TLD.rmk.(none) junk in BitKeeper logs where BK_HOST belongs?

TLD.rmk.(none) junk in BitKeeper logs where BK_HOST belongs?

[Matthias Andree]

Hello,

whose broken script or BitKeeper installation causes all these annoying
de.rmk.(none), au.rmk.(none) and all that to be logged instead of the
real BK_HOST?

The pattern seems to be the original TLD with 2nd level and below of the
domain stripped, plus .rmk.(none) appended.

I don't care to know who it is but would the offending system please be
updated or fixed?

Thanks in advance. Sample, 20 random incidents, one with full log:

ChangeSet@1.1608.65.2, 2004-03-05 20:41:31+00:00, laforge@org.rmk.(none)
  [SERIAL] Fix supprot for AFAVLAB 8port boards in 2.6.x
  
  I didn't yet use one of my AFAVLAB boards with 2.6.x until now.  The
  upper 4 ports are not detected at all.  I suppose the bug was
  introduced while porting the driver from 2.4.x.
  
  Please consider applying the following patch.  It also adds support
  for a new 8 port board called P030.

ChangeSet@1.1608.65.1, 2004-03-05 20:16:43+00:00, bjorn.helgaas@com.rmk.(none)
ChangeSet@1.1608.60.3, 2004-03-04 22:17:06+00:00, dsaxena@net.rmk.(none)
ChangeSet@1.1608.60.2, 2004-03-04 22:13:54+00:00, icampbell@com.rmk.(none)
ChangeSet@1.1608.60.1, 2004-03-04 22:13:37+00:00, dsaxena@net.rmk.(none)
ChangeSet@1.1608.19.1, 2004-02-26 12:15:46+00:00, armcc2000@com.rmk.(none)
ChangeSet@1.1588.4.4, 2004-02-22 17:16:19+00:00, mail@de.rmk.(none)
ChangeSet@1.1588.2.3, 2004-02-21 22:44:43+00:00, bjorn.helgaas@com.rmk.(none)
ChangeSet@1.1588.2.2, 2004-02-21 22:39:59+00:00, bjorn.helgaas@com.rmk.(none)
ChangeSet@1.1588.2.1, 2004-02-21 14:54:36+00:00, mark@net.rmk.(none)
ChangeSet@1.1557.70.117, 2004-02-20 19:24:06+00:00, dsaxena@net.rmk.(none)
ChangeSet@1.1557.86.1, 2004-02-20 16:52:31+00:00, h.schurig@de.rmk.(none)
ChangeSet@1.1557.70.116, 2004-02-20 10:09:26+00:00, tony@com.rmk.(none)
ChangeSet@1.1557.70.115, 2004-02-19 12:47:07+00:00, linux@de.rmk.(none2)
ChangeSet@1.1500.3.4, 2004-01-27 22:11:07+00:00, dirk.behme@com.rmk.(none)
ChangeSet@1.1500.3.3, 2004-01-27 22:07:35+00:00, fb.arm@net.rmk.(none)
ChangeSet@1.1500.3.2, 2004-01-27 22:04:31+00:00, fb.arm@net.rmk.(none)
ChangeSet@1.1500.3.1, 2004-01-27 22:01:27+00:00, fb.arm@net.rmk.(none)
ChangeSet@1.1474.101.10, 2004-01-20 22:26:42+00:00, fb.arm@net.rmk.(none)
ChangeSet@1.1474.101.9, 2004-01-20 22:24:14+00:00, nico@org.rmk.(none)

-- 
Matthias Andree

Encrypt your mail: my GnuPG key ID is 0x052E7D95

Re: TLD.rmk.(none) junk in BitKeeper logs where BK_HOST belongs?

[Russell King]

On Tue, Mar 16, 2004 at 07:44:56PM +0100, Matthias Andree wrote:
> whose broken script or BitKeeper installation causes all these annoying
> de.rmk.(none), au.rmk.(none) and all that to be logged instead of the
> real BK_HOST?

I do it purposely.  Go read:

 http://www.informationcommissioner.gov.uk/

and read through the (flash) bits.  Now consider "are email addresses
information which can be used to identify individuals?"  The answer is
"yes".  Are we storing that in a kind of database?  Yes.  Therefore,
does this fall under the terms of the Data Protection Act?  Yes.

Therefore I myself do not want to store peoples email addresses BK,
thereby avoiding this issue entirely.

> I don't care to know who it is but would the offending system please be
> updated or fixed?

Nope - live with it.

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 PCMCIA      - http://pcmcia.arm.linux.org.uk/
                 2.6 Serial core

Re: TLD.rmk.(none) junk in BitKeeper logs where BK_HOST belongs?

[David Woodhouse]

On Tue, 2004-03-16 at 19:30 +0000, Russell King wrote:
> and read through the (flash) bits.  Now consider "are email addresses
> information which can be used to identify individuals?"  The answer is
> "yes".  Are we storing that in a kind of database?  Yes.  Therefore,
> does this fall under the terms of the Data Protection Act?  Yes.

You already have my email address in your copy of the BK tree. You
achieve nothing by simply not adding _more_ valid email addresses.

I think civil disobedience in this case is appropriate.

-- 
dwmw2

Re: TLD.rmk.(none) junk in BitKeeper logs where BK_HOST belongs?

[Matthias Andree]

On Tue, 16 Mar 2004, David Woodhouse wrote:

> On Tue, 2004-03-16 at 19:30 +0000, Russell King wrote:
> > and read through the (flash) bits.  Now consider "are email addresses
> > information which can be used to identify individuals?"  The answer is
> > "yes".  Are we storing that in a kind of database?  Yes.  Therefore,
> > does this fall under the terms of the Data Protection Act?  Yes.
> 
> You already have my email address in your copy of the BK tree. You
> achieve nothing by simply not adding _more_ valid email addresses.
> 
> I think civil disobedience in this case is appropriate.

Why would it? I see legal ways to provide such information - and after
all, we want the authors' names rather than their addresses in the frist
place, at least in "shortlog".

-- 
Matthias Andree

Encrypt your mail: my GnuPG key ID is 0x052E7D95

Re: TLD.rmk.(none) junk in BitKeeper logs where BK_HOST belongs?

[Matthias Andree]

On Tue, 16 Mar 2004, Linus Torvalds wrote:

> He does it on purpose. Apparently there is some UK law that may make it 
> illegal to export other peoples email addresses without express consent, 
> so rmk corrupts them with a script..

Two notes:
1.
This could be handled by only including patches of those people who
consent to their address being published,

2.
The user does not have to give a routable mail address in
BK_USER/BK_HOST, but he can set BK_HOST to whatever he wants.


If the whole corruption is intentional, then I'd suggest that RMK
participates in the maintenance of the lk-changelog.pl aka. shortlog
script. I have no chance to resolve common names through
google/lbdb/grep -r on the suspect source files unless the address is
there. There are so many people called Jonas Larsson - how do I know if
that fellow has a middle name?

-- 
Matthias Andree

Encrypt your mail: my GnuPG key ID is 0x052E7D95

Re: TLD.rmk.(none) junk in BitKeeper logs where BK_HOST belongs?

[Russell King]

On Tue, Mar 16, 2004 at 08:41:54PM +0100, Matthias Andree wrote:
> On Tue, 16 Mar 2004, Linus Torvalds wrote:
> > He does it on purpose. Apparently there is some UK law that may make it 
> > illegal to export other peoples email addresses without express consent, 
> > so rmk corrupts them with a script..
> 
> Two notes:
> 1.
> This could be handled by only including patches of those people who
> consent to their address being published,

This requires me to keep a database of peoples addresses who have
consented.  No thanks, that's a huge overhead and waste of time.

> 2.
> The user does not have to give a routable mail address in
> BK_USER/BK_HOST, but he can set BK_HOST to whatever he wants.

Indeed, so I set BK_HOST to something else.

> If the whole corruption is intentional, then I'd suggest that RMK
> participates in the maintenance of the lk-changelog.pl aka. shortlog
> script.

Again, no thanks, I'm already busy enough as it is.

> I have no chance to resolve common names through
> google/lbdb/grep -r on the suspect source files unless the address is
> there. There are so many people called Jonas Larsson - how do I know if
> that fellow has a middle name?

Have a look in the changesets themselves - you'll find the line "Patch
from" in them, so you can pick out the real persons name from that -
even automatically via a suitable regexp.

Sorry for being so unco-operative on this issue, but I'm doing my best
already.

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 PCMCIA      - http://pcmcia.arm.linux.org.uk/
                 2.6 Serial core

Re: TLD.rmk.(none) junk in BitKeeper logs where BK_HOST belongs?

[Matthias Andree]

On Tue, 16 Mar 2004, Russell King wrote:

> > Two notes:
> > 1.
> > This could be handled by only including patches of those people who
> > consent to their address being published,
> 
> This requires me to keep a database of peoples addresses who have
> consented.  No thanks, that's a huge overhead and waste of time.

"vacation" style recipes for maildrop or procmail do most of the parts
you'd need - sort people's mail into one folder "no-consent", if they
send a patch, ask them for permission, if they consent, drop their mail
address into the DB so their mail is sorted into
"address-export-allowed". Change names as you will.

Just a suggestion to counted the "huge" overhead. It is overhead and
costs time without doubt.

> Have a look in the changesets themselves - you'll find the line "Patch
> from" in them, so you can pick out the real persons name from that -
> even automatically via a suitable regexp.

I'll try.

-- 
Matthias Andree

Encrypt your mail: my GnuPG key ID is 0x052E7D95