[gentoo-user] triggered by backtracking - what does that mean?

[gentoo-user] triggered by backtracking - what does that mean?

[Helmut Jarausch]

Hi,
would anybody please explain what the following means:

emerge -vp net-libs/libpcap
Calculating dependencies   * waiting for lock on  
/var/db/.pkg.portage_lockfile ...                 [ ok ]
... done!
[ebuild   R    ] net-libs/libpcap-1.5.3  USE="bluetooth dbus ipv6  
-canusb -netlink -static-libs" 0 KiB

Total: 1 package (1 reinstall), Size of downloads: 0 KiB

!!! The following update(s) have been skipped due to unsatisfied  
dependencies
!!! triggered by backtracking:

net-libs/libpcap:0


Many thanks,
Helmut

Re: [gentoo-user] triggered by backtracking - what does that mean?

[Alan McKinnon]

On 24/07/2014 13:47, Helmut Jarausch wrote:
> Hi,
> would anybody please explain what the following means:
> 
> emerge -vp net-libs/libpcap
> Calculating dependencies   * waiting for lock on
> /var/db/.pkg.portage_lockfile ...                 [ ok ]
> ... done!
> [ebuild   R    ] net-libs/libpcap-1.5.3  USE="bluetooth dbus ipv6
> -canusb -netlink -static-libs" 0 KiB
> 
> Total: 1 package (1 reinstall), Size of downloads: 0 KiB
> 
> !!! The following update(s) have been skipped due to unsatisfied
> dependencies
> !!! triggered by backtracking:
> 
> net-libs/libpcap:0


It means you got tripped up by portage's New! Improved! Awesome!
internal invisible magic. Subslots started it all and portage has to
wade through tons of cruft to figure out the entire dependency tree. In
a nutshell, it keeps searching deeper and deeper until it finds an
answer that works, or until it hits a threshold. When it hits that
threshold, portage exits and says it went as far as it should and has
now given up.

A solution is to have portage search deeper:

emerge --backtrack=30

The default is 10 (usually plenty) but sometimes you need more.


-- 
Alan McKinnon
alan.mckinnon@gmail.com

Re: [gentoo-user] triggered by backtracking - what does that mean?

[Neil Bothwick]

[-- Attachment #1: Type: text/plain, Size: 909 bytes --]

On Fri, 25 Jul 2014 11:12:26 +0200, Alan McKinnon wrote:

> > !!! The following update(s) have been skipped due to unsatisfied
> > dependencies
> > !!! triggered by backtracking:
> > 
> > net-libs/libpcap:0  
> 
> 
> It means you got tripped up by portage's New! Improved! Awesome!
> internal invisible magic. Subslots started it all and portage has to
> wade through tons of cruft to figure out the entire dependency tree. In
> a nutshell, it keeps searching deeper and deeper until it finds an
> answer that works, or until it hits a threshold. When it hits that
> threshold, portage exits and says it went as far as it should and has
> now given up.

And in true portage tradition it tells you in a way that is factually
accurate yet totally uninformative unless you already understand what
happened.


-- 
Neil Bothwick

Death is proven to be 99.9% fatal to all laboratory rats.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 181 bytes --]

Re: [gentoo-user] triggered by backtracking - what does that mean?

[Alan McKinnon]

On 25/07/2014 11:54, Neil Bothwick wrote:
> On Fri, 25 Jul 2014 11:12:26 +0200, Alan McKinnon wrote:
> 
>>> !!! The following update(s) have been skipped due to unsatisfied
>>> dependencies
>>> !!! triggered by backtracking:
>>>
>>> net-libs/libpcap:0  
>>
>>
>> It means you got tripped up by portage's New! Improved! Awesome!
>> internal invisible magic. Subslots started it all and portage has to
>> wade through tons of cruft to figure out the entire dependency tree. In
>> a nutshell, it keeps searching deeper and deeper until it finds an
>> answer that works, or until it hits a threshold. When it hits that
>> threshold, portage exits and says it went as far as it should and has
>> now given up.
> 
> And in true portage tradition it tells you in a way that is factually
> accurate yet totally uninformative unless you already understand what
> happened.
> 
> 


portage has always suffered badly from Prime Error #1 in interface design:

1. Never expose the underlying implementation in the interface



-- 
Alan McKinnon
alan.mckinnon@gmail.com