Bug#974565: ITP: python-logassert -- Simple log assertion mechanism for Python unittests

[Inaki Malerba]

Package: wnpp
Severity: wishlist
Owner: Iñaki Malerba 
X-Debbugs-Cc: debian-de...@lists.debian.org

* Package name: python-logassert
  Version : 6
  Upstream Author : Facundo Batista 
* URL : https://github.com/facundobatista/logassert/
* License : GPL-3
  Programming Lang: Python
  Description : Simple log assertion mechanism for Python unittests

 python-logassert provides a simple way to test logging calls on
 a python package tests.

 This is a dependency needed to stop downloading test dependencies on
fades package (Bug#974556)

-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#970896: lintian times out while analyzing package

[Inaki Malerba]

Hey there!

Even though I couldn't reproduce it locally ( :/ ) it was 100%
reproducible on salsa ci.

Just retried a job with lintian==2.98.0 and it seems to be fixed :)

Same job from the original report:
https://salsa.debian.org/ina/mediawiki/-/jobs/1079414

Another failing job:
https://salsa.debian.org/ina/svxlink/-/jobs/1079386


Thanks a lot !

-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#961584: lxc-stop fails with exit code 1

[Inaki Malerba]

El 2/6/20 a las 20:07, Pierre-Elliott Bécue escribió:
> Could you remove the --quiet bit to see if lxc-stop gives us a bit more
> intel?
> 
> If possible, have your testbed call lxc-stop --kill --logpriority trace
> --logfile /dev/stdout --name ${NAME}, so we can have the most expressive
> output. But if it's an error, removing the quiet bit could be enough.
> 
> Thanks.
> 

The POC I attached on a previous email it's not using debci, so you can
modify the lxc-stop call.

After enabling trace log level, the output is the following:

> Starting LXC network bridge: :Starting LXC autoboot containers: :
> lxc-stop autopkgtest-stable-amd64 20200603085059.718 TRACEcommands - 
> commands.c:lxc_cmd_rsp_recv:123 - Command "get_init_pid" received response
> lxc-stop autopkgtest-stable-amd64 20200603085059.718 DEBUGcommands - 
> commands.c:lxc_cmd_rsp_recv:156 - Response data length for command 
> "get_init_pid" is 0
> lxc-stop autopkgtest-stable-amd64 20200603085059.718 TRACEcommands - 
> commands.c:lxc_cmd:293 - Opened new command socket connection fd 4 for 
> command "get_init_pid"
> lxc-stop autopkgtest-stable-amd64 20200603085059.719 TRACEcommands - 
> commands.c:lxc_cmd_rsp_recv:123 - Command "get_state" received response
> lxc-stop autopkgtest-stable-amd64 20200603085059.719 DEBUGcommands - 
> commands.c:lxc_cmd_rsp_recv:156 - Response data length for command 
> "get_state" is 0
> lxc-stop autopkgtest-stable-amd64 20200603085059.719 TRACEcommands - 
> commands.c:lxc_cmd:293 - Opened new command socket connection fd 4 for 
> command "get_state"
> lxc-stop autopkgtest-stable-amd64 20200603085059.719 DEBUGcommands - 
> commands.c:lxc_cmd_get_state:656 - Container "autopkgtest-stable-amd64" is in 
> "RUNNING" state
> lxc-stop autopkgtest-stable-amd64 20200603085059.719 TRACEcommands - 
> commands.c:lxc_cmd_rsp_recv:123 - Command "stop" received response
> lxc-stop autopkgtest-stable-amd64 20200603085059.719 DEBUGcommands - 
> commands.c:lxc_cmd_rsp_recv:156 - Response data length for command "stop" is 0
> lxc-stop autopkgtest-stable-amd64 20200603085059.719 TRACEcommands - 
> commands.c:lxc_cmd:293 - Opened new command socket connection fd 4 for 
> command "stop"
> lxc-stop autopkgtest-stable-amd64 20200603085059.720 ERRORcommands - 
> commands.c:lxc_cmd_stop:707 - No such file or directory - Failed to stop 
> container "autopkgtest-stable-amd64"
> lxc-stop: autopkgtest-stable-amd64: commands.c: lxc_cmd_stop: 707 No such 
> file or directory - Failed to stop container "autopkgtest-stable-amd64"

-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#961584: lxc-stop fails with exit code 1

[Inaki Malerba]

El 26/5/20 a las 23:35, Antonio Terceiro escribió:
> is there an easy/documented way of reproducing the salsa ci environment
> (i.e. lxc working under docker) locally? 

Attached is a Dockerfile. It should be sufficient to reproduce the problem.

Changing the FROM statement from debian:unstable to debian:testing is
enough to make it work. It needs to be run as privileged.

$ docker build -t autopkgtest - < Dockerfile
$ docker run --rm --privileged autopkgtest

-- 
- ina
FROM debian:unstable

ENV SALSA_CI_AUTOPKGTEST_LXC 
https://salsa.debian.org/salsa-ci-team/autopkgtest-lxc
ENV LXC_PATH /lxc

# Download and configure container.
RUN apt-get update && apt-get install -y wget
RUN wget --progress=dot:giga 
${SALSA_CI_AUTOPKGTEST_LXC}/-/jobs/artifacts/master/raw/artifacts/lxc.tar?job=stable
 -O lxc.tar
RUN mkdir ${LXC_PATH} && tar xf lxc.tar -C ${LXC_PATH}
RUN sed -i "/lxc.rootfs.path/ s@dir:.*/lxc/@dir:${LXC_PATH}/@" 
${LXC_PATH}/autopkgtest-stable-amd64/config

# Install lxc.
RUN apt-get update && apt-get install -y eatmydata
RUN eatmydata apt-get install -y lxc iptables

RUN echo "lxc.lxcpath=${LXC_PATH}" | tee -a /etc/lxc/lxc.conf
RUN echo 'USE_LXC_BRIDGE="true"' | tee /etc/default/lxc-net

RUN echo 'tmpfs /sys/fs/cgroup tmpfs rw,relatime,seclabel' | tee /etc/fstab
RUN echo 'cgroup /sys/fs/cgroup/cpuset cgroup rw,relatime,cpuset,x-mount.mkdir' 
| tee -a /etc/fstab
RUN echo 'cgroup /sys/fs/cgroup/devices cgroup 
rw,relatime,devices,x-mount.mkdir' | tee -a /etc/fstab

# This steps need to be run as privileged.
CMD umount -R /sys/fs/cgroup && mount -a && \
/etc/init.d/lxc-net start && \
/etc/init.d/lxc start && \
lxc-start autopkgtest-stable-amd64 && \
lxc-stop autopkgtest-stable-amd64 && \
echo "ok"


signature.asc
Description: OpenPGP digital signature

Bug#961584: lxc-stop fails with exit code 1

[Inaki Malerba]

Source: lxc
Version: 1:4.0.2-1
Severity: important

Dear Maintainer,

Since version 1:4.0.2-1, we've found a change on the behavior of
lxc-stop when running on the Salsa-CI pipeline.

debci calls `lxc-stop --quiet --kill --name $NAME` and it's returning
exit code 1.

This can be reproduced on salsa-ci pipeline, which calls `debci localtest`.

https://salsa.debian.org/salsa-ci-team/pipeline/-/jobs/765946

: failure: ['sudo', 'timeout', '600', 'lxc-stop',
'--quiet', '--kill', '--name', 'ci-147-3f089355'] failed (exit status 1,
stderr '')


-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#960389: ITP: python-configclass -- A Python class to hold configuration values

[Inaki Malerba]

Package: wnpp
Owner: Iñaki Malerba 
Severity: wishlist

* Package name: python-configclass
  Version : 0.2.0
  Upstream Author : Eduardo Naufel Schettino 
* URL : https://github.com/schettino72/configclass
* License : MIT
  Programming Lang: Python
  Description : A Python class to hold configuration values

Config is a dict where existing items can be modified but no items can be 
added. 
It has make() method so you can easily created derived configs where
duplicated values will be merged according to mergedict.ConfigDict.merge().

Needed as a dependency to close an RC bug in src:doit


-- - ina

-- 
- ina




signature.asc
Description: OpenPGP digital signature

Bug#960338: ITP: python-mergedict -- A Python dict with a merge() method

[Inaki Malerba]

Package: wnpp
Owner: Iñaki Malerba 
Severity: wishlist

* Package name: python-mergedict
  Version : 1.0.0
  Upstream Author : Eduardo Naufel Schettino 
* URL : https://github.com/schettino72/mergedict
* License : MIT
  Programming Lang: Python
  Description : A Python dict with a merge() method

A MergeDict is a dict with a merge() method. merge() is like
dict.update(), but it can be subclassed to create custom
"merge" operations based on the type of an item value.

The mergedict module comes with a ConfigDict that will
extend/update lists/sets/dicts.

Needed as a dependency to close an RC bug in src:doit


-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#954151: Please move logbook to Debian Python Modules team (Was: src:logbook: Requires a package outside of Main)

[Inaki Malerba]

El 14/4/20 a las 10:43, Andreas Tille escribió:
> Hi Agustin and Iñaki,
> 
> I wonder whether you would like to maintain the logbook package in
> Debian Python Modules team (by setting the mailing list as Maintainer
> and you two serve as Uploaders.)  This would possibly enhance the number
> of people who are watching this package and would simplify doing a team
> upload for the package.
> 
> Kind regards
> 
>   Andreas.
> 

Hi Andreas,

I'm having troubles to find time to maintain my packages lately, so I
think that's the best way to go.

I'll wait for Agustin's opinion, but it's ok for me!

Thanks!


-- 
- ina

Bug#950847: python-icecream needs new upstream for Python 3.8 support

[Inaki Malerba]

El 7/2/20 a las 12:37, Matthias Klose escribió:
> Package: src:python-icecream
> Version: 1.3.1-1
> Severity: serious
> Tags: sid bullseye
> User: debian-pyt...@lists.debian.org
> Usertags: python3.8
> 
> python-icecream needs new upstream for Python 3.8 support. The tests fail.
> 
> upstream changelog:
> 
> 
>  v2.0.0
> 
> Added: Support for Python 3.8.
> Removed: Support for Python 3.4.
> Changed: Switched core AST parsing engine to Alex Hall's executing
>   (https://github.com/alexmojaki/executing). Huge thank you to Alex Hall.
> Changed: Whitespace in arguments is no longer collapsed. Indentation in
>   multiline arguments is now preserved.
> 
> 
> from https://github.com/gruns/icecream/archive/v2.0.0.tar.gz
> 

Hi !

I'm waiting for #951438 to get accepted in order to be able to package
the new version.

Abrazos

-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#951438: ITP: python-executing -- Get information about what a frame is currently doing

[Inaki Malerba]

Package: wnpp
Severity: wishlist
Owner: Iñaki Malerba 

* Package name: python-executing
  Version : 0.4.1
  Upstream Author : Alex Hall 
* URL : https://github.com/alexmojaki/executing
* License : MIT
  Programming Lang: Python
  Description : Get information about what a frame is currently doing
 particularly the AST node being executed and it translate it into
 source code using asttokens.

 This is a new dependency for the release 2.0.0 of python-icecream.

-- 
- ina




signature.asc
Description: OpenPGP digital signature

Bug#896904: New upstream available

[Inaki Malerba]

Dear Maintainer,

This issue is still present. I tested version 2.2.4 from upstream and
works fine.

Please consider packaging the latest version.


Thanks :)

-- 
- ina




signature.asc
Description: OpenPGP digital signature

Bug#905581: lists.debian.org: Request for new mailing list: debian-salsa-ci

[Inaki Malerba]

Hi Alex,

Do you think it might me possible to open a 'salsa-ci-team' list now
that the project has gained more interest and we need a formal way to
communicate and discuss things?


Cheers

-- 
- ina




signature.asc
Description: OpenPGP digital signature

Bug#901000: [dh-python] Use current locale for encoding the output of readelf on so2pyver function

[Inaki Malerba]

Hi there!

Pinging about this issue.

I think the problem is still happening here
https://salsa.debian.org/debian/logbook/-/jobs/113627


Abrazo,

-- 
- ina

Bug#912578: [devscripts] Package is not installable on unstable

[Inaki Malerba]

Package: devscripts
Version: 2.18.7
Severity: grave

--- Please enter the report below this line. ---

Dear Maintainer,

The package is not installable on unstable due to dependencies problem.

```
The following packages have unmet dependencies:
 devscripts : Depends: libipc-run-perl but it is not going to be installed
  Depends: libmoo-perl but it is not going to be installed
  Depends: libwww-perl but it is not going to be installed
  Recommends: licensecheck but it is not going to be installed
  Recommends: lintian but it is not going to be installed
  Recommends: liblwp-protocol-https-perl but it is not going
to be installed
  Recommends: libsoap-lite-perl but it is not going to be
installed
E: Unable to correct problems, you have held broken packages.*
```*

The issue is reproducible building the following dockerfile:
```
FROM debian:unstable
RUN apt-get update && apt-cache policy devscripts && apt-get install
devscripts -y
```

The output is attached.


--- System information. ---
Architecture:
Kernel: Linux 4.17.0-2-amd64

Debian Release: buster/sid
500 unstable httpredir.debian.org
500 stretch download.docker.com
500 stable repository.spotify.com
500 jessie packagecloud.io

--- Output from package bug script ---

--- /etc/devscripts.conf ---

--- ~/.devscripts ---
Not present

-- 
- ina

Sending build context to Docker daemon  2.048kB
Step 1/2 : FROM debian:unstable
 ---> b291695bd6d4
Step 2/2 : RUN apt-get update && apt-cache policy devscripts && apt-get install 
devscripts -y
 ---> Running in 40561f614f4d
Get:1 http://cdn-fastly.deb.debian.org/debian unstable InRelease [233 kB]
Get:2 http://cdn-fastly.deb.debian.org/debian unstable/main amd64 Packages 
[8296 kB]
Fetched 8529 kB in 4s (2237 kB/s)
Reading package lists...
devscripts:
  Installed: (none)
  Candidate: 2.18.7
  Version table:
 2.18.7 500
500 http://deb.debian.org/debian unstable/main amd64 Packages
Reading package lists...
Building dependency tree...
Reading state information...
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:

The following packages have unmet dependencies:
 devscripts : Depends: libipc-run-perl but it is not going to be installed
  Depends: libmoo-perl but it is not going to be installed
  Depends: libwww-perl but it is not going to be installed
  Recommends: licensecheck but it is not going to be installed
  Recommends: lintian but it is not going to be installed
  Recommends: liblwp-protocol-https-perl but it is not going to be 
installed
  Recommends: libsoap-lite-perl but it is not going to be installed
E: Unable to correct problems, you have held broken packages.
The command '/bin/sh -c apt-get update && apt-cache policy devscripts && 
apt-get install devscripts -y' returned a non-zero code: 100


signature.asc
Description: OpenPGP digital signature

Bug#905514: Reproduced in Gitlab CI

[Inaki Malerba]

Dear maintainer, you can find the bug reproduced on the following
pipeline [1], adding our generic CI script [2].

1_ https://salsa.debian.org/ina-guest/cryptsetup/-/jobs/36815

2_ https://salsa.debian.org/salsa-ci-team/pipeline/


Saludos

-- 
- ina




signature.asc
Description: OpenPGP digital signature

Bug#904149: [python-gitlab] New upstram available

[Inaki Malerba]

Package: python-gitlab
Severity: normal

--- Please enter the report below this line. ---

Hello maintainer.

Would you mind packing the newest version ? 1.5.1 is already available.

Thanks!


---
System information. ---
Architecture:
Kernel: Linux 4.16.0-1-amd64

Debian Release: buster/sid
500 unstable httpredir.debian.org
500 stretch download.docker.com
500 stable repository.spotify.com
500 jessie packagecloud.io

--- Package information. ---
Package's Depends field is empty.

Package's Recommends field is empty.

Package's Suggests field is empty.

-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#855091: aiocoap: FTBFS randomly (failing tests)

[Inaki Malerba]

I won't help much with the problem, but just want to add some info about
this:

On 17/06/18 19:58, Santiago Vila wrote:
> Does this mean we should not worry about this anymore?
As we moved to salsa, CI pipeline was added which builds and tests the
package. Tests were disabled on build time because autopkgtest is used
to run them on a separated job.

You can check the results in here: 
https://salsa.debian.org/debian/aiocoap/pipelines/9851


Saludos!

-- 
- ina




signature.asc
Description: OpenPGP digital signature

Bug#901390: RFP: python-orderedmultidict -- an ordered multivalue dictionary that retains method parity with Python's dict.

[Inaki Malerba]

Edwads,

Thanks for your review.

The package was already accepted but I'll fix that now.
I packaged this as a dependency for furl, also form Gruns, which has
also the same license. I'll fix it there too.

Thanks !

Saludos.

On 14/06/18 04:43, Edward Betts wrote:
> I'd also like to see this package in Debian.
>
> Inaki Malerba  wrote:
>> License: Public Domain
> Just to be a little more specific python-orderedmultidict uses the Unlicense.
>
> The license says the software is released into the public domain.
>
> See https://github.com/gruns/orderedmultidict/blob/master/LICENSE.md
>
> And https://unlicense.org/

-- 
- ina




signature.asc
Description: OpenPGP digital signature

Bug#901390: RFP: python-orderedmultidict -- an ordered multivalue dictionary that retains method parity with Python's dict.

[Inaki Malerba]

Package: wnpp
Severity: wishlist
X-Debbugs-CC: debian-de...@lists.debian.org

--- Please fill out the fields below. ---

Package name: python-orderedmultidict
Version: 1.0
Upstream Author: Ansgar Grunseid 
URL: https://github.com/gruns/orderedmultidict
License: Public Domain
Description:
A multivalue dictionary is a dictionary that can store multiple values per
key. An ordered multivalue dictionary is a multivalue dictionary that
retains the order of insertions and deletions.

-- 
Martin Iñaki Malerba
inakimmale...@gmail.com | +54 02945 15468443
in...@satellogic.com

Bug#901070: RFP: python-furl -- URL parsing and manipulation made easy

[Inaki Malerba]

Package: wnpp
Severity: wishlist
X-Debbugs-CC: debian-de...@lists.debian.org

--- Please fill out the fields below. ---

Package name: python-furl
Version: v1.1
Upstream Author: Ansgar Grunseid 
URL: https://github.com/gruns/furl
License: Public Domain
Description:
Furl is a small Python library that makes parsing and manipulatin URLs
easy.
It's objects let you access and modify the various components of a URL.

||

-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#847693: v1.0 packaging

[Inaki Malerba]

Hi !

The 1.3 version is packaged.

Should we close this bug ?

-- 
- ina




signature.asc
Description: OpenPGP digital signature

Bug#900464: autopkg logbook

[Inaki Malerba]

Hi Paul!

Thanks for noticing !
We were aware of the problem, and were planning to fix it when releasing
the new upstream version (1.4.0).

I agree with the needs-root restriction. If you wanna see, the tests are
running on our pipeline anyway.

https://salsa.debian.org/debian/logbook/pipelines/6698


Saludos!

-- 
- ina




signature.asc
Description: OpenPGP digital signature

Bug#898739: RFP: python-icecream -- A simple python library which makes debugging a little easier.

[Inaki Malerba]

Package: wnpp
Severity: wishlist
X-Debbugs-CC: debian-de...@lists.debian.org

--- Please fill out the fields below. ---

Package name: python-icecream
Version: v1.3
Upstream Author: gruns
URL: https://github.com/gruns/icecream
License: MIT
Description: icecream is a small python library which makes print
debugging easier.
 With arguments, |ic| inspects itself and prints both its own arguments
and the values of those arguments. It can also print the file and line
number where it was called, among other things.

-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#747909: autopkgtest in docker on my pipeline.

[Inaki Malerba]

Hi all !

Here [1] is a link where you can see the docker virt working on my
pipeline.

Hope you can consider adding it :)


[1]_ https://salsa.debian.org/debian/doit/-/jobs/14389

-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#747909: WIP adt-virt-docker

[Inaki Malerba]

Hi Martin

I just brought to salsa [1] the Mathieu's patch with Chris fixes.

I'll do my best to try answering the questions you did.


1_ https://salsa.debian.org/ci-team/autopkgtest/merge_requests/3

>> +
>> +capabilities = ['revert-full-system', 'root-on-testbed',
>> +    'isolation-container']
>
>You need 'revert' here as well; revert-full-system only specifies a
>stricter behaviour of revert.

Fixed on new patch.

>> +    docker_container_id = VirtSubproc.check_exec(['docker', 'run',
'--detach=true',
>
>Is docker-run similar to lxc-attach in the guest, i. e. will that
>always run the command as root?
>
>Can you run adt-virt-docker as normal user, i. e. is it similar to LXC
>user containers? Or always just as root? In the latter case this
>should probably get a --sudo option like adt-virt-lxc, so that you
>don't need to run the entire adt-run as root, just the virt-server?

Fixed on new patch.

>
>> +    '--volume', "{0}:{0}".format(shared_dir)] + args.dockerargs
+ [args.image, 'sh', '-c', 'while true; do sleep 600; done'],
>
>This shell sleep loop looks ugly -- isn't there a command to just
>start a docker container and let it run?
>
>Style issues: Please use '%' instead of format() for code consistency
>(also in the other parts of the code below), and try to avoid such
>overly long lines.

This is fixed. The while+sleep command was turned into an sleep infinity.
While it doesn't look good either, its something necessary to keep the
container running on the background and then executing things in it via
`docker exec`.

Also, the style mismatch was fixed.

>
>So this command always starts the guest with an ephemeral overlay, the
>actual base image is never modified? OOI, what does that use to
>implement this? (tmpfs and overlayfs or similar?)

Exactly.
Base images are not modified when running them. Layers are added on top,
but, if not tagged, the modifications are dropped on exit.

>
>> +    outp=True)
>> +    adtlog.debug('hook_open: got docker container id %s' %
docker_container_id)
>> +    VirtSubproc.auxverb = [
>> +    'docker', 'exec', docker_container_id
>> +    ]
>
>What does this to the environment of the command in the docker
>container? a-v-lxc needs to do quite some extra effort to ensure that
>the guest's /etc/environment, /etc/default/locale, and /etc/profile
>are respected.
>

Docker uses all the locales and configs of the guest image. No problem
on that.

>> +    (status, out, err) = VirtSubproc.execute_timeout(None, 0,
VirtSubproc.auxverb + ['false'], stdout=subprocess.PIPE)
>> +    if status == 0:
>> +    # In Docker < 1.4, docker exec doesn't pass the return value
>> +    # We use nsenter which pass return value
>> +    # See
https://github.com/duglin/docker/commit/90928eb1140fc0394e2
>> +    adtlog.debug('hook_open: using nsenter workaround')
>> +    docker_container_pid = VirtSubproc.check_exec(['docker',
'inspect',
>> +    '--format', '{{.State.Pid}}', docker_container_id],
outp=True)
>> +    VirtSubproc.auxverb = [
>> +    'sudo', 'nsenter', '--target', docker_container_pid,
'--mount', '--uts', '--ipc', '--net',
>> +    '--pid', '--root', '--wd' # '--user'
>> +    ]
>
>Eww :) Given that jessie has docker 1.5, this code path would be
>exercised rather seldomly, and also couldn't easily be covered by
>automatic tests. Would you mind if this would just spit out an error
>message in the case that the exit code is not passed correctly?
>

This is not in the patch anymore.

>> +    (status, out, err) = VirtSubproc.execute_timeout(None, 0,
VirtSubproc.auxverb + ['apt-get', 'update'], stdout=subprocess.PIPE)
>
>Please drop this. It's not the virt-server's business to call apt-get
>update, it's a *very* expensive and also brittle operation (argh the
>dreaded "hash sum mismatch"), and might not even always work. This gets
>controlled by adt-run, possibly after modifying the guest's apt
>sources with --setup-command or --apt-pocket.

This is not in the patch anymore.
Anyway, apt-get update via setup-command is needed, because most (not to
say all) images come without sources cache.

>
>Do you have a concept of "Unix user" in the docker world, or does
>everything just run as root? In the former case, it might make sense
>to adapt the suggested-normal-user= auto-detection logic from LXC?

By default, it runs everything as root.

>
>> +def hook_downtmp(path):
>> +    global capabilities, shared_dir
>> +
>> +    if shared_dir:
>> +    d = os.path.join(shared_dir, 'downtmp')
>> +    # these permissions are ugly, but otherwise we can't clean
up files
>> +    # written by the testbed when running as user
>> +    VirtSubproc.check_exec(['mkdir', '-m', '777', d], downp=True)
>
>Whether this is necessary depends on the answer from above whether you
>can run docker as a user.
>
>> +    capabilities.append('downtmp-host=' + d)
>> +    else:
>> +    d = VirtSubproc.downtmp_mktemp(path)
>> +    return d

Bug#894217: [reprotest] kk_KZ.RK1048 again

[Inaki Malerba]

Hi Ximin
> More generally, I'd argue that build programs shouldn't fail simply when 
> LC_ALL is unrecognised, for example gcc works perfectly fine.
>
> $ LC_ALL=ououi gcc -c /dev/null 2>/dev/null; echo $?
> 0

Totally agree.
Normally, it should fallback to a known config, but, as you can see in
the following example, the problem it's not exactly that the LC_ALL is
unrecognized but the LC itself.
A similar problem happens with Sphinx.


Sorry if that's not what you meant.



```
# LC_ALL=kk_KZ.RK1048 help2man
Unknown encoding 'RK1048' at /usr/bin/help2man line 56.

# LC_ALL=bleble help2man
perl: warning: Setting locale failed.
perl: warning: Please check that your locale settings:
    LANGUAGE = (unset),
    LC_ALL = "bleble",
    LANG = (unset)
    are supported and installed on your system.
perl: warning: Falling back to the standard locale ("C").
`help2man' generates a man page out of `--help' and `--version' output.

Usage: help2man [OPTION]... EXECUTABLE

 -n, --name=STRING   description for the NAME paragraph
 -s, --section=SECTION   section number for manual page (1, 6, 8)
 -m, --manual=TEXT   name of manual (User Commands, ...)
 -S, --source=TEXT   source of program (FSF, Debian, ...)
 -L, --locale=STRING select locale (default "C")
 -i, --include=FILE  include material from `FILE'
 -I, --opt-include=FILE  include material from `FILE' if it exists
 -o, --output=FILE   send output to `FILE'
 -p, --info-page=TEXT    name of Texinfo manual
 -N, --no-info   suppress pointer to Texinfo manual
 -l, --libtool   exclude the `lt-' from the program name
 --help  print this help, then exit
 --version   print version number, then exit

EXECUTABLE should accept `--help' and `--version' options and produce
output on
stdout although alternatives may be specified using:

 -h, --help-option=STRING help option string
 -v, --version-option=STRING  version option string
 --version-string=STRING  version string
 --no-discard-stderr  include stderr when parsing option output

Report bugs to .
```

-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#894217: [reprotest] kk_KZ.RK1048 again

[Inaki Malerba]

Hi Chris!

Thanks for the answer.

On 17/04/18 13:18, Chris Lamb wrote:
> Could you expand the commit message on this patch to include the
> reason why beyond "causes trouble"?

I've done some research about the root problem but couldn't find much
more than it's a relatively new encoding.
My guess is that it's still not supported by some platforms. I found
some threads about its inclusion to python[1] and glibc[2].

I agree that it might seem like it was added on purpose, but rather than
checking for changes on the build, it makes the build tools fail.

If you know somebody that might know more about the nature of this
problem, or why it was chosen, would you cc him/her, please? Maybe
that's better than changing it.



1. https://bugs.python.org/issue22682
2. https://lists.debian.org/debian-glibc/2004/06/msg00080.html


Saludos!

-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#894217: [reprotest] kk_KZ.RK1048 again

[Inaki Malerba]

On 17/04/18 11:08, Inaki Malerba wrote:
> Should//I send a patch to change this encoding to a less problematic one?

Anyway, here's a proposal to change it to another encoding which I
already tested.


Saludos!

-- 
- ina

From ca7afd30754fa7fa3a794db70e00ce290fdd51fe Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?I=C3=B1aki=20Malerba?= <inaki@malerba.space>
Date: Tue, 17 Apr 2018 11:30:04 -0300
Subject: [PATCH] Change kk_KZ.RK1048 locale to vi_VN.UTF-8 Kazakh locale makes
 trouble on some langages like perl.
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Iñaki Malerba <inaki@malerba.space>
---
 reprotest/build.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/reprotest/build.py b/reprotest/build.py
index 966d2f8..d7da274 100644
--- a/reprotest/build.py
+++ b/reprotest/build.py
@@ -332,7 +332,7 @@ def locales(ctx, build, vary):
 else:
 # if there is an issue with this being random, we could instead select it
 # based on a deterministic hash of the inputs
-loc = random.choice(['fr_CH.UTF-8', 'es_ES', 'ru_RU.CP1251', 'kk_KZ.RK1048', 'zh_CN'])
+loc = random.choice(['fr_CH.UTF-8', 'es_ES', 'ru_RU.CP1251', 'vi_VN.UTF-8', 'zh_CN'])
 return build.add_env('LANG', loc).add_env('LC_ALL', loc).add_env('LANGUAGE', '%s:fr' % loc)
 
 def exec_path(ctx, build, vary):
-- 
2.17.0



signature.asc
Description: OpenPGP digital signature

Bug#894217: [reprotest] kk_KZ.RK1048 again

[Inaki Malerba]

Package: reprotest
Version: 0.7.7

--- Please enter the report below this line. ---

Hi there!

It's me again. I came across the same problem while checking another
package.
Sphinx breaks when the kk_KZ.RK1048 locale is chosen.

Should//I send a patch to change this encoding to a less problematic one?


Cheers!
//

```
# build the html pages for the -doc package
set -ex; \
    cd docs; \
    make html
+ cd docs
+ make html
make[2]�: on entre dans le repertoire
��/tmp/reprotest.eK3Bui/build-experiment-1/docs��
sphinx-build -b html -d _build/doctrees   . _build/html
Running Sphinx v1.7.2
making output directory...
loading pickled environment... not yet created
loading intersphinx inventory from
/usr/share/doc/python2.7/html/objects.inv...
building [mo]: targets for 0 po files that are out of date
building [html]: targets for 24 source files that are out of date
updating environment: 24 added, 0 changed, 0 removed
reading sources... [  4%]
api/base
Exception occurred:
  File
"/usr/lib/python2.7/dist-packages/sphinx/ext/autodoc/__init__.py", line
507, in get_sourcename
    sys.getfilesystemencoding(), 'replace')
TypeError: unicode() argument 2 must be string, not None
```


--- System information. ---
Architecture:
Kernel: Linux 4.15.0-2-amd64

Debian Release: buster/sid
500 unstable httpredir.debian.org
500 trusty ppa.launchpad.net
500 stretch download.docker.com
500 stable repository.spotify.com
500 jessie packagecloud.io

--- Package information. ---
Depends (Version) | Installed
-+-===
diffoscope | 93
python3-distro | 1.0.1-2
python3-rstr | 2.2.6-1
python3:any (>= 3.5~) |
python3-debian | 0.1.32
apt-utils | 1.6~rc1
libdpkg-perl | 1.19.0.5
procps | 2:3.3.14-1
python3-pkg-resources | 39.0.1-2


Recommends (Version) | Installed
==-+-===
diffutils | 1:3.6-1
OR diffoscope (>= 84) | 93
disorderfs (>= 0.5.2) |
faketime | 0.9.7-2
locales-all |
sudo | 1.8.21p2-3


Suggests (Version) | Installed
==-+-===
autodep8 | 0.11.1
schroot | 1.6.10-4
qemu-system |
qemu-utils |

-- 
- ina



signature.asc
Description: OpenPGP digital signature

Bug#894217: [reprotest] Problem with kk_KZ.RK1048 encoding.

[Inaki Malerba]

Package: reprotest
Version: 0.7.7
Severity: normal

--- Please enter the report below this line. ---

Hi there!

I've came across a problem triggered when the 'kk_KZ.RK1048' encoding is
chosen.
It's not a reprepro bug, but help2man's.

```
$ LC_ALL=kk_KZ.RK1048 help2man
Unknown encoding 'RK1048' at /usr/bin/help2man line 56.
```

I've already reported it to help2man
(https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=894126) but it seems
to be a perl problem.

As a workaround, i've forced LC_ALL to C on the debian/rules script, but
I think this is something to be taken into consideration.
The problem on this is that it's triggered randomly, when this encoding
is chosen
(https://github.com/ReproducibleBuilds/reprotest/blob/master/reprotest/build.py#L335)

The problem itself can be found on the following pipeline:
https://salsa.debian.org/debian/doit/-/jobs/10105


- Iñaki

--- System information. ---
Architecture:
Kernel: Linux 4.15.0-1-amd64

Debian Release: buster/sid
500 unstable deb.debian.org
500 stretch download.docker.com
500 stable repository.spotify.com
500 stable download.mono-project.com
500 stable dl.google.com
500 jessie packagecloud.io

--- Package information. ---
Package's Depends field is empty.

Package's Recommends field is empty.

Package's Suggests field is empty.

-- 
Martin Iñaki Malerba
inakimmale...@gmail.com
in...@satellogic.com

Bug#891076: docker.io: docker version (1.13.1) in sid reached EOL

[Inaki Malerba]

Hi all!

I came across this problem while trying to make explicit a docker
dependency.

Are there any news on the upgrade to the upstream version?
If not, i'd be willing to spend some time working on the task.

Cheers!

-- 
Iñaki