Skip to contentSkip to conversation

snd_ prefix for module options

DISCUSSION Takashi Iwai 9 participants 23 messages Oct 8, 2002 – Oct 12, 2002

lore
Share & tools
lore
#0 lore raw

Hi,

so far, all alsa modules use snd_ prefix for each module option. iirc, there was a problem regarding namespace at the time of 2.0 kernel, and this was some workaround to avoid confliction. but 2.2 and later kernels have no such a problem at all.

so, i'd like to ask you how do you think to remove this snd_ prefix. of course, there is one and only big problem - compatibility!

the questions are

- whether we should really do it or not? is it worthy?

and

- when? now or after 0.9.0-final release?

opinions, please?

Takashi

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#1 lore raw

On Tuesday 08 Oct 2002 15:04, Takashi Iwai wrote:
[snip]

Quoted 7 lines

so, i'd like to ask you how do you think to remove this snd_ prefix. of course, there is one and only big problem - compatibility!

the questions are

- whether we should really do it or not? is it worthy?

Yes - I find it confusing: it's not like everything else in the kernel.

Quoted 4 lines

and

- when? now or after 0.9.0-final release?

Now, before too many distros have picked up 0.9.0-final.

Quoted 4 lines

opinions, please?

IMHO :-)

-- Peter

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#2 lore raw
In reply to Takashi Iwai

I like it a lot. Due to the prefix I was able to cleanly remove 0.5.x from my system, and we might want to do that later on with 0.9, too :)

And since I don't see any benefit in removing the prefix - other than cosmetic - keeping it will save time for distribution's programmers, sysadmins, writers, etc. to update their setup routines, configs, howtos, etc. :)

Florian

Quoted 31 lines · Takashi Iwai

Hi,

so far, all alsa modules use snd_ prefix for each module option. iirc, there was a problem regarding namespace at the time of 2.0 kernel, and this was some workaround to avoid confliction. but 2.2 and later kernels have no such a problem at all.

so, i'd like to ask you how do you think to remove this snd_ prefix. of course, there is one and only big problem - compatibility!

the questions are

- whether we should really do it or not? is it worthy?

and

- when? now or after 0.9.0-final release?

opinions, please?

Takashi

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Alsa-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

Signature · 10 lines
Florian Bomers
Java Sound
Java Software/Sun Microsystems, Inc.
http://java.sun.com/products/java-media/sound/


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
#3 lore raw

Sorry guys, I thought you were talking about the snd_ prefix of the module NAMES :) for the option names, dropping seems the right thing to do - an error message should be printed when a snd_* option is used.

Florian

Quoted 57 lines · Florian Bomers

I like it a lot. Due to the prefix I was able to cleanly remove 0.5.x from my system, and we might want to do that later on with 0.9, too :)

And since I don't see any benefit in removing the prefix - other than cosmetic - keeping it will save time for distribution's programmers, sysadmins, writers, etc. to update their setup routines, configs, howtos, etc. :)

Florian

Quoted 23 lines · Takashi Iwai

Hi,

so far, all alsa modules use snd_ prefix for each module option. iirc, there was a problem regarding namespace at the time of 2.0 kernel, and this was some workaround to avoid confliction. but 2.2 and later kernels have no such a problem at all.

so, i'd like to ask you how do you think to remove this snd_ prefix. of course, there is one and only big problem - compatibility!

the questions are

- whether we should really do it or not? is it worthy?

and

- when? now or after 0.9.0-final release?

opinions, please?

Takashi

Quoted 8 lines

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Alsa-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

--
Florian Bomers
Java Sound
Java Software/Sun Microsystems, Inc.
http://java.sun.com/products/java-media/sound/

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Alsa-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

Signature · 10 lines
Florian Bomers
Java Sound
Java Software/Sun Microsystems, Inc.
http://java.sun.com/products/java-media/sound/


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
#4 lore raw
In reply to Takashi Iwai

--- Takashi Iwai <[email protected]> wrote: > Hi,

Quoted 4 lines

so, i'd like to ask you how do you think to remove this snd_ prefix.
of course, there is one and only big problem - compatibility!

I think it's a worthy goal. It's just noise in modules.conf, of course. Irritating, but no more than that. What concerns me more is when you compile ALSA directly into the kernel. To pass the options NOW requires a boot-time parameter list:

alsa-module-name=snd_this=x,snd_that=y,snd_the_other=z

Every "snd_" here wastes 4 precious bytes on the kernel's command line. Now I don't know what the maximum length of the command line is, but I'm sure I've reached it before. (And had the trailing options silently ignored.) I think economy is only polite with such a limited resource.

As for "when", I'd say that the longer you wait, the greater the pain. I have a fairly complicated setup, and the extent of my pain would be here:

options snd snd_cards_limit=2
options snd-card-0 snd_index=0
options snd-card-1 snd_index=1 snd_mpu_port=0x330

(Aren't PCI devices nice :-)?) I guess the worst hit would be people with ISA sound cards.

Chris

__________________________________________________
Do You Yahoo!?
Everything you'll ever need on one web page from News and Sport to Email and Music Charts
http://uk.my.yahoo.com

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven.
http://thinkgeek.com/sf

#5 lore raw
Quoted 6 lines · On Tue, 8 Oct 2002, [iso-8859-1] Chris Rankin wrote:

--- Takashi Iwai <[email protected]> wrote: > Hi,

Quoted 4 lines

so, i'd like to ask you how do you think to remove this snd_ prefix.
of course, there is one and only big problem - compatibility!

Quoted 7 lines

I think it's a worthy goal. It's just noise in modules.conf, of course. Irritating, but no more than that. What concerns me more is when you compile ALSA directly into the kernel. To pass the options NOW requires a boot-time parameter list:

alsa-module-name=snd_this=x,snd_that=y,snd_the_other=z

No, it should be: alsa-module-name=x,y,z . The prefix for alsa-module-name is required, because we have collisions with OSS drivers.

Jaroslav

-----
Jaroslav Kysela <[email protected]>
Linux Kernel Sound Maintainer
ALSA Project http://www.alsa-project.org
SuSE Linux http://www.suse.com

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#6 lore raw

At Wed, 9 Oct 2002 08:03:44 +0200 (CEST),

Quoted 15 lines · Jaroslav
Quoted 12 lines · On Tue, 8 Oct 2002, [iso-8859-1] Chris Rankin wrote:

--- Takashi Iwai <[email protected]> wrote: > Hi,

Quoted 4 lines

so, i'd like to ask you how do you think to remove this snd_ prefix.
of course, there is one and only big problem - compatibility!

I think it's a worthy goal. It's just noise in modules.conf, of course. Irritating, but no more than that. What concerns me more is when you compile ALSA directly into the kernel. To pass the options NOW requires a boot-time parameter list:

Quoted 4 lines
Quoted 1 line

alsa-module-name=snd_this=x,snd_that=y,snd_the_other=z

No, it should be: alsa-module-name=x,y,z . The prefix for alsa-module-name is required, because we have collisions with OSS drivers.

yes, the prefix for the module "name" is good, but not for the module "prefix", which is currently discussed.

Takashi

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#7 lore raw
In reply to Jaroslav Kysela

--- Jaroslav Kysela <[email protected]> wrote: > On Tue,
8

Quoted 1 line

alsa-module-name=snd_this=x,snd_that=y,snd_the_other=z

Quoted 5 lines

No, it should be: alsa-module-name=x,y,z . The
prefix for alsa-module-name
is required, because we have collisions with OSS
drivers.

I'm not talking about "snd" on the alsa-module-name, which is why I wrote it as "alsa-module-name" in the above example in the first place. I am referring to all the options "snd_index", "snd_enabled" etc.

If the option names are not required on the kernel command line, then can I assume that you can safely omit any that you don't want to specify by writing:

alsa-module-name=1,0,,,10

for example?

Chris

__________________________________________________
Do You Yahoo!?
Everything you'll ever need on one web page from News and Sport to Email and Music Charts
http://uk.my.yahoo.com

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven.
http://thinkgeek.com/sf

#8 lore raw
Quoted 10 lines · On Wed, 9 Oct 2002, [iso-8859-1] Chris Rankin wrote:

--- Jaroslav Kysela <[email protected]> wrote: > On Tue,
8

Quoted 1 line

alsa-module-name=snd_this=x,snd_that=y,snd_the_other=z

Quoted 5 lines

No, it should be: alsa-module-name=x,y,z . The
prefix for alsa-module-name
is required, because we have collisions with OSS
drivers.

Quoted 4 lines

I'm not talking about "snd" on the alsa-module-name, which is why I wrote it as "alsa-module-name" in the above example in the first place. I am referring to all the options "snd_index", "snd_enabled" etc.

Ok, the you distict between modules and kernel command line. Both have different syntax.

Quoted 2 lines

If the option names are not required on the kernel command line, then can I assume that you can safely

Replace required with allowed. You have to specify all parameters "in order".

Quoted 5 lines

omit any that you don't want to specify by writing:

alsa-module-name=1,0,,,10

for example?

It's the right example.

Jaroslav

-----
Jaroslav Kysela <[email protected]>
Linux Kernel Sound Maintainer
ALSA Project http://www.alsa-project.org
SuSE Linux http://www.suse.com

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#9 lore raw
In reply to Takashi Iwai
Quoted 6 lines · Takashi Iwai

so, i'd like to ask you how do you think to remove this snd_ prefix. of course, there is one and only big problem - compatibility!

the questions are

- whether we should really do it or not? is it worthy?

I'd say yes.

Quoted 1 line

- when? now or after 0.9.0-final release?

Better now than later, to minimize the number of installation which must be changed. But the change will invalidate all modules.conf files and every piece of documentation about it, so it should be announced loudly and well in advance (and many people won't notice anyway until their drivers refuse to load).

IMHO it would be nice if "make install" looks into modules.conf and complains loudly if old parameters are found. And maybe snd.o should detect the usage of old parameters, too, at least for some time.

Regards
Clemens

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#10 lore raw
In reply to Takashi Iwai
Quoted 7 lines

Hi,

so far, all alsa modules use snd_ prefix for each module option. iirc, there was a problem regarding namespace at the time of 2.0 kernel, and this was some workaround to avoid confliction. but 2.2 and later kernels have no such a problem at all.

Quoted 8 lines

so, i'd like to ask you how do you think to remove this snd_ prefix. of course, there is one and only big problem - compatibility!

the questions are

- whether we should really do it or not? is it worthy?

leave it like it is.
changing would break lots of users module.conf entries, which might be the fruits of hours of work.

Quoted 11 lines

and

- when? now or after 0.9.0-final release?

opinions, please?

Takashi

Quoted 8 lines

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Alsa-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

__________________________________________________________________

Gesendet von Yahoo! Mail - http://mail.yahoo.de
Yahoo! präsentiert als offizieller Sponsor das Fußball-Highlight des
Jahres: - http://www.FIFAworldcup.com

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven.
http://thinkgeek.com/sf

#11 lore raw

At Wed, 9 Oct 2002 12:15:17 +0200,

Quoted 12 lines · Karsten Wiese
Quoted 12 lines

Hi,

so far, all alsa modules use snd_ prefix for each module option. iirc, there was a problem regarding namespace at the time of 2.0 kernel, and this was some workaround to avoid confliction. but 2.2 and later kernels have no such a problem at all.

so, i'd like to ask you how do you think to remove this snd_ prefix. of course, there is one and only big problem - compatibility!

Quoted 7 lines
Quoted 4 lines

the questions are

- whether we should really do it or not? is it worthy?

leave it like it is.
changing would break lots of users module.conf entries, which might be the fruits of hours of work.

a convenient method is to check and rewrite /etc/modules.conf automatically when alsa-driver is installed. (btw, in the case of debian, do we need to check another path, too?)

i'm not sure whether it's good manner or not, though.

Takashi

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#12 lore raw
Quoted 2 lines

<snip> (proposal to remove snd_ prefix for each module option)

Quoted 3 lines · Takashi Iwai

a convenient method is to check and rewrite /etc/modules.conf automatically when alsa-driver is installed. (btw, in the case of debian, do we need to check another path, too?)

Yes.

I'm a recent (happy) convert to Debian 3.0, Woody. I'm working at learning it, but not an expert. I recently studied the packaging of ALSA, which is quite complex, and I believe this is correct...

Debian installs pieces of modules.conf in the /etc/modutils subdirectory on a per-package basis. The install scripts run "update-modules" to combine these files into /etc/modules.conf. In the case of ALSA, /etc/modutils/alsa is actually a symlink to either /etc/alsa/modutils/0.9 or /etc/alsa/modutils/0.5 since both ALSA versions are supported and the contents are incompatible.

Debian has a policy of handling user-editable configuration files *very* carefully. When installing an update, the package scripts are expected to preserve user edits and somehow deal with any format changes, too. When a package is removed, the config files are left in place so user changes will not be lost in case the package is re-installed in the future. To actually erase config files, one must use the --purge option. In general, every package is expected to be capable of upgrading both its previous version and the one before that.

These are far higher standards of compatibility than ALSA has so far tried to achieve. My impression is that the Debian maintainers have been working very hard to deal with ALSA's perennial instability.

Quoted 1 line

i'm not sure whether it's good manner or not, though.

Best to leave this alone. Anything else causes many people to hassle with fixing something that's not broken in the first place.

Regards,

Signature · 8 lines
  Jack O'Quin
  Austin, Texas, USA


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
#13 lore raw
In reply to Takashi Iwai
Quoted 6 lines · Takashi Iwai

(...)
a convenient method is to check and rewrite /etc/modules.conf automatically when alsa-driver is installed. (btw, in the case of debian, do we need to check another path, too?)

i'm not sure whether it's good manner or not, though.

I guess providing a script in the utils dir and pointing to it after compilation would be nice.

Florian

Quoted 11 lines

Takashi

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Alsa-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

Signature · 10 lines
Florian Bomers
Java Sound
Java Software/Sun Microsystems, Inc.
http://java.sun.com/products/java-media/sound/


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
#14 lore raw

At Wed, 09 Oct 2002 10:41:18 -0700,

Quoted 11 lines · Florian Bomers
Quoted 6 lines · Takashi Iwai

(...)
a convenient method is to check and rewrite /etc/modules.conf automatically when alsa-driver is installed. (btw, in the case of debian, do we need to check another path, too?)

i'm not sure whether it's good manner or not, though.

I guess providing a script in the utils dir and pointing to it after compilation would be nice.

I second it.

Takashi

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#15 lore raw
In reply to Takashi Iwai
Quoted 2 lines · Takashi Iwai

At Wed, 9 Oct 2002 12:15:17 +0200,

[snip]

Quoted 4 lines

a convenient method is to check and rewrite /etc/modules.conf automatically when alsa-driver is installed. (btw, in the case of debian, do we need to check another path, too?)

Yes.
/etc/modutils/alsa -> /etc/alsa/modutils/0.9
And then run update-modules.

This would normally be done when upgrading the .deb to the new version, so it shouldn't be something the driver or init.d file should worry about.

(My opinion only...)

Quoted 2 lines

i'm not sure whether it's good manner or not, though.

I think it's not so nice. Anyone running ALSA 0.9 should know it's not a "release" version and be happy to have (some breakage). If all that's needed is a quick edit of modules.conf, it shouldn't cause too many power-users grief. Those using distro upgrade tools should let their distro maintainer sort it out for them.

I don't think it's an ALSA developer problem.

Quoted 5 lines

Takashi

Quoted 8 lines

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Alsa-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#16 lore raw
Quoted 5 lines · Peter L Jones

I think it's not so nice. Anyone running ALSA 0.9 should know it's not a "release" version and be happy to have (some breakage). If all that's needed is a quick edit of modules.conf, it shouldn't cause too many power-users grief. Those using distro upgrade tools should let their distro maintainer sort it out for them.

This whole thread makes no sense to me at all.

Can someone please explain what terrible problem we're trying to solve that justifies introducing *any* breakage at all?

I don't mean to single out Peter for this one statement. But, I am totally frustrated with the attitude that ALSA is only for power users and that it's OK to introduce spurious incompatibilities on a whim.

The excuse that 0.9 is not a "release" is wearing thin these days. Since 0.5, which *is* a release, is "no longer supported", what are ordinary users expected to run? OSS?

ALSA is part of the 2.5 kernel now. It is mainstream Linux software, good technology, needed by many users. Isn't it about time to start thinking and acting that way?

Signature · 9 lines
  Jack O'Quin
  Austin, Texas, USA



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
#17 lore raw

On Thursday 10 Oct 2002 17:11, Jack O'Quin wrote:
[snip]

Quoted 11 lines

I don't mean to single out Peter for this one statement. But, I am totally frustrated with the attitude that ALSA is only for power users and that it's OK to introduce spurious incompatibilities on a whim.

The excuse that 0.9 is not a "release" is wearing thin these days. Since 0.5, which *is* a release, is "no longer supported", what are ordinary users expected to run? OSS?

ALSA is part of the 2.5 kernel now. It is mainstream Linux software, good technology, needed by many users. Isn't it about time to start thinking and acting that way?

Hi Jack - don't mind singling me out!

In principle, I'd agree entirely with you but

(a) ALSA 0.9.0-rc3 is not ALSA 0.9.0. Once it is, it will really be too late until ALSA 1.0.0

(b) Kernel 2.5.x is really a poor argument :-). Again, if ALSA has module parms begining snd_ in 2.6.x, then it's stuck with it for a long time.

(I'm hoping (a) == (b) myself.)

I agree, any breakage is bad breakage but waiting will make it worse and I think it should be done (should have been done earlier but it should still be done).

-- Peter

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#18 lore raw
In reply to Jack O'Quin
Quoted 5 lines · Jack O'Quin

(...)
I don't mean to single out Peter for this one statement. But, I am totally frustrated with the attitude that ALSA is only for power users and that it's OK to introduce spurious incompatibilities on a whim.

We'd all be happy if ALSA was already a finalized API and finalized implementation, established, proven, "mainstream", etc. But it's not! So if you want it or not, at the moment ALSA 0.9 is NOT meant for mainstream, because it IS still in development!

Quoted 3 lines

The excuse that 0.9 is not a "release" is wearing thin these days. Since 0.5, which *is* a release, is "no longer supported", what are ordinary users expected to run? OSS?

You claim any responsibility of the ALSA developers for linux users. They have not! This is open source. And I don't see what's wrong with using 0.5 until 0.9 is released.

Florian

Quoted 16 lines

ALSA is part of the 2.5 kernel now. It is mainstream Linux software, good technology, needed by many users. Isn't it about time to start thinking and acting that way?
--
Jack O'Quin
Austin, Texas, USA

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
_______________________________________________
Alsa-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/alsa-devel

Signature · 10 lines
Florian Bomers
Java Sound
Java Software/Sun Microsystems, Inc.
http://java.sun.com/products/java-media/sound/


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
#19 lore raw
In reply to Jack O'Quin

At 10 Oct 2002 11:11:06 -0500,

Quoted 10 lines · Jack O'Quin
Quoted 5 lines · Peter L Jones

I think it's not so nice. Anyone running ALSA 0.9 should know it's not a "release" version and be happy to have (some breakage). If all that's needed is a quick edit of modules.conf, it shouldn't cause too many power-users grief. Those using distro upgrade tools should let their distro maintainer sort it out for them.

Quoted 4 lines

This whole thread makes no sense to me at all.

Can someone please explain what terrible problem we're trying to solve that justifies introducing *any* breakage at all?

good to hear the opossite opinion, too.
the compatibility-breakage should be avoided, of course.

Quoted 11 lines

I don't mean to single out Peter for this one statement. But, I am totally frustrated with the attitude that ALSA is only for power users and that it's OK to introduce spurious incompatibilities on a whim.

The excuse that 0.9 is not a "release" is wearing thin these days. Since 0.5, which *is* a release, is "no longer supported", what are ordinary users expected to run? OSS?

ALSA is part of the 2.5 kernel now. It is mainstream Linux software, good technology, needed by many users. Isn't it about time to start thinking and acting that way?

this is the very reason why "now" i asked.
since the ALSA is now on the official kernel tree, we must follow to the standard rules, too. and the snd_ prefix is obviously ugly from this perspective.

if we live alone as external, then we don't have to care - we could do as we like. but the mainstream means that such exception has to be excluded, too, at least before 2.5 turns into the "real mainstream", i.e. 2.6 (or 3.0).

i guess the pain of this change wouldn't be so much once if there is a converter-script or if the distributor takes care of it. but, of course, my guess might be wrong, underestimated.

Takashi

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#20 lore raw
Quoted 8 lines
Quoted 6 lines · Jack O'Quin

Can someone please explain what terrible problem we're trying to solve that justifies introducing *any* breakage at all?

ALSA is part of the 2.5 kernel now. It is mainstream Linux software, good technology, needed by many users. Isn't it about time to start thinking and acting that way?

Quoted 9 lines · Takashi Iwai

this is the very reason why "now" i asked.
since the ALSA is now on the official kernel tree, we must follow to the standard rules, too. and the snd_ prefix is obviously ugly from this perspective.

if we live alone as external, then we don't have to care - we could do as we like. but the mainstream means that such exception has to be excluded, too, at least before 2.5 turns into the "real mainstream", i.e. 2.6 (or 3.0).

Are you saying that the "snd_" prefix violates standard Linux kernel rules for device driver options? Did the kernel developers request this change?

If so, I withdraw my objection.

I thought the change was merely cosmetic. :-)

Regards,

Signature · 9 lines
  Jack O'Quin
  Austin, Texas, USA



-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf
#21 lore raw
Quoted 16 lines · Jack O'Quin
Quoted 8 lines
Quoted 6 lines · Jack O'Quin

Can someone please explain what terrible problem we're trying to solve that justifies introducing *any* breakage at all?

ALSA is part of the 2.5 kernel now. It is mainstream Linux software, good technology, needed by many users. Isn't it about time to start thinking and acting that way?

Quoted 5 lines · Takashi Iwai

this is the very reason why "now" i asked.
since the ALSA is now on the official kernel tree, we must follow to the standard rules, too. and the snd_ prefix is obviously ugly from this perspective.

Quoted 8 lines
Quoted 4 lines

if we live alone as external, then we don't have to care - we could do as we like. but the mainstream means that such exception has to be excluded, too, at least before 2.5 turns into the "real mainstream", i.e. 2.6 (or 3.0).

Are you saying that the "snd_" prefix violates standard Linux kernel rules for device driver options? Did the kernel developers request this change?

Well, they had objections but not major to disallow inclusion of ALSA to 2.5. I think that it would be good to follow standard kernel rules in this case.

Jaroslav

-----
Jaroslav Kysela <[email protected]>
Linux Kernel Sound Maintainer
ALSA Project http://www.alsa-project.org
SuSE Linux http://www.suse.com

-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf

#22 lore raw
Quoted 6 lines · Jack O'Quin
Quoted 4 lines

Are you saying that the "snd_" prefix violates standard Linux kernel rules for device driver options? Did the kernel developers request this change?

Quoted 3 lines · Jaroslav Kysela

Well, they had objections but not major to disallow inclusion of ALSA to 2.5. I think that it would be good to follow standard kernel rules in this case.

I agree.

Signature · 8 lines
  Jack O'Quin
  Austin, Texas, USA


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf