From: Pavel Roskin <proski@gnu.org>
To: Faidon Liambotis <paravoid@debian.org>
Cc: linux-wireless@vger.kernel.org, hermes@gibson.dropbear.id.au,
Jouni Malinen <j@w1.fi>
Subject: Re: [PATCH 2.6.23 3/3] [wireless] orinoco: create a Kconfig option for Prism2
Date: Wed, 25 Jul 2007 12:32:24 -0400 [thread overview]
Message-ID: <1185381144.3236.12.camel@dv> (raw)
In-Reply-To: <46A6EB79.5040403@debian.org>
On Wed, 2007-07-25 at 09:19 +0300, Faidon Liambotis wrote:
> Pavel Roskin wrote:
> > I'm sorry, but considering your original patch, I just cannot be sure
> > that you will get the PCMCIA IDs right. I cannot be sure I'll get it
> > right myself. It requires a lot of searching and detective work.
> We can always begin with the set that is common to both (PCMCIA) drivers.
> You're proposing to do nothing but at the same time you say that HostAP
> claims it can support some cards when it actually doesn't, which is a
> rather important bug IMHO.
> I understand that you're maintaining orinoco and not hostap but this is
> an issue that affects your "users" too, not to mention the overall
> quality of the kernel.
I understand, but I don't have all possible cards, and I depend on what
users are reporting. In absence of reports, I prefer to work on things
I can actually test.
> My patch affected a douzine of PCMCIA IDs, don't you think we can safely
> correct these?
0156:0002 alone is going to be a big headache. Basically, you would
need to replace it with textual IDs for all Lucent cards and remove the
numeric ID.
> > On the other hand, I haven't heard many complains about the ID clash
> > recently. It seems to me that users learned how to deal with it.
> > Distributions do a great job too. For instance, Fedora renames network
> > devices based on the MAC addresses, so the same configuration will work
> > with either orinoco of hostap.
> >
> > Of course, those who want to run an 802.11b AP know that they should
> > choose hostap, but most users don't need that.
> You haven't heard many complaints recently because there aren't many
> users recently...
>
> And this is not about network interface names or AP mode.
> You may disagree, but IMO HostAP is a *much* better driver for Prism2
> devices in all modes.
Well, I agree that there are many things in hostap that are not in
orinoco, such as firmware download and WPA.
> And if you have both drivers compiled as modules (as most distributions
> do) you have to either blacklist orinoco or manually unbind the driver
> from the hardware using /sys.
The later is true for the drivers compiled into the kernel as well.
> I'm maintaining hostap-utils for Debian and we are shipping a blacklist
> for orinoco for years because of the numerous reports of users who
> weren't be able to use HostAP.
> This is suboptimal though, since it breaks (= no driver loaded) when the
> user actually has a Lucent Orinoco card.
I agree, that's suboptimal.
> I understand your reluctance but I think it's way past the time you
> should have passed Prism2 to HostAP.
I actually hoped that some userspace solution would appear that would
make it unnecessary to handle it in the drive.
--
Regards,
Pavel Roskin
next prev parent reply other threads:[~2007-07-25 16:32 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-07-22 13:17 [PATCH 2.6.23 3/3] [wireless] orinoco: create a Kconfig option for Prism2 Faidon Liambotis
2007-07-22 17:14 ` Pavel Roskin
2007-07-23 22:07 ` Faidon Liambotis
2007-07-25 5:30 ` Pavel Roskin
2007-07-25 6:19 ` Faidon Liambotis
2007-07-25 16:32 ` Pavel Roskin [this message]
2007-08-01 3:33 ` Jouni Malinen
2007-08-01 5:55 ` Pavel Roskin
2007-07-23 1:42 ` David Gibson
2007-07-23 6:10 ` Pavel Roskin
2007-07-23 6:20 ` David Gibson
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1185381144.3236.12.camel@dv \
--to=proski@gnu.org \
--cc=hermes@gibson.dropbear.id.au \
--cc=j@w1.fi \
--cc=linux-wireless@vger.kernel.org \
--cc=paravoid@debian.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.