[OpenAFS-win32-devel] Re: Use of the Windows Registry
Jeffrey Altman
jaltman@columbia.edu
Tue, 09 Mar 2004 14:19:20 -0500
This is a multi-part message in MIME format.
--------------020609010303030908090001
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
Meghadri Ghosh wrote:
> Perhaps the afsdbmt.ini containing the drive mappings could be moved
> to be per user.
>
> I believe the current code could be maintained by simply create the
> following key: (The default value can be set to some suitable key,
> perhaps under 'Software\\OpenAFS.org\\Client'?)
>
> HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows
> NT\CurrentVersion\IniFileMapping\afsdbmt.ini]
> @="#USR:Software\\OpenAFS.org\\afsdbmt
>
> Thus per user mapping settings could be maintained.
Interesting. I will have to try this.
> By the way, I installed v1.2.10 on XP and noticed that it installs the
> loopback adapter and binds to it. This is good, but the MS loopback
> adapter was already installed on the system and a second instance was
> inserted by the installation. Additionally, on removing the new
> instance of the loopback adapter, and restarting the client service,
> the binding was to the (real) NIC instead of the existing loopback
> adapter.
The loopback adapter is found by its MAC address. If the binding was to
a real NIC it is because
the MAC address of the loopback adapter which is installed on your
machine is not the default
value. Perhaps that is why the second loopback adapter was installed.
In the future, OpenAFS will look for the adapter to bind to by name.
Loopback adapters (if
installed by the installer) will be assigned the name "AFS".
> Some time back I had emailed a URL that contained the diffs of a local
> version we have at Stanford that installed the loopback adaptor, if
> necessary, and if automatic binding was specificed, detected any
> existing instance of the MS loopback adaptor and bound to its LANA.
The proper way to submit patches is to use the OpenAFS Request Tracker
which is linked off of
the OpenAFS.ORG web page.
> The installer has been written with Installshield Developer 8. The
> code diffs and the install project can be found at
>
> http://www.stanford.edu/~meghadri/OpenAFS/
Install Shield is no longer being supported by OpenAFS.ORG. None of
those who have volunteered to build installers have licenses to the
Install Shield products any longer. Instead we are supporting
installers built with NSIS 2.0. Please see the 1.3.52 build. The
1.2.10 source tree is very
out of date and in my opinion is extremely unstable.
> Please let me know if there are any questions, or these could be
> otherwise modified to be amenable for useful integration into OpenAFS.
>
> --
> megh
>
>
>
> _______________________________________________
> OpenAFS-Win32-devel mailing list
> OpenAFS-Win32-devel@openafs.org
> http://lists.openafs.org/mailman/listinfo/openafs-win32-devel
--------------020609010303030908090001
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
<title></title>
</head>
<body bgcolor="#ffffff" text="#000000">
<font face="Bitstream Cyberbit">Meghadri Ghosh wrote:<br>
</font>
<blockquote cite="mid4043A9E7.6040905@stanford.edu" type="cite"><font
face="Bitstream Cyberbit">Perhaps the afsdbmt.ini containing the drive
mappings could be moved to be per user.
<br>
<br>
I believe the current code could be maintained by simply create the
following key: (The default value can be set to some suitable key,
perhaps under 'Software\\OpenAFS.org\\Client'?)
<br>
<br>
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows
NT\CurrentVersion\IniFileMapping\afsdbmt.ini]
<br>
@="#USR:Software\\OpenAFS.org\\afsdbmt
<br>
<br>
Thus per user mapping settings could be maintained.
<br>
</font></blockquote>
Interesting. I will have to try this.<br>
<br>
<blockquote cite="mid4043A9E7.6040905@stanford.edu" type="cite"><font
face="Bitstream Cyberbit">By the way, I installed v1.2.10 on XP and
noticed that it installs the loopback adapter and binds to it. This is
good, but the MS loopback adapter was already installed on the system
and a second instance was inserted by the installation. Additionally,
on removing the new instance of the loopback adapter, and restarting
the client service, the binding was to the (real) NIC instead of the
existing loopback adapter.
<br>
</font></blockquote>
The loopback adapter is found by its MAC address. If the binding was
to a real NIC it is because<br>
the MAC address of the loopback adapter which is installed on your
machine is not the default <br>
value. Perhaps that is why the second loopback adapter was installed.<br>
<br>
In the future, OpenAFS will look for the adapter to bind to by name.
Loopback adapters (if <br>
installed by the installer) will be assigned the name "AFS".<br>
<blockquote cite="mid4043A9E7.6040905@stanford.edu" type="cite"><font
face="Bitstream Cyberbit">Some time back I had emailed a URL that
contained the diffs of a local version we have at Stanford that
installed the loopback adaptor, if necessary, and if automatic binding
was specificed, detected any existing instance of the MS loopback
adaptor and bound to its LANA.
<br>
</font></blockquote>
The proper way to submit patches is to use the OpenAFS Request Tracker
which is linked off of <br>
the OpenAFS.ORG web page.<br>
<blockquote cite="mid4043A9E7.6040905@stanford.edu" type="cite"><font
face="Bitstream Cyberbit">The installer has been written with
Installshield Developer 8. The code diffs and the install project can
be found at
<br>
<br>
<a class="moz-txt-link-freetext" href="http://www.stanford.edu/~meghadri/OpenAFS/">http://www.stanford.edu/~meghadri/OpenAFS/</a>
<br>
</font></blockquote>
Install Shield is no longer being supported by OpenAFS.ORG. None of
those who have volunteered to build installers have licenses to the
Install Shield products any longer. Instead we are supporting<br>
installers built with NSIS 2.0. Please see the 1.3.52 build. The
1.2.10 source tree is very <br>
out of date and in my opinion is extremely unstable.<br>
<br>
<blockquote cite="mid4043A9E7.6040905@stanford.edu" type="cite"><font
face="Bitstream Cyberbit">Please let me know if there are any
questions, or these could be otherwise modified to be amenable for
useful integration into OpenAFS.
<br>
<br>
--
<br>
megh
<br>
<br>
<br>
<br>
_______________________________________________
<br>
OpenAFS-Win32-devel mailing list
<br>
<a class="moz-txt-link-abbreviated" href="mailto:OpenAFS-Win32-devel@openafs.org">OpenAFS-Win32-devel@openafs.org</a>
<br>
<a class="moz-txt-link-freetext" href="http://lists.openafs.org/mailman/listinfo/openafs-win32-devel">http://lists.openafs.org/mailman/listinfo/openafs-win32-devel</a>
<br>
</font>
</blockquote>
</body>
</html>
--------------020609010303030908090001--