On Wed, 2024-05-15 at 10:58 +0100, Chris Lamb wrote:
> Ah, I foolishly didn't check back with the original test case. The
> root cause here, if I can call it that, is that we were calling "xz
> --list --verbose" and not specifying a second "--verbose".
That would explain it indeed.
> This has no
On Fri, 2024-05-10 at 14:49 +, Chris Lamb wrote:
> * Use "xz --list" to supplement the output when comparing .xz archives;
> essential when some underlying metadata differs. (Closes: #1069329)
> * Actually append the xz --list after the container differences, as it
> simplifies
Package: diffoscope
Severity: wishlist
X-Debbugs-Cc: Sam James
When comparing two XZ compressed files that decompress to identical
data, please compare the metadata before comparing compressed data.
The xz --list option can be used for this, from the manual page:
Print information about comp
Package: diffoscope
Version: 224
Severity: wishlist
It would be nice if diffoscope could help highlight that text data
differ only in the whitespace or if they differ in the text too.
The proposal would be that by default diffoscope would use wdiff (or
similar) to compare all line based text buff
Package: diffoscope
Version: 224
Severity: wishlist
It would be nice if diffoscope could help highlight that HTML files
differ in the text output or if they differ only in the non-text HTML
bytes like the page title, the stylesheet etc.
The proposal would be that by default diffoscope would conve
Package: diffoscope
Version: 224
Severity: wishlist
It would be nice if diffoscope could detect that XML files are only
different because of ordering differences, currently it cannot:
$ head -vn-0 *.xml
==> bar.xml <==
==> foo.xml <==
Package: diffoscope
Version: 224
Severity: wishlist
It would be nice if diffoscope could detect that text files are only
different because of ordering differences, currently it cannot:
$ (echo foo ; echo bar) > foo
$ (echo bar ; echo foo) > bar
$ diffoscope foo bar
--- foo
+++
On Thu, 2022-06-09 at 13:25 +0200, Mattia Rizzolo wrote:
> Oh, interesting. TIL.
The docs are relatively new, at one point we had a few folks who had
never contributed to Debian show up wanting to convert their company's
internal port of Debian for their company's CPU arch into an official
Debia
On Thu, 2022-06-09 at 11:00 +0100, Chris Lamb wrote:
> For some unknown reason, I thought that riscv64 *was* an official
> Debian architecture, and the tone of my previous emails should be
> read with that (incorrect) assumption in mind.
riscv64 is in the process of becoming an official port and
On Thu, 2022-05-26 at 14:00 +0100, Chris Lamb wrote:
> What I was trying to say is that we would need enough riscv64
> computing power to build all of Debian in good time... where "good
> time" is loosely defined as being able to keep up with the current
> architectures.
Here are the statistics o
Simon McVittie wrote:
> Relatedly, I would like to be able to capture some information about
> builds even if (perhaps especially if) the build fails.
That is a good point that I hadn't considered.
> so that failing builds can also produce artifacts, to help the
> maintainer and/or porters to fi
On Sun, 2022-02-13 at 14:13 -0800, Vagrant Cascadian wrote:
> * Split build metadata into a separate file or archive
>
> Some of the debian-installer packages generate tarballs that are not
> .deb files and are included in the .changes files when uploading to
> the archive; making a similar gener
Vagrant Cascadian wrote:
> Over the last several months, I and others have found quite a few
> packages that embed build paths via rpath when building with cmake.
This seems like the sort of thing that will be an ongoing problem, so
if it is detectable statically then a lintian warning might be g
Package: diffoscope
Version: 197
Severity: wishlist
I recently sponsored memtest86+ and used diffoscope to compare the
Debian source packages before uploading. The debian/ directory had been
subject to wrap-and-sort, which also affects the Architecture field.
The architecture list hadn't been alte
Package: diffoscope
Version: 190
Severity: normal
There are two issues with XML files not named *.xml:
They don't get reformatted before comparison, resulting in a diff of
the plain text, instead of a diff of the reformatted XML.
When comparing them with XML files named *.xml, a comparison of th
Package: diffoscope
Version: 177
Severity: wishlist
diffoscope currently prints a hex dump when comparing a TTF font with a
.ttx XML dump of a TTF font. If it used ttx (from fonttools) to convert
that TTF font to .ttx instead, then the comparison would become an XML
diff, which diffoscope treats a
Package: diffoscope
Version: 174
Severity: wishlist
When comparing directories and files that are essentially directories
serialised into a file (such as tar, ZIP, ISO etc), currently
diffoscope just says one is a file while one is a directory.
$ rm -rf ./* ; mkdir foo ; echo foo > foo/bar ; t
On Fri, Mar 5, 2021 at 8:26 AM Frédéric Pierret wrote:
> Is it possible to have an exported version somewhere for starting working on
> the mirroring? To not bothering you currently, just a one shot export until
> we are sure that would be absolutely necessary to have this daily export.
Please
On Thu, Mar 4, 2021 at 8:57 AM Frédéric Pierret wrote:
> Following discussion on #debian-reproducible, would this be feasible from the
> snapshot team to allow something like a daily export of the database on a
> public exposed server?
Looking on the database server, the raw PostgreSQL tables a
On Tue, 2021-03-02 at 10:58 +0100, Lucas Nussbaum wrote:
> - queries the snapshot DB to identify the files and hashes for each
> source package
> - fetches and analyses Sources files to identify (source,version) of
> interest, and thus hashes to transfer
You can probably replace this with the
On Mon, Mar 1, 2021 at 5:25 PM Holger Levsen wrote:
> > How would the mirroring work?
>
> to be discussed, but my raw idea would be to use rsync with excluding the
> years
> before 2015 or 2017. or can't this work? 8-)
That won't work, since the filesystem storing the data is hash (SHA1)
based,
On Sun, 2020-12-06 at 11:51 +0100, Mattia Rizzolo wrote:
> Note that the formal explenation for those usertags is at
> https://reproducible-builds.org/contribute/#how-to-report-bugs-in-debian
> and https://wiki.debian.org/ReproducibleBuilds/UserCategory (though
> I'm not sure anymore if they are
Hi all,
[Please CC me on any replies you would like me to read]
As part of my efforts to do some QA on Debian usertags, I noticed a few
typos and other issues with the reproducible builds Debian usertags,
which I fixed using a script and the attached commands to the bts bot.
https://salsa.debian
Package: diffoscope
Version: 150
Severity: normal
When comparing squashfs filesystems I get an error saying the first
filesystem could not be found, even though I created it just before.
$ mkdir foo bar
$ touch foo/foo bar/foo
$ chronic mksquashfs foo foo.squashfs
$ chronic mksquashfs bar bar.sq
On Mon, 2020-06-01 at 20:53 +0200, Mattia Rizzolo wrote:
> But so, do you think the first step lies in pbuilder/sbuild to decide an
> interface for the package maintainers to use, or for the buildd managers
> (wanna-build/buildd and launchpad) to decide on an interface that
> sbuild needs to provi
On Mon, 2020-06-01 at 11:12 +0200, Mattia Rizzolo wrote:
> I know this has been linked in multiple places, but I'll link it
> again:
> https://bugs.launchpad.net/launchpad/+bug/1845159
> there has also been a relevant thread in debian-devel a few months
> back, spawned off the debian/.build th
On Sun, 2020-05-17 at 13:12 +0200, Matthias Klose wrote:
> Any idea which package this should be filed for?
I think this is going to require a bunch of different changes in
various places including dpkg-dev, debhelper, reprepro, dak, launchpad
and any other apt repos that have an incoming system
Package: diffoscope
Version: 144
Severity: minor
When I compare the attached files (one unsigned buildinfo file, one
signed buildinfo file) I get a hex diff. When I strip away the
signature, the comparison is a text diff as normal. The optimal
comparison would be to compare the missing container w
On Sat, 2020-05-16 at 22:58 +, Chris Lamb wrote:
> My gut feeling is that this is the avenue we want to explore. Having a
> separate mechanism to capture this build-specific metadata would be an
> elegant solution and, as you imply, having the logs would have QA
> advantages as well as permit
On Mon, 24 Feb 2020 14:09:23 -0800 Vagrant Cascadian wrote:
> Exploring avenues to put files like this into some separate artifact for
> things that are not reproducible might be one avenue
There is already the BYHAND (and automatic BYHAND) mechanisms for files
that get installed outside of pool/
On Tue, 2018-11-27 at 03:11 -0500, Daniel Kahn Gillmor wrote:
> for machine-parseable stuff or automated processing, you should also
> probably supply --with-colons.
In this case we want human-readable output since this is for diffs.
--
bye,
pabs
https://wiki.debian.org/PaulWise
signature.a
On Mon, 2018-11-26 at 16:31 -0500, Chris Lamb wrote:
> I added support for pgpdump in diffoscope version 82 via #859034.
> Had you seen this?
I hadn't noticed that yet, thanks for that.
> I do note that your bug mentions some other tools (name gpg --list-
> packets, hot dump and hot dearmor) how
On Mon, 2018-11-26 at 16:58 -0500, Chris Lamb wrote:
> I think this could be made to work.
> As a demo, I just knocked together:
> ... which resulted in the attached.
Looks good.
> One difficulty however will be in passing in a sane value of `cols` to
> ConsoleDiff; not because it's difficult t
Package: diffoscope
Version: 101
Severity: wishlist
When comparing OpenPGP keys or signatures (binary and ASCII-armoured),
either as raw files or embedded in other formats (like .dsc files),
it would be nice to compare the internal data of those keys and sigs.
There are some different ways to out
Package: diffoscope
Version: 101
Severity: wishlist
diffoscope uses colour to highlight intra-line differences in the HTML
output but the terminal output only does a colordiff-style output. It
would be nice if the terminal mode could highlight intra-line changes.
The package ccdiff contains a too
Package: diffoscope
Version: 101
Severity: wishlist
The HTML output uses a side-by-side diff but the terminal output uses
`diff -u` style output. It would be nice to have an option for
side-by-side diffs with the terminal output. The icdiff package
contains a Python 3 module that could be used for
Package: diffoscope
Version: 101
Severity: normal
diffoscope currently diffs a hex dump of ASCII-armoured OpenPGP
upstream signature files instead of comparing the plain text.
The ASCII-armoured OpenPGP signature files are plain text containing
base64-encoded signatures so it makes more sense to
Package: qa.debian.org
Severity: wishlist
User: qa.debian@packages.debian.org
Usertags: bls
X-Debbugs-CC: reproducible-bui...@lists.alioth.debian.org, brl...@debian.org
Control: user reproducible-bui...@lists.alioth.debian.org
Control: usertags -1 + infrastructure
Control: clone -1 -2
Control:
On Sat, 2018-07-07 at 11:26 +0100, Chris Lamb wrote:
> Thanks, applied.
Attached a couple more patches.
--
bye,
pabs
https://wiki.debian.org/PaulWise
From 2c95bd3ec8e17c54fc2f8a1c4a1735d17df7ab55 Mon Sep 17 00:00:00 2001
From: Paul Wise
Date: Sat, 7 Jul 2018 16:31:24 +0800
Subject: [PATC
there.
I was thinking diffoscope &> output shell redirection.
> In that case, please could you provide a merge request?
I'm not interested in touching gitlab, patches attached.
--
bye,
pabs
https://wiki.debian.org/PaulWise
From 9fa755c81033148cc1a5a49bc4d2fe865ec6e0c8 Mon Sep 17
On Fri, 06 Jul 2018 18:50:44 +0100 Chris Lamb wrote:
> https://salsa.debian.org/reproducible-builds/diffoscope/commit/68c8263a430aca779b0c44fd31e729ae9bd59e0b.patch
> +# Clear the progress bar after completion (#901758)
I don't think mentioning the bug in the comment is useful he
Package: diffoscope
Version: 96
Severity: normal
Usertags: crash
diffoscope crashes when comparing ext4 filesystems:
$ truncate -s 1M foo.ext4.initial
$ /sbin/mkfs.ext4 foo.ext4.initial
mke2fs 1.44.2 (14-May-2018)
Filesystem too small for a journal
Discarding device blocks: done
Package: diffoscope
Version: 95
Severity: minor
diffoscope leaves a giant progress bar on the terminal after it
completes. It would be nice if the progress bar could disappear after
it has served its purpose. In case the progress bar code you use
doesn't support this, I've included code below from
Package: diffoscope
Version: 95
Severity: wishlist
It would be nice if comparing ZIP files with diffoscope could compare
the comments embedded in the CDFH and EOCD structures:
https://en.wikipedia.org/wiki/Zip_(file_format)#Central_directory_file_header
This can be done using both zipnote or zip
Package: qa.debian.org
Severity: wishlist
User: qa.debian@packages.debian.org
Usertags: bls
X-Debbugs-CC: reproducible-bui...@lists.alioth.debian.org, brl...@debian.org
Control: user reproducible-bui...@lists.alioth.debian.org
Control: usertags -1 + infrastructure
Control: clone -1 -2
Control:
On Sat, 2018-05-26 at 08:37 +0100, Chris Lamb wrote:
> Is this something you think you can drive or should someone take
> over..?
If someone wants to take over, please feel free.
Otherwise I'll get to it at some point.
--
bye,
pabs
https://wiki.debian.org/PaulWise
signature.asc
Description:
On Thu, 2018-04-26 at 09:08 +0200, Mattia Rizzolo wrote:
> * add a test
EDID data is almost always proprietary so I'm not sure how to do that.
> * expand that TODO item, I can't understand what it is about
diffoscope does not compare the output of the commands when one of them
returns an exit
5.2.2-1.3
Versions of packages diffoscope suggests:
ii libjs-jquery 3.2.1-1
-- no debconf information
--
bye,
pabs
https://wiki.debian.org/PaulWise
From 526a76729430897b4d5115f8b7d85efd26b094af Mon Sep 17 00:00:00 2001
From: Paul Wise
Date: Mon, 23 Apr 2018 13:54:35 +080
48 matches
Mail list logo
