Re: POLL: for/against dropping support for kernels < 2.6.22

Re: POLL: for/against dropping support for kernels < 2.6.22

[Jean Delvare]

Hi Hans,

> There are lot's of discussions, but it can be hard sometimes to actually 
> determine someone's opinion.
> 
> So here is a quick poll, please reply either to the list or directly to me 
> with your yes/no answer and (optional but welcome) a short explanation to 
> your standpoint. It doesn't matter if you are a user or developer, I'd like 
> to see your opinion regardless.
> 
> Please DO NOT reply to the replies, I'll summarize the results in a week's 
> time and then we can discuss it further.
> 
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
> 

X: Yes

> _: No
> 
> Optional question:
> 
> Why:

The cost to preserve backwards compatibility for these old kernels is
much too high compared to the remaining user-base. I can only repeat
the points I have made in the past week:
* Maintained distributions aimed at home users (Fedora, openSUSE) run
  kernels >= 2.6.22 by now.
* Enterprise-class distributions (RHEL, SLED) are not the right target
  for the v4l-dvb repository, so we don't care which kernels these are
  running.
* Engineering time which is put into backwards compatibility would be
  better spent on improving the drivers upstream and adding support
  for new hardware faster.
* v4l-dvb depends on subsystems which do evolve, and when these changes
  are too important (e.g. new i2c device driver binding model)
  backwards compatibility comes are an unbearable complexity and cost.
  That kind of cost sucks the time of current developers, might turn
  them into ex-developers when they realize they lost all the fun, and
  prevents new developers from joining the project because of the
  complexity of the compatibility layer.

So let's just drop support for kernels < 2.6.22 and focus on better
supporting upstream and recent kernels.

Thanks,
-- 
Jean Delvare

Re: POLL: for/against dropping support for kernels < 2.6.22

[Trent Piepho]

On Mon, 23 Feb 2009, Jean Delvare wrote:
> > There are lot's of discussions, but it can be hard sometimes to actually
> > determine someone's opinion.
> >
> > So here is a quick poll, please reply either to the list or directly to me
> > with your yes/no answer and (optional but welcome) a short explanation to
> > your standpoint. It doesn't matter if you are a user or developer, I'd like
> > to see your opinion regardless.
> >
> > Please DO NOT reply to the replies, I'll summarize the results in a week's
> > time and then we can discuss it further.
> >
> > Should we drop support for kernels <2.6.22 in our v4l-dvb repository?

Does this mean keep our current system and move the backward compatibility
point to 2.6.22?

Or not have any backward compatibilty at all?

Re: POLL: for/against dropping support for kernels < 2.6.22

[John Pilkington]

Jean Delvare wrote:

> * Enterprise-class distributions (RHEL, SLED) are not the right target
>   for the v4l-dvb repository, so we don't care which kernels these are
>   running.


I think you should be aware that the mythtv and ATrpms communities 
include a significant number of people who have chosen to use the 
CentOS_5 series in the hope of getting systems that do not need to be 
reinstalled every few months.  I hope you won't disappoint them.

John P

Re: POLL: for/against dropping support for kernels < 2.6.22

[Simon Kenyon]

John Pilkington wrote:
> Jean Delvare wrote:
>
>> * Enterprise-class distributions (RHEL, SLED) are not the right target
>>   for the v4l-dvb repository, so we don't care which kernels these are
>>   running.
>
>
> I think you should be aware that the mythtv and ATrpms communities 
> include a significant number of people who have chosen to use the 
> CentOS_5 series in the hope of getting systems that do not need to be 
> reinstalled every few months.  I hope you won't disappoint them.
>
> John P
just had a quick look at mythtv-users
there are a handful using centos
--
simon

Re: POLL: for/against dropping support for kernels < 2.6.22

[Jean Delvare]

Hi John,

(re-adding linux-media on Cc as I doubt you dropped it on purpose...)

On Tue, 24 Feb 2009 11:46:04 +0000, John Pilkington wrote:
> Jean Delvare wrote:
> 
> > * Enterprise-class distributions (RHEL, SLED) are not the right target
> >   for the v4l-dvb repository, so we don't care which kernels these are
> >   running.
> 
> I think you should be aware that the mythtv and ATrpms communities 
> include a significant number of people who have chosen to use the 
> CentOS_5 series in the hope of getting systems that do not need to be 
> reinstalled every few months.  I hope you won't disappoint them.

CentOS is a parasite, if it dies I can't care less. CentOS users have
the recurrent habit to expect professional support from the community
without giving anything in return. Even worse: they consider that
running an antediluvian OS is the default and they don't understand why
upstream developers won't help them.

You said it yourself: they expect to be able to keep the same system
for a long time. This is a service you normally get from Red Hat or
Novell, and you pay for it. This is something the community is
generally not willing to offer for free, because it is not fun.

If the MythTV community cares that much about the v4l-dvb tree, they are
free to fork it right before support for kernel 2.6.18 is dropped, and
maintain that copy themselves. But their model is broken to start with:
sticking to a several-year-old kernel and OS, and OTOH picking critical
(for their use case) kernel drivers from a development tree which
evolves continuously by definition, makes little sense. Then again, I
would be happy to keep support for them if the cost wasn't too high.
But right now, the cost _is_ too high.

Your view of community distributions is a bit too negative BTW. You
don't need to go to the extreme CentOS_5 is to not have to reinstall
every few months. openSUSE distributions are maintained for 2 years for
example.

-- 
Jean Delvare

Re: POLL: for/against dropping support for kernels < 2.6.22

[John Pilkington]

Hi Jean:
Jean Delvare wrote:
> Hi John,
> 
> (re-adding linux-media on Cc as I doubt you dropped it on purpose...)

Right: I do subconsciously expect 'reply' to go to the list, and don't 
always remember to check first.  No doubt you got two copies.
> 
> On Tue, 24 Feb 2009 11:46:04 +0000, John Pilkington wrote:
>> Jean Delvare wrote:
>>
>>> * Enterprise-class distributions (RHEL, SLED) are not the right target
>>>   for the v4l-dvb repository, so we don't care which kernels these are
>>>   running.
>> I think you should be aware that the mythtv and ATrpms communities 
>> include a significant number of people who have chosen to use the 
>> CentOS_5 series in the hope of getting systems that do not need to be 
>> reinstalled every few months.  I hope you won't disappoint them.
> 
> CentOS is a parasite, if it dies I can't care less. CentOS users have
> the recurrent habit to expect professional support from the community
> without giving anything in return. Even worse: they consider that
> running an antediluvian OS is the default and they don't understand why
> upstream developers won't help them.
> 
> You said it yourself: they expect to be able to keep the same system
> for a long time. This is a service you normally get from Red Hat or
> Novell, and you pay for it. This is something the community is
> generally not willing to offer for free, because it is not fun.

I know there is a range of visions from linux-as-playground to 
linux-as-tool. Sometimes just learning to use the tool is not a trivial 
matter, and shows enthusiasm too. I recognize that enthusiasm is the 
great driver of development, but would still hope that users might 
expect to be able to slow their personal treadmills without actually 
falling off. :-)
> 
> If the MythTV community cares that much about the v4l-dvb tree, they are
> free to fork it right before support for kernel 2.6.18 is dropped, and
> maintain that copy themselves. But their model is broken to start with:
> sticking to a several-year-old kernel and OS, and OTOH picking critical
> (for their use case) kernel drivers from a development tree which
> evolves continuously by definition, makes little sense. Then again, I
> would be happy to keep support for them if the cost wasn't too high.
> But right now, the cost _is_ too high.
> 
> Your view of community distributions is a bit too negative BTW. You
> don't need to go to the extreme CentOS_5 is to not have to reinstall
> every few months. openSUSE distributions are maintained for 2 years for
> example.
> 

Re: POLL: for/against dropping support for kernels < 2.6.22

[Hans Verkuil]

> Hans Verkuil wrote:
>> Hi all,
>>
>> There are lot's of discussions, but it can be hard sometimes to actually
>> determine someone's opinion.
>>
>> So here is a quick poll, please reply either to the list or directly to
>> me
>> with your yes/no answer and (optional but welcome) a short explanation
>> to
>> your standpoint. It doesn't matter if you are a user or developer, I'd
>> like
>> to see your opinion regardless.
>>
>> Please DO NOT reply to the replies, I'll summarize the results in a
>> week's
>> time and then we can discuss it further.
>>
>> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>>
>> _: Yes
>> _: No
>>
> No
>> Optional question:
>>
>> Why:
>>
>>
> i don't have a vote as i'm only a user and not a developer

Users *especially* have a vote. This poll is among others meant to get a
feeling for how important people think the backwards compat is. I think it
is of limited importance, but I've no way of knowing that for certain
unless I get feedback. So I invite anyone, developer or user, to give
their opinion!

And it's not a case of 'most votes count'. It's not that type of a poll.
It's really a survey. (Hmm, that would have been a better name for this
anyway. So sue me :-) ).

Regards,

         Hans

> but i thought i would just make one point
>
> as far as i can see, the v4l-dvb tree exists to create support for a
> particular class of hardware within the linux kernel
> the separate tree is very useful to lots of people (i include myself in
> that) - but it is a byproduct of the development methodology
>
> so if you think this group's mission is to provide support for
> distributions then you should vote no
> and if you think this group's mission is to provide support for the
> linux kernel then you should vote yes
>
>>
>> Thanks,
>>
>> 	Hans
>>
>>
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-media" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>


-- 
Hans Verkuil - video4linux developer - sponsored by TANDBERG

Re: POLL: for/against dropping support for kernels < 2.6.22

[Hans Verkuil]

Hi Trent,

> On Mon, 23 Feb 2009, Jean Delvare wrote:
>> > There are lot's of discussions, but it can be hard sometimes to
>> actually
>> > determine someone's opinion.
>> >
>> > So here is a quick poll, please reply either to the list or directly
>> to me
>> > with your yes/no answer and (optional but welcome) a short explanation
>> to
>> > your standpoint. It doesn't matter if you are a user or developer, I'd
>> like
>> > to see your opinion regardless.
>> >
>> > Please DO NOT reply to the replies, I'll summarize the results in a
>> week's
>> > time and then we can discuss it further.
>> >
>> > Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> Does this mean keep our current system and move the backward compatibility
> point to 2.6.22?
>
> Or not have any backward compatibilty at all?

It was a bit imprecise, perhaps. With 'drop support' I mean in practice
that we:

1) do a one time effort to make everything compile from 2.6.16 onwards
(there are several compile issues right now with older kernels).

2) when it's OK, make a copy of the master repository, call it v4l-dvb-old
or whatever, and people who are still on old kernels can use that to at
least have the possibility to get something newer. We might even do the
occasional update if some important bug is found.

3) strip our master repository from any compatibility code needed to
support the pre-2.6.22 kernels and continue development based on that
code.

We still need to support kernels from 2.6.22 onwards. Although I think the
minimum supported kernel is something that needs a regular sanity check,
right now there are no technical reasons that I am aware of to go to
something newer than 2.6.22.

Whether we keep our current system or not is a separate discussion:
whatever development system you choose there will be work involved in
keeping up the backwards compatibility.

Hope this explains it,

          Hans

-- 
Hans Verkuil - video4linux developer - sponsored by TANDBERG

Re: POLL: for/against dropping support for kernels < 2.6.22

[sonofzev]

Yes... 


On Mon Feb 23 12:13 , hermann pitton  sent:

>
>Am Sonntag, den 22.02.2009, 11:15 +0100 schrieb Hans Verkuil:
>> Hi all,
>> 
>> There are lot's of discussions, but it can be hard sometimes to actually 
>> determine someone's opinion.
>> 
>> So here is a quick poll, please reply either to the list or directly to me 
>> with your yes/no answer and (optional but welcome) a short explanation to 
>> your standpoint. It doesn't matter if you are a user or developer, I'd like 
>> to see your opinion regardless.
>> 
>> Please DO NOT reply to the replies, I'll summarize the results in a week's 
>> time and then we can discuss it further.
>> 
>> Should we drop support for kernels 
>> 
>> _: Yes
>> _: No
>
>Yes.
>
>> Optional question:
>> 
>> Why:
>
>Keeping too old kernels supported makes others lazy and in worst case
>they ask you to support v4l2 version one. (happened)
>
>Our user base for new devices is covered with down to 2.6.22 for now, we
>likely never got anything from those on old commercial distribution
>kernels, same for Debian and stuff derived from there.
>
>Since new drivers actually prefer to avoid the compat work and are happy
>to make it just into the latest rc1 during the merge window and further
>from there, there is no loss either.
>
>Some new devices we likely get on already established drivers should not
>be hard to add to a v4l-dvb tar ball we leave with support for the even
>older kernels.
>
>Cheers,
>Hermann
> 
>
>--
>To unsubscribe from this list: send the line "unsubscribe linux-media" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at  http://vger.kernel.org/majordomo-info.html
>)

POLL: for/against dropping support for kernels < 2.6.22

[Hans Verkuil]

Hi all,

There are lot's of discussions, but it can be hard sometimes to actually 
determine someone's opinion.

So here is a quick poll, please reply either to the list or directly to me 
with your yes/no answer and (optional but welcome) a short explanation to 
your standpoint. It doesn't matter if you are a user or developer, I'd like 
to see your opinion regardless.

Please DO NOT reply to the replies, I'll summarize the results in a week's 
time and then we can discuss it further.

Should we drop support for kernels <2.6.22 in our v4l-dvb repository?

_: Yes
_: No

Optional question:

Why:



Thanks,

	Hans

-- 
Hans Verkuil - video4linux developer - sponsored by TANDBERG

Re: POLL: for/against dropping support for kernels < 2.6.22

[CityK]

Hans Verkuil wrote:
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>   

Yes


> Optional question:
>
> Why:
>   

Its causing skilled developers to waste time that would be better served
in other areas.  Because of that, these skilled volunteers are becoming
frustrated and losing their interest in pressing forth.  

It causes unnecessary complexity.  The golden rule is to keep things as
simple as possible.

It presents a hurdle to attracting new development talent (both
corporate and individual).

When upstream technical changes (such as i2c subsystem changes) have
made backporting downstream a nightmare, it is time to seriously
evaluate why you are even bothering doing such.  The salient point is
that it is absolutely illogical for volunteers to be catering to narrow
commercial interests. 
- Arguments about appeasing the needs of Enterprise distro's are moot.  
V4L-DVB owes them nothing.  Enterprise distro's are specifically that --
an enterprise's  work; if they crave support, then they can put Hans (or
whomever) on the payroll to backport for their specific needs.
- Arguments about appeasing the needs of embedded distros/platforms are
moot.   V4L-DVB owes them nothing.  Let those groups figure out and/or
support such device needs on their own; else they can put Hans (or
whomever) on the payroll.   Those manufactures releasing products within
this space will adapt to whatever V4L-DVB does.    This space will not
suddenly fall apart because of our decision.   These entrepreneurs have
entered this space specifically to exploit a market opportunity.  If
they exit, someone else will move in.  Its simple free market
dynamics.   (As it is, they are getting a free lunch ... seriously, I
think that when the embedded space looks at how bent over accommodating
we currently are, they must be rubbing their hands together and
gleefully repeating Flounders statement: Oh boy, is this great!
(http://www.acmewebpages.com/midi/great.wav))

The V4L-DVB is lacking in strategic direction.  Yesterday was the time
to adopt one; so lets pick up one today!

I believe the plan to currently backport to 2.6.22 but to bump/narrow
the kernel support window to the ideal/easier_to_maintain 2.6.25, once
express support from the big 3 desktop distos ends, is the most logical
choice and the one which will have the most beneficial impact on the
project's future.

Re: POLL: for/against dropping support for kernels < 2.6.22

[Andy Walls]

On Sun, 2009-02-22 at 14:12 -0500, CityK wrote:

> The V4L-DVB is lacking in strategic direction.  Yesterday was the time
> to adopt one; so lets pick up one today!

CityK,

I see you've been reading (or channeling) my blathering:

http://www.linuxtv.org/irc/v4l/index.php?date=2009-02-20

([19:42] to [20:21])

Regards,
Andy

Re: POLL: for/against dropping support for kernels < 2.6.22

[Guennadi Liakhovetski]

On Sun, 22 Feb 2009, Hans Verkuil wrote:

> Hi all,
> 
> There are lot's of discussions, but it can be hard sometimes to actually 
> determine someone's opinion.
> 
> So here is a quick poll, please reply either to the list or directly to me 
> with your yes/no answer and (optional but welcome) a short explanation to 
> your standpoint. It doesn't matter if you are a user or developer, I'd like 
> to see your opinion regardless.
> 
> Please DO NOT reply to the replies, I'll summarize the results in a week's 
> time and then we can discuss it further.
> 
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
> 
> _: Yes
> _: No

Yes

> 
> Optional question:
> 
> Why:

This shall free human resources necessary for performing the switch to the 
full-kernel development model.

Thanks
Guennadi
---
Guennadi Liakhovetski, Ph.D.
Freelance Open-Source Software Developer

Re: POLL: for/against dropping support for kernels < 2.6.22

[kilgota]

On Sun, 22 Feb 2009, Hans Verkuil wrote:

> Hi all,
>
> There are lot's of discussions, but it can be hard sometimes to actually
> determine someone's opinion.
>
> So here is a quick poll, please reply either to the list or directly to me
> with your yes/no answer and (optional but welcome) a short explanation to
> your standpoint. It doesn't matter if you are a user or developer, I'd like
> to see your opinion regardless.
>
> Please DO NOT reply to the replies, I'll summarize the results in a week's
> time and then we can discuss it further.
>
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
X _: Yes
> _: No
>
> Optional question:
>
> Why:

After a certain point it becomes in practical terms impossible to support 
old versions of anything. There are too many dependencies on too many 
things that have to be changed all at once. The resulting problems do not 
pertain only to kernel-related development but to all development, as I 
have tried to make clear in other posts. I do not know the gory details of 
just what has become too difficult, as I am new to this area of kernel 
development, but I am quite willing, based upon a general description, and 
based upon other experience, to believe that there are problems.

I think it is obvious that a version cutoff has to be made somewhere, and 
seven minor versions behind the kernel which is about to come out does not 
at all appear to me to be an unreasonable restriction.

Re: POLL: for/against dropping support for kernels < 2.6.22

[hermann pitton]

Am Sonntag, den 22.02.2009, 11:15 +0100 schrieb Hans Verkuil:
> Hi all,
> 
> There are lot's of discussions, but it can be hard sometimes to actually 
> determine someone's opinion.
> 
> So here is a quick poll, please reply either to the list or directly to me 
> with your yes/no answer and (optional but welcome) a short explanation to 
> your standpoint. It doesn't matter if you are a user or developer, I'd like 
> to see your opinion regardless.
> 
> Please DO NOT reply to the replies, I'll summarize the results in a week's 
> time and then we can discuss it further.
> 
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
> 
> _: Yes
> _: No

Yes.

> Optional question:
> 
> Why:

Keeping too old kernels supported makes others lazy and in worst case
they ask you to support v4l2 version one. (happened)

Our user base for new devices is covered with down to 2.6.22 for now, we
likely never got anything from those on old commercial distribution
kernels, same for Debian and stuff derived from there.

Since new drivers actually prefer to avoid the compat work and are happy
to make it just into the latest rc1 during the merge window and further
from there, there is no loss either.

Some new devices we likely get on already established drivers should not
be hard to add to a v4l-dvb tar ball we leave with support for the even
older kernels.

Cheers,
Hermann
 

Re: POLL: for/against dropping support for kernels < 2.6.22

[Robert Golding]

2009/2/22 Hans Verkuil <hverkuil@xs4all.nl>:
> Hi all,
>
> There are lot's of discussions, but it can be hard sometimes to actually
> determine someone's opinion.
>
> So here is a quick poll, please reply either to the list or directly to me
> with your yes/no answer and (optional but welcome) a short explanation to
> your standpoint. It doesn't matter if you are a user or developer, I'd like
> to see your opinion regardless.
>
> Please DO NOT reply to the replies, I'll summarize the results in a week's
> time and then we can discuss it further.
>
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>

 ** Yes **


> _: No
>
> Optional question:
>
> Why: (from a non-coder, I have failed miserably to learn how to code in anything other than shell scripting)

I think the development of later drivers suffers because of the work
needed for backwards kernel compatibility.

In any case, I think most home users (like me) will usually be very up
to date with their kernels (we so like to tinker where we can) so the
drivers for older kernels would only be of any use to those using
enterprise kernels, and I think those should be addressed by the
people being paid!  Isn't that what they're being paid for?

>
> Thanks,
>
>        Hans
>
> --
> Hans Verkuil - video4linux developer - sponsored by TANDBERG
> --
> To unsubscribe from this list: send the line "unsubscribe linux-media" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>



-- 
Regards,	Robert

..... Some people can tell what time it is by looking at the sun, but
I have never been able to make out the numbers.
---
Errata: Spelling mistakes are not intentional, however, I don't use
spell checkers because it's too easy to allow the spell checker to
make the decisions and use words that are out of context for that
being written, i.e. their/there, your/you're, threw/through and even
accept/except, not to mention foreign (I'm Australian) English
spelling, i.e. colour/color, socks/sox, etc,.

Re: POLL: for/against dropping support for kernels < 2.6.22

[VDR User]

On Sun, Feb 22, 2009 at 2:15 AM, Hans Verkuil <hverkuil@xs4all.nl> wrote:
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> _: Yes
> _: No

Yes.

> Optional question:
>
> Why:

The reasons already stated, those resources could be better used doing
other things.  Aside of that, of the devs/users how many people
actually _need_ to remain on an old kernel.  I could be wrong in my
assumption that most people using old kernels are doing so simply by
choice and not necessity.  You want to maximize developer productivity
and if that means some people will need to update their kernel, is
that so horrible?

Re: POLL: for/against dropping support for kernels < 2.6.22

[Mauro Carvalho Chehab]

On Sun, 22 Feb 2009 11:15:01 +0100
Hans Verkuil <hverkuil@xs4all.nl> wrote:

> Hi all,
> 
> There are lot's of discussions, but it can be hard sometimes to actually 
> determine someone's opinion.
> 
> So here is a quick poll, please reply either to the list or directly to me 
> with your yes/no answer and (optional but welcome) a short explanation to 
> your standpoint. It doesn't matter if you are a user or developer, I'd like 
> to see your opinion regardless.
> 
> Please DO NOT reply to the replies, I'll summarize the results in a week's 
> time and then we can discuss it further.
> 
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
> 
> _: Yes
> _: No

No.

> 
> Optional question:
> 
> Why:

For a couple of reasons:

1) This will remove testers from our user database;
2) The current way of backporting is not scaling. Just dropping support for a
random version is just postponing the question that we need to re-think about
the way for backport;
3) This doesn't solve the development issues we have of not using -git. This
causes lots of work when sending patches uptreaming, on when someone changes
something upstream and a backport is needed.

So, in practice, this won't solve any real problem.

I'm right now working on another way of allowing backport that will better
scale, and will allow developers to use -git, without losing backport for users.

Cheers,
Mauro

Re: POLL: for/against dropping support for kernels < 2.6.22

[Mauro Carvalho Chehab]

On Mon, 23 Feb 2009 09:26:57 -0300
Mauro Carvalho Chehab <mchehab@infradead.org> wrote:


> I'm right now working on another way of allowing backport that will better
> scale, and will allow developers to use -git, without losing backport for users.

I have an incomplete skeleton for the backport scripts, available at:

http://linuxtv.org/hg/~mchehab/backport

For now, it is very dumb (it recompiles all drivers every time) and requires
much more hacking to cleanup the Makefiles.

The current version just removes a very simple check for linux version, but it
is not hard to use this way for all cases where backport is needed. After
having this working fine and supporting all backports, people can develop using
-git as basis for development, without needing to take care of backport anymore.

Cheers,
Mauro

Re: POLL: for/against dropping support for kernels < 2.6.22

[Douglas Schilling Landgraf]

Hello Hans,

On Sun, 22 Feb 2009 11:15:01 +0100
Hans Verkuil <hverkuil@xs4all.nl> wrote:

> Hi all,
> 
> There are lot's of discussions, but it can be hard sometimes to
> actually determine someone's opinion.
> 
> So here is a quick poll, please reply either to the list or directly
> to me with your yes/no answer and (optional but welcome) a short
> explanation to your standpoint. It doesn't matter if you are a user
> or developer, I'd like to see your opinion regardless.
> 
> Please DO NOT reply to the replies, I'll summarize the results in a
> week's time and then we can discuss it further.
> 
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
> 
> _: Yes
> _: No
> 

No

> Optional question:
> 
> Why:

I know it's not easy task keep this support working... but we
still have *users* around the world using kernel < 2.6.22 (as
some of them already reported this). 

Cheers,
Douglas

Re: POLL: for/against dropping support for kernels < 2.6.22

[David Ellingsworth]

On Sun, Feb 22, 2009 at 5:15 AM, Hans Verkuil <hverkuil@xs4all.nl> wrote:
> Hi all,
>
> There are lot's of discussions, but it can be hard sometimes to actually
> determine someone's opinion.
>
> So here is a quick poll, please reply either to the list or directly to me
> with your yes/no answer and (optional but welcome) a short explanation to
> your standpoint. It doesn't matter if you are a user or developer, I'd like
> to see your opinion regardless.
>
> Please DO NOT reply to the replies, I'll summarize the results in a week's
> time and then we can discuss it further.
>
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> _: Yes
> _: No

YES

>
> Optional question:

Why can't we drop support for all but the latest kernel?

>
> Why:

As others have already pointed out, it is a waste of time for
developers who volunteer their time to work on supporting prior kernel
revisions for use by enterprise distributions. The task of
back-porting driver modifications to earlier kernels lessens the
amount of time developers can focus on improving the quality and
stability of new and existing drivers. Furthermore, it deters driver
development since  there an expectation that they will back-port their
driver to earlier kernel versions. Finally, as a developer, I have
have little interest in what was new yesterday. I usually run the
latest kernel whenever possible and for a number of different reasons.
Some of those reasons include better hardware support, bug detection,
and stability testing. All services greatly valued by other kernel
developers.

Regards,

David Ellingsworth

Re: POLL: for/against dropping support for kernels < 2.6.22

[Trent Piepho]

On Mon, 23 Feb 2009, David Ellingsworth wrote:
> On Sun, Feb 22, 2009 at 5:15 AM, Hans Verkuil <hverkuil@xs4all.nl> wrote:
> > Optional question:
>
> Why can't we drop support for all but the latest kernel?
>
> >
> > Why:
>
> As others have already pointed out, it is a waste of time for
> developers who volunteer their time to work on supporting prior kernel
> revisions for use by enterprise distributions. The task of
> back-porting driver modifications to earlier kernels lessens the
> amount of time developers can focus on improving the quality and
> stability of new and existing drivers. Furthermore, it deters driver
> development since  there an expectation that they will back-port their
> driver to earlier kernel versions. Finally, as a developer, I have

We don't backport the drivers to older kernels.  That's what drivers kept
in a full kernel tree end up doing.

Generally there is just the code for the newest kernel to think about.
Most of the driver code doesn't have backward compatibility ifdefs.  Most
of the compat issues are handled transparently by compat.h and only those
developers who patch compat.h ever need to know they exist.

When a developer does need to deal with some compat ifdef in a driver,
almost all the time it's something trivial and obvious.  Change the
variable name in both branches.  Copy in a couple lines of boilerplate.

Sometimes a bigger issue comes up.  IIRC, around 2.6.16 there was a major
class_device change in the kernel and backward compat code for it ended up
being a nightmare.  So we didn't do it.  We stopped supporting back to
~2.6.11 and moved up the target past the problem change.

Maybe this has happened again with the changes to i2c?  I don't think
it's that hard, but I've yet to do it myself, so maybe it is.

Re: POLL: for/against dropping support for kernels < 2.6.22

[Hans Verkuil]

On Tuesday 24 February 2009 06:04:48 Trent Piepho wrote:
> On Mon, 23 Feb 2009, David Ellingsworth wrote:
> > On Sun, Feb 22, 2009 at 5:15 AM, Hans Verkuil <hverkuil@xs4all.nl> 
wrote:
> > > Optional question:
> >
> > Why can't we drop support for all but the latest kernel?
> >
> > > Why:
> >
> > As others have already pointed out, it is a waste of time for
> > developers who volunteer their time to work on supporting prior kernel
> > revisions for use by enterprise distributions. The task of
> > back-porting driver modifications to earlier kernels lessens the
> > amount of time developers can focus on improving the quality and
> > stability of new and existing drivers. Furthermore, it deters driver
> > development since  there an expectation that they will back-port their
> > driver to earlier kernel versions. Finally, as a developer, I have
>
> We don't backport the drivers to older kernels.  That's what drivers kept
> in a full kernel tree end up doing.
>
> Generally there is just the code for the newest kernel to think about.
> Most of the driver code doesn't have backward compatibility ifdefs.  Most
> of the compat issues are handled transparently by compat.h and only those
> developers who patch compat.h ever need to know they exist.
>
> When a developer does need to deal with some compat ifdef in a driver,
> almost all the time it's something trivial and obvious.  Change the
> variable name in both branches.  Copy in a couple lines of boilerplate.
>
> Sometimes a bigger issue comes up.  IIRC, around 2.6.16 there was a major
> class_device change in the kernel and backward compat code for it ended
> up being a nightmare.  So we didn't do it.  We stopped supporting back to
> ~2.6.11 and moved up the target past the problem change.

Actually that was in 2.6.19. The class_device #ifs are still in e.g. 
v4l2-dev.c. It would be a nice bonus when we can drop that as well. It 
could be that there were additional changes as well in pre-2.6.16 kernels. 
If so, then we definitely implemented the backwards compat for it at the 
time.

> Maybe this has happened again with the changes to i2c?  I don't think
> it's that hard, but I've yet to do it myself, so maybe it is.

I've been working on this since around 2.6.24 (and been involved with i2c in 
one way or another for quite a bit longer) and I say it's hard. Jean 
Delvare made the i2c core changes in 2.6.22 and he says it's hard. So 
perhaps if the two people who know most about the topic say it's hard and 
not solvable with a compat.h change, or the occasional #if, or a regexp as 
Mauro seems to be attempting now, then it really IS hard.

Regards,

	Hans

-- 
Hans Verkuil - video4linux developer - sponsored by TANDBERG

Re: POLL: for/against dropping support for kernels < 2.6.22

[Ales Jurik]

On Sunday 22 of February 2009, Hans Verkuil wrote:
> Hi all,
>
> There are lot's of discussions, but it can be hard sometimes to actually
> determine someone's opinion.
>
> So here is a quick poll, please reply either to the list or directly to me
> with your yes/no answer and (optional but welcome) a short explanation to
> your standpoint. It doesn't matter if you are a user or developer, I'd like
> to see your opinion regardless.
>
> Please DO NOT reply to the replies, I'll summarize the results in a week's
> time and then we can discuss it further.
>
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> _: Yes
> _: No
>
Yes

> Optional question:
>
> Why:
>
>
>
> Thanks,
>
> 	Hans

Re: POLL: for/against dropping support for kernels < 2.6.22

[Hans Verkuil]

Please reply to this poll if you haven't done so yet. I only count clear 
yes/no answers, so if you replied earlier to this in order to discuss a 
point then I haven't counted that.

Currently it's 16 yes and 2 no votes, but I'd really like to see some more 
input. I want to post the final results on Sunday.

Regards,

	Hans

On Sunday 22 February 2009 11:15:01 you wrote:
> Hi all,
>
> There are lot's of discussions, but it can be hard sometimes to actually
> determine someone's opinion.
>
> So here is a quick poll, please reply either to the list or directly to
> me with your yes/no answer and (optional but welcome) a short explanation
> to your standpoint. It doesn't matter if you are a user or developer, I'd
> like to see your opinion regardless.
>
> Please DO NOT reply to the replies, I'll summarize the results in a
> week's time and then we can discuss it further.
>
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> _: Yes
> _: No
>
> Optional question:
>
> Why:
>
>
>
> Thanks,
>
> 	Hans



-- 
Hans Verkuil - video4linux developer - sponsored by TANDBERG

Re: POLL: for/against dropping support for kernels < 2.6.22

[Jonathan Johnson]

Hello all,

My vote is YES, why haven't we done this already??

My understanding is that we are just drop old kernel support and retaining the vast majority of the drivers.
If anyone tallied the total number CVE listed vulernabilites and other problems fixed since then they would probably be shocked.
Unless for some reason your hardware is so old that it does support 2.6.28.7, this is the version you should run.

Later,
Jonathan

>>> Hans Verkuil <hverkuil@xs4all.nl> 2/24/2009 2:19 PM >>>
Please reply to this poll if you haven't done so yet. I only count clear 
yes/no answers, so if you replied earlier to this in order to discuss a 
point then I haven't counted that.

Currently it's 16 yes and 2 no votes, but I'd really like to see some more 
input. I want to post the final results on Sunday.

Regards,

	Hans

On Sunday 22 February 2009 11:15:01 you wrote:
> Hi all,
>
> There are lot's of discussions, but it can be hard sometimes to actually
> determine someone's opinion.
>
> So here is a quick poll, please reply either to the list or directly to
> me with your yes/no answer and (optional but welcome) a short explanation
> to your standpoint. It doesn't matter if you are a user or developer, I'd
> like to see your opinion regardless.
>
> Please DO NOT reply to the replies, I'll summarize the results in a
> week's time and then we can discuss it further.
>
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> _: Yes
> _: No
>
> Optional question:
>
> Why:
>
>
>
> Thanks,
>
> 	Hans



-- 
Hans Verkuil - video4linux developer - sponsored by TANDBERG
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@vger.kernel.org 
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: POLL: for/against dropping support for kernels < 2.6.22

[Markus Rechberger]

On Tue, Feb 24, 2009 at 9:43 PM, Jonathan Johnson <johnsonn@matc.edu> wrote:
> Hello all,
>
> My vote is YES, why haven't we done this already??
>
> My understanding is that we are just drop old kernel support and retaining the vast majority of the drivers.
> If anyone tallied the total number CVE listed vulernabilites and other problems fixed since then they would probably be shocked.
> Unless for some reason your hardware is so old that it does support 2.6.28.7, this is the version you should run.
>
> Later,
> Jonathan
>
>>>> Hans Verkuil <hverkuil@xs4all.nl> 2/24/2009 2:19 PM >>>
> Please reply to this poll if you haven't done so yet. I only count clear
> yes/no answers, so if you replied earlier to this in order to discuss a
> point then I haven't counted that.
>
> Currently it's 16 yes and 2 no votes, but I'd really like to see some more
> input. I want to post the final results on Sunday.
>
> Regards,
>
>        Hans
>
> On Sunday 22 February 2009 11:15:01 you wrote:
>> Hi all,
>>
>> There are lot's of discussions, but it can be hard sometimes to actually
>> determine someone's opinion.
>>
>> So here is a quick poll, please reply either to the list or directly to
>> me with your yes/no answer and (optional but welcome) a short explanation
>> to your standpoint. It doesn't matter if you are a user or developer, I'd
>> like to see your opinion regardless.
>>
>> Please DO NOT reply to the replies, I'll summarize the results in a
>> week's time and then we can discuss it further.
>>
>> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>>
>> _: Yes
>> _: No
>>

don't care

>> Optional question:
>>
>> Why:
>>

analog TV didn't work for me anyway < 2.6.18 not sure if it got fixed up
timers are horribly modified within 2.6.18 - now, something that
worked back then might not work properly anymore now.
remote control support has a horrible bug, I reported something a half
year ago..
although the solution I use to provide is to move the frontend and
configuration layer to userland and only having the data transfer API
in the kernel (this includes v4l2 - latest API and the entire dvb core
API).
http://mcentral.de/wiki/index.php5/DVBConfigFramework it's infront of
the API access and the applications make use of it using LD_PRELOAD.
Customers mostly use to compile the em28xx driver from mcentral.de
against the currently installed drivers which come with the native
kernel, or I use to deliver binaries. Now the DVB config framework
makes use of usbfs (similar like libusbfs) in my case and also works
on OSX with exactly the same drivers.

Markus

>>
>>
>> Thanks,
>>
>>       Hans
>
>
>
> --
> Hans Verkuil - video4linux developer - sponsored by TANDBERG
> --
> To unsubscribe from this list: send the line "unsubscribe linux-media" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-media" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Re: POLL: for/against dropping support for kernels < 2.6.22

[Markus Rechberger]

On Wed, Feb 25, 2009 at 1:06 AM, Markus Rechberger
<mrechberger@gmail.com> wrote:
> On Tue, Feb 24, 2009 at 9:43 PM, Jonathan Johnson <johnsonn@matc.edu> wrote:
>> Hello all,
>>
>> My vote is YES, why haven't we done this already??
>>
>> My understanding is that we are just drop old kernel support and retaining the vast majority of the drivers.
>> If anyone tallied the total number CVE listed vulernabilites and other problems fixed since then they would probably be shocked.
>> Unless for some reason your hardware is so old that it does support 2.6.28.7, this is the version you should run.
>>
>> Later,
>> Jonathan
>>
>>>>> Hans Verkuil <hverkuil@xs4all.nl> 2/24/2009 2:19 PM >>>
>> Please reply to this poll if you haven't done so yet. I only count clear
>> yes/no answers, so if you replied earlier to this in order to discuss a
>> point then I haven't counted that.
>>
>> Currently it's 16 yes and 2 no votes, but I'd really like to see some more
>> input. I want to post the final results on Sunday.
>>
>> Regards,
>>
>>        Hans
>>
>> On Sunday 22 February 2009 11:15:01 you wrote:
>>> Hi all,
>>>
>>> There are lot's of discussions, but it can be hard sometimes to actually
>>> determine someone's opinion.
>>>
>>> So here is a quick poll, please reply either to the list or directly to
>>> me with your yes/no answer and (optional but welcome) a short explanation
>>> to your standpoint. It doesn't matter if you are a user or developer, I'd
>>> like to see your opinion regardless.
>>>
>>> Please DO NOT reply to the replies, I'll summarize the results in a
>>> week's time and then we can discuss it further.
>>>
>>> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>>>
>>> _: Yes
>>> _: No
>>>
>
> don't care
>
>>> Optional question:
>>>
>>> Why:
>>>
>
> analog TV didn't work for me anyway < 2.6.18 not sure if it got fixed up
> timers are horribly modified within 2.6.18 - now, something that
> worked back then might not work properly anymore now.
> remote control support has a horrible bug, I reported something a half
> year ago..
> although the solution I use to provide is to move the frontend and
> configuration layer to userland and only having the data transfer API
> in the kernel (this includes v4l2 - latest API and the entire dvb core
> API).
> http://mcentral.de/wiki/index.php5/DVBConfigFramework it's infront of
> the API access and the applications make use of it using LD_PRELOAD.
> Customers mostly use to compile the em28xx driver from mcentral.de
> against the currently installed drivers which come with the native
> kernel, or I use to deliver binaries. Now the DVB config framework
> makes use of usbfs (similar like libusbfs) in my case and also works
> on OSX with exactly the same drivers.
>

one more thing to add here,

the usb devices always have the same USB transfer, but different
configuration block.
there's libv4l2 which I think of inheriting within the userland daemon
since it also intercepts the calls.
A good interface would more or less be to add custom vendor ids and
custom product ids including custom
capabilities eg. enable transfer for analog TV, DVB-T, DVB-C, ATSC,
DMB-TH, ISDB-T...

Since I now have that framework which can more or less also work with
the existing em28xx kernel driver it might be an interesting question.
And since it's in userland the compatibility is also pretty good
(basically everything that supports dumping the data to userland for
v4l and the same for DVB) hybrid or non hybrid doesn't really matter
the daemon and lib will take care about this.

We also have libv4l2 in userland which knows about device specific
decoding algorithms, so there's already a kernel-userland dependency
here.

http://mcentral.de/wiki/index.php5/Terratec_HTC_XS

this is the first device which only relies on the data dump capability
of the em28xx driver, and overrides all the configuration issues (and
again there's no extra kernel module needed to override all this).
Basic dump capability of data exists since a very long time in the
v4l-dvb API - this is the main reason why backward compatibility is no
issue for me anymore.

http://mcentral.de/wiki/index.php5/DVBConfigFramework

interested people currently have access to the core userland API
already feel free to drop me a line if you want to know more about it.

regards,
Markus

Re: POLL: for/against dropping support for kernels < 2.6.22

[Simeon Simeonov]

Yes



      

Re: POLL: for/against dropping support for kernels < 2.6.22

[Rudy Zijlstra]

On Sun, 2009-02-22 at 11:15 +0100, Hans Verkuil wrote:
> Hi all,
> 
> There are lot's of discussions, but it can be hard sometimes to actually 
> determine someone's opinion.
> 
> So here is a quick poll, please reply either to the list or directly to me 
> with your yes/no answer and (optional but welcome) a short explanation to 
> your standpoint. It doesn't matter if you are a user or developer, I'd like 
> to see your opinion regardless.
> 
> Please DO NOT reply to the replies, I'll summarize the results in a week's 
> time and then we can discuss it further.
> 
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
> 
> _: Yes

YES

> _: No
> 
> Optional question:
> 
> Why:

>From what i see, i2c is causing trouble, also still in in 2.6.28. I
prefer attention on that in stead of trying to get the old i2c working. 

I've seen a remark seemed to imply that the Mythtv community is using
CentOS a lot. In my experience that is a minority in the mythtv group. 


-- 
Cheers,


Rudy

Re: POLL: for/against dropping support for kernels < 2.6.22

[Michael Krufky]

On Sun, Feb 22, 2009 at 5:15 AM, Hans Verkuil <hverkuil@xs4all.nl> wrote:
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> _: Yes
> _: No


NO.


> Optional question:
>
> Why:


Dropping support for older kernels means dropping support for MOST testers.

Sure, it's an inconvenience for the maintainers.  This does *not* have
to cause a hindrance for new drivers.  At first, new drivers can be
added to the repository, and set to require only the latest kernels,
via versions.txt .  When somebody has time to fix backwards compat for
that driver, simply update versions.txt with the new kernel version
dependency for the driver in question.


Additionally, we all know what upstream kernel development is like --
new kernel does *not* mean new stability.  More likely, new kernels
bring new bugs.  (this isnt always the case, but it's good to be
skeptical when it comes to production systems)

If I build an embedded system to use as a dedicated TV streaming box,
I will not want to update my kernel JUST so that I can use the new
driver required for my new TV tuner device.

Being able to build the v4l-dvb development repository against a
reasonable set of stable kernels, including kernels as old as 2.6.16,
is a critical feature for users of the v4l-dvb driver repository.

Regards,

Mike Krufky

Re: POLL: for/against dropping support for kernels < 2.6.22

[hermann pitton]

Hi,

Am Dienstag, den 24.02.2009, 16:02 -0500 schrieb Michael Krufky:
> On Sun, Feb 22, 2009 at 5:15 AM, Hans Verkuil <hverkuil@xs4all.nl> wrote:
> > Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
> >
> > _: Yes
> > _: No
> 
> 
> NO.
> 

this is an unwanted reply to Hans' polling, but I also stated previously
that I would leave it in the end to those who contributed and might
further do so. So this polling, for me, only means that neither Hans nor
Jean have to take care for how difficult backward compat would be for <
2.6.22.

Any common sense here?

> > Optional question:
> >
> > Why:
> 
> 
> Dropping support for older kernels means dropping support for MOST testers.

I seriously doubt this, I think I can count every single one reporting
issues below 2.6.22 and 2.6.18 during the last year, but it does not
even matter.

> Sure, it's an inconvenience for the maintainers.  This does *not* have
> to cause a hindrance for new drivers.  At first, new drivers can be
> added to the repository, and set to require only the latest kernels,
> via versions.txt .  When somebody has time to fix backwards compat for
> that driver, simply update versions.txt with the new kernel version
> dependency for the driver in question.

All agreed.

> Additionally, we all know what upstream kernel development is like --
> new kernel does *not* mean new stability.  More likely, new kernels
> bring new bugs.  (this isnt always the case, but it's good to be
> skeptical when it comes to production systems)

That is all true. But we start lacking testers on the recent rcx kernels
and unfortunately this includes me after years ...

> If I build an embedded system to use as a dedicated TV streaming box,
> I will not want to update my kernel JUST so that I can use the new
> driver required for my new TV tuner device.

Yes.

> Being able to build the v4l-dvb development repository against a
> reasonable set of stable kernels, including kernels as old as 2.6.16,
> is a critical feature for users of the v4l-dvb driver repository.

It is at least fun for them and we are great in that :)

Question is only, if Hans or Jean do to have to care for any of that
below 2.6.22 and I say no. Or?

> Regards,
> 
> Mike Krufky

Cheers,
Hermann

Re: POLL: for/against dropping support for kernels < 2.6.22

[Ant]

Hans Verkuil wrote:
> Hi all,
>
> There are lot's of discussions, but it can be hard sometimes to actually 
> determine someone's opinion.
>
> So here is a quick poll, please reply either to the list or directly to me 
> with your yes/no answer and (optional but welcome) a short explanation to 
> your standpoint. It doesn't matter if you are a user or developer, I'd like 
> to see your opinion regardless.
>
> Please DO NOT reply to the replies, I'll summarize the results in a week's 
> time and then we can discuss it further.
>
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> _: Yes
>   
Yes
> _: No
>
> Optional question:
>
> Why:
Firstly let me state that I am not a v4l developer. I have been lurking 
on this list and its predecessor for about 3 years as I find low level 
hardware programming very interesting. The main concern for the no camp 
seemed to be support for EL5. I use EL 3, 4 and 5 for different purposes 
to this day, and I would like to add my viewpoint. I know the older 
releases have inferior hardware support compared to the newer ones, but 
to me this is not a problem, just a consideration when selecting the 
hardware I wish to use. If v4l stops supporting kernels < 2.6.22 then it 
is not like EL5 based on 2.6.18 will instantly be useless. It just means 
that you will need to find a camera or dvr card that is already 
supported. I dont see that this is a problem, and think that this trade 
off is worth it so as to not complicate life for future development more 
than it needs to be.

Ant

Re: POLL: for/against dropping support for kernels < 2.6.22

[Simon Kenyon]

Hans Verkuil wrote:
> Hi all,
>
> There are lot's of discussions, but it can be hard sometimes to actually 
> determine someone's opinion.
>
> So here is a quick poll, please reply either to the list or directly to me 
> with your yes/no answer and (optional but welcome) a short explanation to 
> your standpoint. It doesn't matter if you are a user or developer, I'd like 
> to see your opinion regardless.
>
> Please DO NOT reply to the replies, I'll summarize the results in a week's 
> time and then we can discuss it further.
>
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> _: Yes
> _: No
>   
No
> Optional question:
>
> Why:
>
>   
i don't have a vote as i'm only a user and not a developer

but i thought i would just make one point

as far as i can see, the v4l-dvb tree exists to create support for a 
particular class of hardware within the linux kernel
the separate tree is very useful to lots of people (i include myself in 
that) - but it is a byproduct of the development methodology

so if you think this group's mission is to provide support for 
distributions then you should vote no
and if you think this group's mission is to provide support for the 
linux kernel then you should vote yes

>
> Thanks,
>
> 	Hans
>
>   

Re: POLL: for/against dropping support for kernels < 2.6.22

[Simon Kenyon]

Simon Kenyon wrote:
> Hans Verkuil wrote:
>> Hi all,
>>
>> There are lot's of discussions, but it can be hard sometimes to 
>> actually determine someone's opinion.
>>
>> So here is a quick poll, please reply either to the list or directly 
>> to me with your yes/no answer and (optional but welcome) a short 
>> explanation to your standpoint. It doesn't matter if you are a user 
>> or developer, I'd like to see your opinion regardless.
>>
>> Please DO NOT reply to the replies, I'll summarize the results in a 
>> week's time and then we can discuss it further.
>>
>> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>>
>> _: Yes
>> _: No
>>   
> No
i am an idiot - i read the question backwards

that really should be "YES"
i don't think old kernels should be maintained within the v4l-dvb tree

that is the job of the distributions - for which people pay money
let the distribution vendors pick that one up

>> Optional question:
>>
>> Why:
>>
>>   
> i don't have a vote as i'm only a user and not a developer
>
> but i thought i would just make one point
>
> as far as i can see, the v4l-dvb tree exists to create support for a 
> particular class of hardware within the linux kernel
> the separate tree is very useful to lots of people (i include myself 
> in that) - but it is a byproduct of the development methodology
>
> so if you think this group's mission is to provide support for 
> distributions then you should vote no
> and if you think this group's mission is to provide support for the 
> linux kernel then you should vote yes
>
>>
>> Thanks,
>>
>>     Hans
>>
>>   
>
> -- 
> To unsubscribe from this list: send the line "unsubscribe linux-media" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Re: POLL: for/against dropping support for kernels < 2.6.22

[wk]

> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> _: Yes
> _: No
>
>   
YES.

> Optional question:
>
> Why:
>   
I assume that the main goal should be development of linux v4l/dvb 
drivers to be included in *new* kernel versions. These dont need compat 
code.
But beside of the main goal there are requirements and other goals

- simplify development and save time (skip)
- keep code as easy as possible (skip)
- having as many testers as needed (don't skip or choose kernel version 
suitable)
- support of linux users who aren't able to update (either dont skip or 
provide backports in regular intervals. still easier to implement)


looking at this it will hurt only users from embedded hardwrae, but at 
least a bunch of them cannot compile modules anyway.
Might be solved by (i.e. yearly) backports.

Would be also interesting which kernel versions are used by list members.

Re: POLL: for/against dropping support for kernels < 2.6.22

[Jean-Francois Moine]

> Should we drop support for kernels <2.6.22 in our v4l-dvb
> repository?
>
> _: Yes
> _: No
>

Yes.

> Optional question:
>
> Why:
>

Most distros offer the lastest kernels. I had requests for old kernels,
but these ones were very old (2.6.10!).

-- 
Ken ar c'hentañ	|	      ** Breizh ha Linux atav! **
Jef		|		http://moinejf.free.fr/

Re: POLL: for/against dropping support for kernels < 2.6.22

[Hans Werner]

 
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> _: Yes
> _: No
>
>   

YES (but I would go even further)

> 
> Optional question:
>
> Why:
>   

The aim should be to bring improvements to the released Linux kernel.

The *only* point relevant for both development and testing is the current
latest development kernel, currently 2.6.29-rc6.

So I would go further : development should be moved to git and support
for all previous kernels should be dropped allowing concentration of
development resources on making patches which will be applied to the head 
of the git tree.

I think it is completely wrong-headed to have "v4l-dvb" as a thing which 
can be installed on top of old kernels to add new driver support to old
kernels. It is a waste of time to create such a thing, and a drain on 
resources to support.

As for users/testers, the message should be made crystal clear: if you
want to try running bleeding-edge code to get the latest support your
hardware the first thing to do is upgrade to the latest kernel. It will
be easier to communicate with developers who are (or should be!) working
on improving the latest development kernel.

This is better for everyone : time wasted on backporting and talking
about/debugging old kernel issues will be eliminated and drivers will 
released in the mainline kernels faster.

It will also clarify to distros and users where the "coal face" is:
new hardware support comes from new kernels, not v4l-dvb or (usually) 
backports or anything else.

Fixes for bugs in last stable kernel (currently 2.6.28.7) should be
pushed in if known, but never new features.

Distros or those with special commercial reasons can work on backports
if they really feel they can justify the use of their time, money and
other resources. They are also the only ones who can properly take account
of all the userland consequences of making a backport because they see
the whole system.

Regards,
Hans
-- 
Release early, release often.

Psssst! Schon vom neuen GMX MultiMessenger gehört? Der kann`s mit allen: http://www.gmx.net/de/go/multimessenger01

Re: POLL: for/against dropping support for kernels < 2.6.22

[Lars Hanisch]

> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
> 
> _: Yes
> _: No

  Yes.

> Why:

  I'm a v4l-user, I use my VDR for a couple of years now. These were the 
steps I took, before I assembled my box:

- I have analog cable, so what hardware does exist, that is capable to 
record video on an old PC (even my desktop had only a 400MHz Celeron)?
- Which of these pieces are supported by Linux?

  For me it ended up with a PVR150 and an DXR3, later replaced by a 
PVR350. I started with kernel 2.6.9, that time ivtv wasn't part of the 
kernel, it was even outside v4l-dvb (am I correct?). Without a large 
amount of help from the ivtv-lists and VDR forum, that would have been a 
disaster for me. I can't say how glad I was, when I read the news, that 
ivtv was integrated in the kernel.
  What I'm trying to say is: when you need support for hardware, you 
have to upgrade your kernel and there are many other people beside the 
main driver developer which can help you. In the "hot" time of 
integrating ivtv in the kernel, I back off asking Hans for supporting an 
older kernel, since all I wanted was a working driver. And if that means 
I have to upgrade the kernel, I just have to do it.

  I get paid for developing and maintaining some specialized desktop 
applications since ~15 years now (~200 users), and from that point of 
view, sometimes you have to drop support for older installations 
respectively have to upgrade those to some level, because it's just a 
pain. I can remember what a relief it was, to be able to drop support 
for Windows 98 and base my company's (rather complex and large) ERP-app 
on some "real" Windows (>= 2000). (right now we're right in the middle 
of porting from Win32/C++ to .Net3.5/C#, guess who will make a jig when 
it's done...)

  Reading the diverse postings and from my point of knowledge and 
experience, I think it's best to swap the development model to an "in 
kernel"-tree, that feeds a compat-tree, which supports kernel-versions 
that are reasonable. And if someone has fun backporting (i2c-related) 
drivers below 2.6.22, than let him do it. But let the main developer do 
their work in keeping uptodate with new hardware and new kernels. They 
get old soon enough. (the kernel, not the developers...) ;-)

Lars.

Re: POLL: for/against dropping support for kernels < 2.6.22

[Mike Isely]

On Sun, 22 Feb 2009, Hans Verkuil wrote:

> Hi all,
> 
> There are lot's of discussions, but it can be hard sometimes to actually 
> determine someone's opinion.
> 
> So here is a quick poll, please reply either to the list or directly to me 
> with your yes/no answer and (optional but welcome) a short explanation to 
> your standpoint. It doesn't matter if you are a user or developer, I'd like 
> to see your opinion regardless.
> 
> Please DO NOT reply to the replies, I'll summarize the results in a week's 
> time and then we can discuss it further.
> 
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
> 
> _: Yes
> _: No

Yes (see below)


> 
> Optional question:
> 
> Why:
> 

I'm always for backwards compatibility in general.  I have an 
out-of-tree "standalone" pvrusb2 driver which includes extra stuff that 
at least compiles correctly all the way back to 2.6.12 (extra - but old 
- i2c modules are also included with the driver for kernels of that 
vintage).

However, that's just my one driver and I think trying to maintain that 
sort of (in)sanity over the entire v4l-dvb tree is going to be a major 
morale-sucking headache.

I'm working right now on v4l2-subdev support and it's my intention that 
I will be ripping out all the old I2C adaptation stuff as part of this 
effort.  (I am actually going to at least try to make the old stuff 
still work as a compile-time switch in the standalone pvrusb2 driver but 
I don't realistically expect that to remain practical with the driver as 
it currently resides in v4l-dvb.)

So even if the decision is made to keep v4l-dvb as a whole compatible 
all the way back to 2.6.16, the pvrusb2 driver will still in the end 
have to be excluded in v4l-dvb builds for anything older than 2.6.22.  
I really can't vote "no" above with a straight face while doing this 
v4l2-subdev related work in the driver.

  -Mike

-- 

Mike Isely
isely @ pobox (dot) com
PGP: 03 54 43 4D 75 E5 CC 92 71 16 01 E2 B5 F5 C1 E8

Re: POLL: for/against dropping support for kernels < 2.6.22

[Magnus Damm]

On Sun, Feb 22, 2009 at 7:15 PM, Hans Verkuil <hverkuil@xs4all.nl> wrote:
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
>
> _: Yes
> _: No

Yes.

> Optional question:
>
> Why:

Focus on moving forward instead of looking backwards. Keeping user
space compatibilty is of course a good idea, but I see no reason why
V4L needs to be special compared to the rest of the kernel.

I don't have the full picture though and I'm not the one who spend
energy and time on keeping backwards compatibility. If the cost is
small enough then it may of course be worth it. But my gut feeling is
that there is no point in being special with these things, just do
like the rest of the kernel subsystems and you will be fine.

/ magnus

Re: POLL: for/against dropping support for kernels < 2.6.22

[Tobias Stoeber]

Hi all,

Hans Verkuil schrieb:
> So here is a quick poll, please reply either to the list or directly to me 
> with your yes/no answer and (optional but welcome) a short explanation to 
> your standpoint. It doesn't matter if you are a user or developer, I'd like 
> to see your opinion regardless.
> 
> Please DO NOT reply to the replies, I'll summarize the results in a week's 
> time and then we can discuss it further.
> 
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?
> 

X : Yes

> _: No
> 
> Optional question:
> 
> Why:

I am voting for "Yes" because ...

a) developers complain about technical reasons (changes to i2c etc.) 
which - in the words of the developers - make it harder to provide a 
backward compatibility. This is a valid reason.

b) I hope, that the resources freed will leed to a better v4l / dvb 
subsystem (and to surrounding tasks as, for instance, having 
documentation that is in line with the state of development etc.)

Maybe this also frees time, to really think about the direction of 
subsystem and whether it must be really depended that much to specific 
kernel versions, especially for devices like USB.

c) in the end: it would also not be of any value to insist on support 
for older kernels, because the users are dependend on developers and in 
some way in "their hands"  ...

d) I furthermore hope, that there will be snapshots (as archives) 
provided to allow for more recent device support to pre-2.6.22 kernel 
users (as provided by the kernel itself)

It should a least be stated (in the wiki) which devices are supported 
from what kernel version upwards, so users then can select - as it 
always has been - older hardware.

In the end, the process of having a poll including the question of "Why" 
gives an interesting view into how - at least some - developers and 
"users" see the facts and what hopes expactations exist.

In a way I am a bit shocked about some of the replies, which - just my 
personal opinion - placed ease and development speed above the users 
(and I don't have in mind those "users" here, who have a broad technical 
knowledge and resources to *really* know what they are doing when 
upgrading to recent or even git kernels!) of this "product" of development.

Regards, Tobias

Re: POLL: for/against dropping support for kernels < 2.6.22

[Adrian Pardini]

On 22/02/2009, Hans Verkuil <hverkuil@xs4all.nl> wrote:
>
> Should we drop support for kernels <2.6.22 in our v4l-dvb repository?

YES



> Optional question:
>
> Why:

I'm not going to repeat everything again, but i'd like to know that
most of the developer's time to code is used adding new features or
improving the existing ones instead of trying to make things work
under very old kernels.

If it helps, I'm mostly a user. Long ago I sent a small patch to get
my card working; also hacked a bit the driver to suit my needs.

cheers!

-- 
Adrian.
http://solar.org.ar