think it's a
>>> bug...
>>>
>>> commercial -- there are some commercial requirements necessary to
>>> use this recipe in a commercial device... you are responsible for
>>> understanding them and doing whatever is necessary... (for
>>> non-commercial devices you can likely use it...)
>>>
>>> non-commercial -- this item is restricted to non-commercial users
>>> only.  As in the case of netperf, the license says it's only for
>>> non-commercial use.
>>>
>>> So anyway, my suggestion is to simply change the value of the flag.
>>
>> Ugh.  I had a quick look around and the first thing I found was this:
>>
>> http://www.netperf.org/svn/netperf4/trunk/include/netperf.h
>
> This is netperf4 though which I gather is distinct from the version of netperf
> we are building.
>
>> So I had a look at the netperf source tree fetched by bitbake during a
>> build.  The COPYING file says this:
>>
>>    The enclosed software and documentation includes copyrighted works
>>    of Hewlett-Packard Co. For as long as you comply with the following
>>    limitations, you are hereby authorized to (i) use, reproduce, and
>>    modify the software and documentation, and to (ii) distribute the
>>    software and documentation, including modifications, for
>>    non-commercial purposes only.
>>
>> So I'm inclined to agree, this sounds like the definition of
>> non-commercial above to me.
>
> It's a subtlety - in either case you can't just go ahead and use it for
> commercial use. I wouldn't object to changing it to "non-commercial" though if
> it makes more sense.
>
> Would it be practical for us to move to netperf4 which is GPLv2+ licensed? I'm
> not sure if it is a complete replacement for netperf 2.x and we'd need to
> confirm that first, but at least the license is more reasonable...

There are things that netperf4 can not do, that netperf 2.x can.  So I would 
recommend that we add a 'netperf4_<PV>.bb' recipe as well.  This way the user 
can select either or both of the netperf's depending on their requirements.

According to the netperf home page, if you want something "more open source" 
they suggest you use netperf4.

--Mark

> Cheers,
> Paul
>




^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: netperf in meta-networking
  2012-11-30 16:55   ` Paul Eggleton
  2012-11-30 17:11     ` Mark Hatle
@ 2012-11-30 17:56     ` Joe MacDonald
  1 sibling, 0 replies; 6+ messages in thread
From: Joe MacDonald @ 2012-11-30 17:56 UTC (permalink / raw)
  To: Paul Eggleton; +Cc: openembedded-devel

[-- Attachment #1: Type: text/plain, Size: 3181 bytes --]

[Re: [oe] netperf in meta-networking] On 12.11.30 (Fri 16:55) Paul Eggleton wrote:

> On Friday 30 November 2012 11:34:11 Joe MacDonald wrote:
> > [netperf in meta-networking] On 12.11.30 (Fri 10:18) Mark Hatle wrote:
> > > I was helping someone with building netperf from meta-networking
> > > last night, and stumbled on what I think is a bug.
> > > 
> > > SUMMARY = "A networking benchmarking tool"
> > > DESCRIPTION = "Network performance benchmark including tests for
> > > TCP, UDP, sockets, ATM and more."
> > > SECTION = "console/network"
> > > HOMEPAGE = "http://www.netperf.org/"
> > > LICENSE = "netperf"
> > > LICENSE_FLAGS = "commercial"
> > > 
> > > In the above, the LICENSE_FLAGS are set to 'commercial'.  I think
> > > this is incorrect.  It should be set to 'non-commercial'.
> > > 
> > > There is a subtle difference between them which is why I think it's a
> > > bug...
> > > 
> > > commercial -- there are some commercial requirements necessary to
> > > use this recipe in a commercial device... you are responsible for
> > > understanding them and doing whatever is necessary... (for
> > > non-commercial devices you can likely use it...)
> > > 
> > > non-commercial -- this item is restricted to non-commercial users
> > > only.  As in the case of netperf, the license says it's only for
> > > non-commercial use.
> > > 
> > > So anyway, my suggestion is to simply change the value of the flag.
> > 
> > Ugh.  I had a quick look around and the first thing I found was this:
> > 
> > http://www.netperf.org/svn/netperf4/trunk/include/netperf.h
> 
> This is netperf4 though which I gather is distinct from the version of
> netperf we are building.

Yeah, that's what had me going back to the documentation in the tarball,
since given the previous license, this one is worlds apart even though
it's coming from the same folks.

> > So I had a look at the netperf source tree fetched by bitbake during a
> > build.  The COPYING file says this:
> > 
> >   The enclosed software and documentation includes copyrighted works
> >   of Hewlett-Packard Co. For as long as you comply with the following
> >   limitations, you are hereby authorized to (i) use, reproduce, and
> >   modify the software and documentation, and to (ii) distribute the
> >   software and documentation, including modifications, for
> >   non-commercial purposes only.
> > 
> > So I'm inclined to agree, this sounds like the definition of
> > non-commercial above to me.
> 
> It's a subtlety - in either case you can't just go ahead and use it for 
> commercial use. I wouldn't object to changing it to "non-commercial" though if 
> it makes more sense.
> 
> Would it be practical for us to move to netperf4 which is GPLv2+ licensed? I'm 
> not sure if it is a complete replacement for netperf 2.x and we'd need to 
> confirm that first, but at least the license is more reasonable...

I would be happy to see a netperf4-based recipe show up, but I don't
currently have any plans for one.  It would certainly be nice to have
both available, though, since 4 is not a complete replacement for 2.x.

-- 
-Joe MacDonald.
:wq

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 205 bytes --]

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2012-11-30 20:14 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2012-11-30 16:18 netperf in meta-networking Mark Hatle
2012-11-30 16:34 ` Joe MacDonald
2012-11-30 16:54   ` Joe MacDonald
2012-11-30 16:55   ` Paul Eggleton
2012-11-30 17:11     ` Mark Hatle
2012-11-30 17:56     ` Joe MacDonald

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.