Re: [linux-dvb] Merge multiproto tree

Re: [linux-dvb] Merge multiproto tree

[Grégoire FAVRE]

On Sat, Aug 30, 2008 at 1:59 PM, Manu Abraham <abraham.manu@gmail.com> wrote:
> Hello Mauro,
>
> Please pull from http://jusst.de/hg/multiproto_api_merge/
> to merge the following Changesets from the multiproto tree.

Is it a complete up to date tree (like lipliandvb which contains a
real working solution for HVR-4000) ?

Even with limited support, multiproto is good to finally have in
kernel :-)
-- 
Grégoire FAVRE
http://picasaweb.google.com/Gregoire.Favre
http://gregoire.favre.googlepages.com/
http://fr.wikipedia.org/wiki/N%C3%A9tiquette
http://en.wikipedia.org/wiki/Netiquette

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[Mauro Carvalho Chehab]

Hello, Manu,

On Sat, 30 Aug 2008, Manu Abraham wrote:

>  Hello Mauro,
>
>  Please pull from http://jusst.de/hg/multiproto_api_merge/
>  to merge the following Changesets from the multiproto tree.

The need for supporting newer DTV protocols increases day by day, since when
the first multiproto proposal started to be discussed, about two years ago.

At the end of the last year, Steven send one email to the ML with a different 
API proposal. Yet, people decided to wait for your work to be done. People then 
pinged you, from time to time, asking about the completion of multiproto. All 
the times, your answer were that multiproto were not ready yet for production.

I'm aware that your solution seems to be more code-complete than Steven's 
proposal.

But the recent activity on the mailing list regarding his idea (and its, 
so far, positive feedback) and the fact that I was anyway planning to 
have a discussion about the future of the DVB-API at the Linux Plumbers 
Conference 2008 are supporting me in my idea of post-poning such a pull to 
a point in time shortly after this event.

Cheers,
Mauro

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[Jelle De Loecker]

[-- Attachment #1.1: Type: text/plain, Size: 2446 bytes --]

At the risk of getting into this thing, I must admit I didn't know about 
these rivalries on the list.
I have only followed the production of multiproto for the past 6 months, 
not as a developer but as someone who just wants to get his DVB-S2 card 
going.

You say a lot of people back this new API-idea, but I honestly can't 
wait another 2 years to get proper support for my card. Am I wrong into 
thinking that, should multiproto be merged today, major applications 
will start to really support it long before this new api will get 
finalised?

I believe Manu also shot down the theory that multiproto wouldn't be 
able to support future modulations?

It might be so that designing a whole new api might be easier for these 
new standards in the future, but haven't we waited long enough? If it's 
true what Manu's saying, tweaking multiproto to handle these new 
standards wouldn't be that hard, and since there is no hardware for 
these new modulations anyway, shouldn't dvb-s2 get some priority?


/Met vriendelijke groeten,/

*Jelle De Loecker*
Kipdola Studios - Tomberg


Mauro Carvalho Chehab schreef:
> Hello, Manu,
>
> On Sat, 30 Aug 2008, Manu Abraham wrote:
>
>   
>>  Hello Mauro,
>>
>>  Please pull from http://jusst.de/hg/multiproto_api_merge/
>>  to merge the following Changesets from the multiproto tree.
>>     
>
> The need for supporting newer DTV protocols increases day by day, since when
> the first multiproto proposal started to be discussed, about two years ago.
>
> At the end of the last year, Steven send one email to the ML with a different 
> API proposal. Yet, people decided to wait for your work to be done. People then 
> pinged you, from time to time, asking about the completion of multiproto. All 
> the times, your answer were that multiproto were not ready yet for production.
>
> I'm aware that your solution seems to be more code-complete than Steven's 
> proposal.
>
> But the recent activity on the mailing list regarding his idea (and its, 
> so far, positive feedback) and the fact that I was anyway planning to 
> have a discussion about the future of the DVB-API at the Linux Plumbers 
> Conference 2008 are supporting me in my idea of post-poning such a pull to 
> a point in time shortly after this event.
>
> Cheers,
> Mauro
>
> _______________________________________________
> linux-dvb mailing list
> linux-dvb@linuxtv.org
> http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
>   

[-- Attachment #1.2: Type: text/html, Size: 3178 bytes --]

[-- Attachment #2: Type: text/plain, Size: 150 bytes --]

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[VDR User]

On Mon, Sep 1, 2008 at 8:06 AM, Mauro Carvalho Chehab
<mchehab@infradead.org> wrote:
> I'm aware that your solution seems to be more code-complete than Steven's
> proposal.
>
> But the recent activity on the mailing list regarding his idea (and its,
> so far, positive feedback) and the fact that I was anyway planning to
> have a discussion about the future of the DVB-API at the Linux Plumbers
> Conference 2008 are supporting me in my idea of post-poning such a pull to
> a point in time shortly after this event.

I understand peoples frustration in waiting for multiproto since I'm
one of them as well.  However, I believe the support for Steven's
proposal is largely because people aren't aware that multiproto is now
in a ready-state and has a pull request pending.  Over the last
several months I've seen many question when/if multiproto will be
done, or if it is dead...  I think we all agree that it has taken
quite some time for multiproto to get to a point where it's ready but
that time has come.

Multiproto -can- be pulled in right now, and if that happened, drivers
could be written immediately, finally providing users with what
they've needed for so long.  In my opinion it makes no sense to throw
out a robust api that is ready right now just because of frustration
and past personal grudges (whether anyone will openly admit to this or
not, it -is- a part of this).  The questions for consideration -should
be-... Is the code ready?  Can it handle future specs?  Is it missing
anything that should be included?  If the code is ready and is robust,
then the final question is what benefit is there in making people wait
yet longer for another api to be written?  Will this new api proposal
offer anything that multiproto doesn't already?

It seems we can finally move forward and now instead of incomplete
code stopping it, politics are.

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[P. van Gaans]

On 09/01/2008 06:00 PM, VDR User wrote:
> On Mon, Sep 1, 2008 at 8:06 AM, Mauro Carvalho Chehab
> <mchehab@infradead.org> wrote:
>> I'm aware that your solution seems to be more code-complete than Steven's
>> proposal.
>>
>> But the recent activity on the mailing list regarding his idea (and its,
>> so far, positive feedback) and the fact that I was anyway planning to
>> have a discussion about the future of the DVB-API at the Linux Plumbers
>> Conference 2008 are supporting me in my idea of post-poning such a pull to
>> a point in time shortly after this event.
> 
> I understand peoples frustration in waiting for multiproto since I'm
> one of them as well.  However, I believe the support for Steven's
> proposal is largely because people aren't aware that multiproto is now
> in a ready-state and has a pull request pending.  Over the last
> several months I've seen many question when/if multiproto will be
> done, or if it is dead...  I think we all agree that it has taken
> quite some time for multiproto to get to a point where it's ready but
> that time has come.
> 
> Multiproto -can- be pulled in right now, and if that happened, drivers
> could be written immediately, finally providing users with what
> they've needed for so long.  In my opinion it makes no sense to throw
> out a robust api that is ready right now just because of frustration
> and past personal grudges (whether anyone will openly admit to this or
> not, it -is- a part of this).  The questions for consideration -should
> be-... Is the code ready?  Can it handle future specs?  Is it missing
> anything that should be included?  If the code is ready and is robust,
> then the final question is what benefit is there in making people wait
> yet longer for another api to be written?  Will this new api proposal
> offer anything that multiproto doesn't already?
> 
> It seems we can finally move forward and now instead of incomplete
> code stopping it, politics are.
> 
> _______________________________________________
> linux-dvb mailing list
> linux-dvb@linuxtv.org
> http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb
> 

If what VDR User and Jelle De Loecker are saying is right, I don't see a 
reason for not going with multiproto either.

I'm an enduser and I would be happy with multiproto as well. I think I'm 
not alone if I say I care less *what* the solution is going to be, as 
long as we get DVB-S2 support in-kernel.

P. van Gaans

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

[linux-dvb] Merge multiproto tree

[lucian orasanu]

Hy.

I'am using vdr with mulltiproto for 7 monts from now, and even plugins got patches and support for multiproto, even mythtv has patches to support multiproto and I'am not hapy to wait another 2 years for new API. 
So I pull for merge multiproto.


      

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[Hans Werner]

> Hello Mauro,
> 
> Please pull from http://jusst.de/hg/multiproto_api_merge/
> to merge the following Changesets from the multiproto tree.
> 
> 8876
> 8875
> 8874
> 8873
> 8872
> 8871
> 8870
> 8869
> 8868
> 8867
> 8866
> 8865
> 8864
> 8863
> 8862
> 8861
> 8860
> 8859
> 8858
> 8857
> 8856
> 8855
> 8854
> 8853
> 8852
> 8851
> 8850
> 8849
> 8848
> 8847
> 8846
> 
> The tree is being pushed out, it might take a little while for the
> repository to be populated due to the slow connect.
> 
> Regards,
> Manu

that pull request does not include the following code, does it?

stb0899
stb6100
mantis

Regards,
Hans


-- 
Release early, release often.

Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten 
Browser-Versionen downloaden: http://www.gmx.net/de/go/browser

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[Manu Abraham]

Hans Werner wrote:
> that pull request does not include the following code, does it?
> 
> stb0899
> stb6100
> mantis

The above mentioned requires the API change and the relevant callbacks
to exist, without which it wouldn't even compile. So it makes sense to
merge the API changes initially prior to anything else.

Regards,
Manu


_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[Christophe Thommeret]

Le Monday 01 September 2008 22:33:40 Manu Abraham, vous avez écrit :

> Something seems to be stinking. There is no reason to legitimately delay
> so much. If the intention is for a merge, please do so within the next
> few days.

Hi Manu,

Funny to see that you are now giving multiproto merge a very high priority :)

Anyway, as an application dev, i prefer Steve's proposal. And porting drivers 
to this new API shouldn't take ages.

-- 
Christophe Thommeret


_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[Manu Abraham]

Christophe Thommeret wrote:
> Le Monday 01 September 2008 22:33:40 Manu Abraham, vous avez écrit :
> 
>> Something seems to be stinking. There is no reason to legitimately delay
>> so much. If the intention is for a merge, please do so within the next
>> few days.
> 
> Hi Manu,
> 
> Funny to see that you are now giving multiproto merge a very high priority :)
> 

Once it is ready to be merged, there's no reason to hold back a merge
and hence ..

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[Jelle De Loecker]

[-- Attachment #1.1: Type: text/plain, Size: 782 bytes --]


Christophe Thommeret schreef:
> Le Monday 01 September 2008 22:33:40 Manu Abraham, vous avez écrit :
>   
>> Something seems to be stinking. There is no reason to legitimately delay
>> so much. If the intention is for a merge, please do so within the next
>> few days.
>>     
> Hi Manu,
>
> Funny to see that you are now giving multiproto merge a very high priority :)
>
> Anyway, as an application dev, i prefer Steve's proposal. And porting drivers 
> to this new API shouldn't take ages.
>   
I'm sorry for my ignorance, all this drivers and api scene is still new 
territory for me, as I said before :)
but could you explain why designing a whole new api is preferable to an 
application developer?


/Met vriendelijke groeten,/

*Jelle De Loecker*
Kipdola Studios - Tomberg


[-- Attachment #1.2: Type: text/html, Size: 1277 bytes --]

[-- Attachment #2: Type: text/plain, Size: 150 bytes --]

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[Mauro Carvalho Chehab]

Hi Marcel,

On Mon, 01 Sep 2008 18:22:59 +0200
Marcel Siegert <mws@linuxtv.org> wrote:

> hello mauro,
> 
> Mauro Carvalho Chehab schrieb:
> > Hello, Manu,
> > 
> > On Sat, 30 Aug 2008, Manu Abraham wrote:
> > 
> >>  Hello Mauro,
> >>
> >>  Please pull from http://jusst.de/hg/multiproto_api_merge/
> >>  to merge the following Changesets from the multiproto tree.
> > 
> > The need for supporting newer DTV protocols increases day by day, since when
> > the first multiproto proposal started to be discussed, about two years ago.
> > 
> > At the end of the last year, Steven send one email to the ML with a different 
> > API proposal. Yet, people decided to wait for your work to be done. People then 
> > pinged you, from time to time, asking about the completion of multiproto. All 
> > the times, your answer were that multiproto were not ready yet for production.
> > 
> how is the actual state of his proposal? 

Steven can explain the details. Basically: it works, it is simple to
understand and to work with, and it seems to be flexible enough to support all
current needs and seems to be flexible enough to support future protocols. It is
currently under review based on community feedback.

I didn't have time yet to do carefully inspect the latest version of the
multiproto. I suspect that nobody did it yet, since I didn't see any technical
analysis of the current proposal.

I'll carefully review multiproto proposal during my trip, and compare it
with the Stoth's proposal.

> > I'm aware that your solution seems to be more code-complete than Steven's 
> > proposal.
> 
> > 
> > But the recent activity on the mailing list regarding his idea (and its, 
> > so far, positive feedback) and the fact that I was anyway planning to 
> > have a discussion about the future of the DVB-API at the Linux Plumbers 
> > Conference 2008 are supporting me in my idea of post-poning such a pull to 
> > a point in time shortly after this event.
> with whom? the linuxtv developers or with an attending audience?

We'll have some sort of panel or speech for the attending audience of the Video input
infrastructure miniconf. The API analysis will happen with the Linuxtv
developers that will be there.

> i havent been on the list and that active since the "nearly two years war" begun,
> thus i am not the one to ack or nack in this pull request.
> 
> what i did in the past was just watching what was happening and what was taking
> progress - even if it is slow. 
> 
> users on the linuxtv list started to ask over and over again, when multiproto is
> going to be merged. vdr started to support it in a kind of experiment.
> 
> due to the lack of spare time and devices :) i do not actually know how productive 
> the vdr multiproto implementation is, but we should not start the whole discussions again.
> 
> if we do not merge multiproto now, we will never do it, i am afraid. 
> 
> for the future of this project it is more than neccessary to get some progress.
> 
> if there are no serious objections to multiproto, mauro, please pull/merge it 
> within the next few days. to wait until the end of september is time spending without
> any sense.

As I've already explained, merging today or after Plumbers won't make any
practical difference. The upstream merge will only happen after 2.6.27 being
released, due to the next kernel window for merging patches. This is a major
change at API and claims for a carefully analysis.

Cheers,
Mauro

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[Christophe Thommeret]

Le Monday 01 September 2008 23:09:10 Jelle De Loecker, vous avez écrit :
> Christophe Thommeret schreef:
> > Le Monday 01 September 2008 22:33:40 Manu Abraham, vous avez écrit :
> >> Something seems to be stinking. There is no reason to legitimately delay
> >> so much. If the intention is for a merge, please do so within the next
> >> few days.
> >
> > Hi Manu,
> >
> > Funny to see that you are now giving multiproto merge a very high
> > priority :)
> >
> > Anyway, as an application dev, i prefer Steve's proposal. And porting
> > drivers to this new API shouldn't take ages.
>
> I'm sorry for my ignorance, all this drivers and api scene is still new
> territory for me, as I said before :)
> but could you explain why designing a whole new api is preferable to an
> application developer?

Both Manu's and Steve's API proposals are new since none are approved/merged 
yet.
I know that some guys have made multiproto patches for various applications, 
but this should not be a major argument in favor of multiproto (tuning code 
is a very very little part of a DVB app).


-- 
Christophe Thommeret


_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb

Re: [linux-dvb] Merge multiproto tree

[Simon Kenyon]

VDR User wrote:
> It's no secret Mauro and company are trying to sweep multiproto under
> the rug so I too am curious to hear about how those multiproto talks
> went.
>
> By the way, maybe someone will have the decency to remove the
> misleading b.s. and lies from
> http://www.linuxtv.org/wiki/index.php/Multiproto
>   
mister sombody is currently on holiday
perhaps mister anybody will do it for him

_______________________________________________
linux-dvb mailing list
linux-dvb@linuxtv.org
http://www.linuxtv.org/cgi-bin/mailman/listinfo/linux-dvb