Re: Has Copyright summarizing outlived its usefulness?
Quoting Markus Koschany : Why don't we add all DFSG-free licenses to /usr/share/common-licenses or /usr/share/free-licenses instead? It would save a lot of developer and maintenance time ... IMHO using links and references is just common sense and reduces unnecessary make work. +1 with "all DFSG-free licenses" == \ "all DFSG-free licenses used by at least 0x10 packages".
Re: Which files should go in ‘/usr/share/common-licenses/’? (was: Has Copyright summarizing outlived its usefulness?)
Quoting Ben Finney : If I understand correctly, the justification of putting a file there must include that it is overwhelmingly more likely to save *storage space* overall (by reducing the space in a corresponding number of ‘/usr/share/doc/…/copyright’ files), especially on machines that have low disk space in ‘/usr/share/’. Yes. This comes from times, when disk space was expensive. Times have changed for most systems, and those who work an small embedded systems (I do!) have brutal methods to solve the problem, mainly `rm -rf /usr/share/common-licenses/`.
I
U Sent from my iPhone
custom packages and schroot workflow
Hi, being new to the Debian schroot setup on Debian machines, I tried debugging some package. I found the crash happening in a library pulled as a runtime dependency. My idea was to recompile that library with some debug enabled and install those custom .deb's within the current schroot, to rerun the initial binary (with debug as well). Using dd-schroot-cmd -c $sessionid, I realized that this is limited to apt-get and not dpkg, and thus can't install .deb's not in the source.list ( https://dsa.debian.org/doc/schroot/ ) Jumping as root in the schroot is not possible too. Did I miss something ? Am I following the wrong workflow with Debian machines or generally speaking ? :) How do you work for this kind of issue? Thanks, F. pgp2uOSPbAyYS.pgp Description: PGP signature
Re: custom packages and schroot workflow
On 08/12/17 11:02, Frédéric Bonnard wrote: > Hi, > being new to the Debian schroot setup on Debian machines, I tried > debugging some package. I found the crash happening in a library pulled > as a runtime dependency. > My idea was to recompile that library with some debug enabled and install > those custom .deb's within the current schroot, to rerun the initial > binary (with debug as well). > Using dd-schroot-cmd -c $sessionid, I realized that this is limited to > apt-get and not dpkg, and thus can't install .deb's not in the > source.list ( https://dsa.debian.org/doc/schroot/ ) > Jumping as root in the schroot is not possible too. > Did I miss something ? > Am I following the wrong workflow with Debian machines or generally > speaking ? :) > How do you work for this kind of issue? What library is that? Does it not have a -dbg or -dbgsym package? Otherwise one option is to build the library and load it with LD_LIBRARY_PATH, that way you don't have to install random packages with dpkg (which is not allowed in porterboxes as you found). Cheers, Emilio
Re: custom packages and schroot workflow
Hi Emilio, On Fri, 8 Dec 2017 14:01:54 +0100, Emilio Pozuelo Monfort wrote: > On 08/12/17 11:02, Frédéric Bonnard wrote: > > Hi, > > being new to the Debian schroot setup on Debian machines, I tried > > debugging some package. I found the crash happening in a library pulled > > as a runtime dependency. > > My idea was to recompile that library with some debug enabled and install > > those custom .deb's within the current schroot, to rerun the initial > > binary (with debug as well). > > Using dd-schroot-cmd -c $sessionid, I realized that this is limited to > > apt-get and not dpkg, and thus can't install .deb's not in the > > source.list ( https://dsa.debian.org/doc/schroot/ ) > > Jumping as root in the schroot is not possible too. > > Did I miss something ? > > Am I following the wrong workflow with Debian machines or generally > > speaking ? :) > > How do you work for this kind of issue? > > What library is that? Does it not have a -dbg or -dbgsym package? luajit ; it has one, and maybe it is enough. But I wondered in case I need to add some other debug flags or want to instrument it (which I know, I'll do some day for some other package ) or just, simply, fix it and try it with the initial package. > Otherwise one option is to build the library and load it with LD_LIBRARY_PATH, > that way you don't have to install random packages with dpkg (which is not > allowed in porterboxes as you found). That's a useful hack I may use, right. Maybe limited though :-/ Thanks Emilio. At least, I know, I'm not missing some feature :) F. pgp4KEnfbC55a.pgp Description: PGP signature
Re: Which files should go in ‘/usr/share/common-licenses/’?
On Thu, Dec 7, 2017 at 11:18 PM, Russ Allbery wrote: > +1. I'd love some guidance on this. I'm not convinced that our current > Policy approach is best here. Let me ping about https://bugs.debian.org/859649 which is much simpler than the other CC bug it was (erroneously?) marked a duplicate of (because the other bug is partly blocked by there being multiple versions and variants of the CC licenses.). CC0-1.0 is now used by hundreds of Debian packages and most of the time it is not listed in the package's debian/copyright! https://codesearch.debian.net/search?q=CC0-1%5C.0&perpkg=1 Thanks, Jeremy Bicha
Re: Should tasks be considered harmful?
On Sun, 2017-12-03 at 23:43 +0100, Bjørn Mork wrote: > tl;dr: Desktop tasks have unexpected (from the user point of view) > side > effects due to dependencies. This can be considered harmful since the > installer task selection can easily can trick a user into installing > a > "substandard" system. I think the issue is we have a loose definition of "desktop", and expect task-gnome-desktop, task-lxde-desktop etc. to correspond to it. One solution might be firming up what our definition of "desktop" is, or at least, what minimum functionality should be provided by a task- desktop-* installation. If modemmanager/working 3G/3G configuration via GUI/[however this requirement is encoded] is considered part of that functionality, then the way forward would be one of a) adjust dependency chain so task-lxde-desktop preferred network- manager over wicd b) enhance wicd to manage 3G stuff c) add separate dependency on modemmanager (or whatever) | network- manager ... N) drop task-lxde-desktop (if none of the above could be achieved) However broadly I think this is tractable so I disagree with the premise that we need to dump desktop tasks.
Re: Should tasks be considered harmful?
On Sun, 2017-12-03 at 23:43 +0100, Bjørn Mork wrote: > tl;dr: Desktop tasks have unexpected (from the user point of view) > side > effects due to dependencies. This can be considered harmful since the > installer task selection can easily can trick a user into installing > a > "substandard" system. I think the issue is we have a loose definition of "desktop", and expect task-gnome-desktop, task-lxde-desktop etc. to correspond to it. One solution might be firming up what our definition of "desktop" is, or at least, what minimum functionality should be provided by a task- desktop-* installation. If modemmanager/working 3G/3G configuration via GUI/[however this requirement is encoded] is considered part of that functionality, then the way forward would be one of a) adjust dependency chain so task-lxde-desktop preferred network- manager over wicd b) enhance wicd to manage 3G stuff c) add separate dependency on modemmanager (or whatever) | network- manager ... N) drop task-lxde-desktop (if none of the above could be achieved) However broadly I think this is tractable so I disagree with the premise that we need to dump desktop tasks. -- Jonathan Dowland
Bug#883887: ITP: journalwatch -- Simple log monitoring utility for the systemd journal
Package: wnpp Severity: wishlist Owner: Ralf Jung * Package name: journalwatch Version : 1.1.0 Upstream Author : Florian Bruhin * URL : https://github.com/The-Compiler/journalwatch * License : GPL-3.0 Programming Lang: Python Description : Simple log monitoring utility for the systemd journal journalwatch regularily checks the systemd journal and reports all entries above a given priority via email. The tool supports a regexp-based per-service whitelist. It is similiar to tools like logwatch or logcheck, except it's much more KISS and only works with the systemd journal. In particular, since it can take the message priority into account, it needs much less of a whitelist. I want to package this tool because I am going to use it on my servers, and I think it may be useful for others as well. I am looking for a sponsor.
Bug#883898: ITP: golang-github-mwitkow-go-conntrack -- Go middleware for net.Conn tracking (Prometheus/trace)
Package: wnpp Severity: wishlist Owner: Martín Ferrari * Package name: golang-github-mwitkow-go-conntrack Version : 0.0~git20161129.cc309e4-1 Upstream Author : Michal Witkowski * URL : https://github.com/mwitkow/go-conntrack * License : Apache-2.0 Programming Lang: Go Description : Go middleware for net.Conn tracking Prometheus (https://prometheus.io/) monitoring and x/net/trace (https://godoc.org/golang.org/x/net/trace#EventLog) tracing wrappers for net.Conn, both inbound (net.Listener) and outbound (net.Dialer). . Go standard library does a great job of doing "the right" things with your connections: http.Transport pools outbound ones, and http.Server sets good Keep Alive defaults. However, it is still easy to get it wrong. . That's why you should be able to monitor (using Prometheus) how many connections your Go frontend servers have inbound, and how big are the connection pools to your backends. You should also be able to inspect your connection without ssh and netstat. This is a new dependency for prometheus 2.0
Bug#883900: ITP: golang-github-prometheus-tsdb --
Package: wnpp Severity: wishlist Owner: Martín Ferrari * Package name: golang-github-prometheus-tsdb Version : 0.0~git20171208.e103f21-1 Upstream Author : Prometheus * URL : https://github.com/prometheus/tsdb * License : Apache-2.0 Programming Lang: Go Description : Prometheus storage layer library This package contains the new Prometheus storage layer that will be used in its 2.0 release.
Re: Should tasks be considered harmful?
[ Sorry for the duplicate. So began, and I guess now finishes, my brief exploration of the Evolution MUA. ] -- ⢀⣴⠾⠻⢶⣦⠀ ⣾⠁⢠⠒⠀⣿⡁ Jonathan Dowland ⢿⡄⠘⠷⠚⠋⠀ https://jmtd.net ⠈⠳⣄ Please do not CC me, I am subscribed to the list.
