From megver83 at hyperbola.info Mon Jun 1 02:09:40 2020 From: megver83 at hyperbola.info (Megver83) Date: Sun, 31 May 2020 22:09:40 -0400 Subject: [Dev] [RFC] Let's solve the Chromium freedom issues, definitively In-Reply-To: <20200531072502.4904ec58@primarylaptop.localdomain> References: <4dcf3ee7-c081-94d5-0a0d-e07475b97d04@hyperbola.info> <20200529214310.0eab61d7@parabola> <20200531072502.4904ec58@primarylaptop.localdomain> Message-ID: <9b6b1c3c-0beb-a269-3a2e-7a6c6e166fa6@hyperbola.info> El 31-05-20 a las 01:25, Denis 'GNUtoo' Carikli escribi?: > As I understand the issue, if we end up being obligated to rely on > Chromium to get more and more things working, it would would mean that > we would be accepting Google's design decisions, and patching around > them would probably require a (constant) amount of work that is way > bigger than the constant amount of work that would be needed to fix > Gecko based browsers. > Now that you mention this, it sounds much better. The thing with Chromium is that there are every time more programs and platforms adopting it (Android Webview, Qt through webengine, electron, etc.). Maybe right now it isn't too critical, yet, but that might change in the future, and we might have to sacrifice many packages (like we are doing now with some qt5-webengine dependant packages, which are mostly KDE software). However, we have to first evaluate if it's more feasible to free Chromium or free its "derivatives" separately (Android Webview and qt5-webengine for instance, which, at first look, seems to be easier) Regarding the qt5-webengine specific case, we could ask their devs if they can consider moving it from Chromium web renderer (Blink) to Gecko. Otherwise, we should do a research to find the problems to be solved with it, and eventually patch it. Hopefully its devs will help us to find any licensing issues, at least, or solve any doubt we have if we encounter a potential freedom issue. -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From megver83 at hyperbola.info Wed Jun 3 01:01:19 2020 From: megver83 at hyperbola.info (Megver83) Date: Tue, 2 Jun 2020 21:01:19 -0400 Subject: [Dev] [RFC] Let's solve the Chromium freedom issues, definitively In-Reply-To: <5827880e567ead1d967d5336010972de@cock.email> References: <4dcf3ee7-c081-94d5-0a0d-e07475b97d04@hyperbola.info> <5827880e567ead1d967d5336010972de@cock.email> Message-ID: you forgot to send this to the mailing list El 02-06-20 a las 20:03, eliotreyna at cock.email escribi?: > Honestly, I'm using Chromium from Arch Linux (installed in my netbook > due to my wifi card that requires a blob to work), and it doesn't > include Widevine support by default. Actually the only one widely > avariable function in Chromium is the Media Source Extension feature, > but EME support is strictly limited to Widevine in the stable branch. > > Exist a community of Chromium builds (which is focused mainly on > Windows) that are really dedicated to offer vay editions of Chromium, > since "Chrome-like" until "ungoogled". > > The website is here >> https://chromium.woolyss.com/ > > On 2020-05-29 19:03, Megver83 wrote: >> I think that one of the greatest problems with this is that users and >> devs of FSDG distros are not joining forces, or better say: we are not >> doing a coordinated effort to address this topic, but I think we all >> know that. Coordinating efforts is in fact easier that how it sounds. >> Someone (me? when I've the time, I'm super busy now, but just wanted to >> share the idea) could create a GitLab/Gitea/Gogs repo (called >> "chromium-fsdg-research" for example, and using sth. like Gitea give us >> the ability to report issues in such project) with text files with all >> of what we know about the issues related to Chromium, possible >> solutions, and what is needed to solve this from once of all. >> >> Here I'd link (as a start point): >> * useful Parabola issues >> * useful Debian/PureOS issues >> * FSF pages >> * Chromium documentation >> * *concrete known issues* (e.g. the non-free widevine extension[0]) with >> references and elaborated explanation (not just "license is unclear" or >> simplistic stuff) >> >> And one of the most important things: planning. My proposal for a plan >> is to 1) collect and centralize all the needed info we can in this git >> repo, 2) report the known issues in this repo's issue tracker, so we can >> track the specific issues one by one, and look for a way to solve them, >> for which we need to 3) invite experienced free software contributors >> and programmers to actively participate in this project (looking for >> issues and solving them), or anyone with the will and experience to help. >> >> The purpose with this is to create a distro-agnostic project to solve >> the Chromium problem, so then each FSDG distro will implement the >> solutions in their specific ways. I think that ppl from Parabola, PureOS >> and the FSF/GNU project will be the most interested and active ones >> (Trisquel is developed by one person only, and idk if Hyperbola devs may >> even care about this, since they are very strict with their browsers' >> security and Chromium isn't the most secure one) >> >> I'd like to know what do you think about this, and if you have another >> idea or if you would change sth. of my proposal. This is just a draft of >> an idea, if wanted to go more in detail I'd say, for example, that the >> requirements for reporting an issue is to put the corresponding >> references and proofs of that is being reported, or whatever. But first, >> I want some feedback from the community, to see if there's some interest >> in this, at least in Parabola :) >> >> [0] >> https://salsa.debian.org/chromium-team/chromium/-/blob/master/debian/README.debian#L32 >> >> >> >> _______________________________________________ >> Dev mailing list >> Dev at lists.parabola.nu >> https://lists.parabola.nu/mailman/listinfo/dev -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From bill-auger at peers.community Wed Jun 3 20:29:57 2020 From: bill-auger at peers.community (bill-auger) Date: Wed, 3 Jun 2020 16:29:57 -0400 Subject: [Dev] [RFC] Let's solve the Chromium freedom issues, definitively In-Reply-To: References: <4dcf3ee7-c081-94d5-0a0d-e07475b97d04@hyperbola.info> <5827880e567ead1d967d5336010972de@cock.email> Message-ID: <20200603162957.6cd5511b@parabola> On Tue, 2 Jun 2020 21:01:19 -0400 Megver83 wrote: > you forgot to send this to the mailing list > > El 02-06-20 a las 20:03, eliotreyna escribi?: > > Honestly, I'm using Chromium from Arch Linux (installed in > > my netbook due to my wifi card that requires a blob to > > work), that immediate (perceived) problem may be solved with a single command: # pacman -U https://www.parabola.nu/packages/libre/x86_64/iceweasel/download/ it is the general procedure for installing foreign packages - alternatively, one could have run parabola, and installed the blob packages and/or chromium - the difference would be an arch system with a few parabola packages, vs a parabola system with a few arch packages; but either configuration should be functional, albeit unsupported, because parabola promises to be compatible with arch all of #1167 is mainly a workload concern, of far less of importance to users; because ideally, every parabola user would know these essential things: * hardware support does not prevent anyone from running a parabola system; though it is a common notion - almost any individual arch package, including the kernel and hardware blobs, could be installed onto a parabola system - no one is compelled to run a complete arch system, if only to gain hardware support * for any software with a maintained PKGBUILD, the primary value is in the maintenance of the PKGBUILD - if the PKGBUILD is well-maintained, then anyone should be able to compile and install it easily - the value of the binary packages in repos, is mainly in the conveniences of having them pre-compiled, and having octopi managing them automatically * if every parabola user becomes familiar with makepkg and a few pacman tricks, then most packaging decisions are rarely impose more than a matter of minor conveniences lost or gained upon users From grizzlyuser at protonmail.com Wed Jun 3 21:04:29 2020 From: grizzlyuser at protonmail.com (grizzlyuser) Date: Wed, 03 Jun 2020 21:04:29 +0000 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update Message-ID: Please find change descriptions in commit messages from the attached patches. Please apply iceweasel-branding-77.0-1.patch to [1]. Built for x86_64 and i686. armv7h build fails with message: fatal error: 'cmath' file not found As Arch Linux ARM have already built their firefox package of the latest version, I suppose this error should resolve by itself after the next sync of the repos. [1] https://git.parabola.nu/packages/iceweasel.git/ -------------- next part -------------- A non-text attachment was scrubbed... Name: iceweasel_77.0.1-1.patch Type: text/x-patch Size: 32462 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: iceweasel-branding-77.0-1.patch Type: text/x-patch Size: 9322 bytes Desc: not available URL: From nobody at parabola.nu Mon Jun 8 23:19:33 2020 From: nobody at parabola.nu (Parabola Website Notification) Date: Mon, 08 Jun 2020 23:19:33 -0000 Subject: [Dev] Orphan Libre package [qutebrowser] marked out-of-date Message-ID: <20200608231933.1188.84917@winston.parabola.nu> eliotreyna at disroot.org wants to notify you that the following packages may be out-of-date: * qutebrowser 1.11.0-1.par1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/qutebrowser/ * qutebrowser 1.11.1-1.parabola1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/qutebrowser/ The user provided the following additional text: Qutebrowser has been updated to the version 1.12.0. For more info, please follow this link: https://github.com/qutebrowser/qutebrowser/releases/tag/v1.12.0 Thanks. From nobody at parabola.nu Wed Jun 10 23:58:54 2020 From: nobody at parabola.nu (Parabola Website Notification) Date: Wed, 10 Jun 2020 23:58:54 -0000 Subject: [Dev] Orphan nonsystemd package [util-linux] marked out-of-date Message-ID: <20200610235854.1191.68856@winston.parabola.nu> ugw.ugw at gmail.com wants to notify you that the following packages may be out-of-date: The user provided the following additional text: 2.35.2 is out and should contain a fix for the hwclock issue where settimeofday() fails in OpenRC From nobody at parabola.nu Thu Jun 11 07:02:53 2020 From: nobody at parabola.nu (Parabola Website Notification) Date: Thu, 11 Jun 2020 07:02:53 -0000 Subject: [Dev] Orphan Libre package [cpupower] marked out-of-date Message-ID: <20200611070253.1191.22432@winston.parabola.nu> grizzlyuser at protonmail.com wants to notify you that the following packages may be out-of-date: * cgroup_event_listener 5.6.1_gnu-1 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/cgroup_event_listener/ * cgroup_event_listener 5.6.1_gnu-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/cgroup_event_listener/ * cgroup_event_listener 5.6.1_gnu-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/cgroup_event_listener/ * cpupower 5.6.1_gnu-1 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/cpupower/ * cpupower 5.6.1_gnu-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/cpupower/ * cpupower 5.6.1_gnu-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/cpupower/ * libtraceevent 5.6.1_gnu-1 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/libtraceevent/ * libtraceevent 5.6.1_gnu-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/libtraceevent/ * libtraceevent 5.6.1_gnu-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/libtraceevent/ * linux-libre-tools-meta 5.6.1_gnu-1 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/linux-libre-tools-meta/ * linux-libre-tools-meta 5.6.1_gnu-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/linux-libre-tools-meta/ * linux-libre-tools-meta 5.6.1_gnu-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/linux-libre-tools-meta/ * perf 5.6.1_gnu-1 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/perf/ * perf 5.6.1_gnu-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/perf/ * perf 5.6.1_gnu-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/perf/ * tmon 5.6.1_gnu-1 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/tmon/ * tmon 5.6.1_gnu-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/tmon/ * tmon 5.6.1_gnu-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/tmon/ * turbostat 5.6.1_gnu-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/turbostat/ * turbostat 5.6.1_gnu-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/turbostat/ * usbip 5.6.1_gnu-1 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/usbip/ * usbip 5.6.1_gnu-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/usbip/ * usbip 5.6.1_gnu-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/usbip/ * x86_energy_perf_policy 5.6.1_gnu-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/x86_energy_perf_policy/ * x86_energy_perf_policy 5.6.1_gnu-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/x86_energy_perf_policy/ The user provided the following additional text: Thank you for maintaining this package! A new version of Linux-libre on which it is based is available: http://linux-libre.fsfla.org/pub/linux-libre/releases/ From nobody at parabola.nu Thu Jun 11 07:05:48 2020 From: nobody at parabola.nu (Parabola Website Notification) Date: Thu, 11 Jun 2020 07:05:48 -0000 Subject: [Dev] Orphan Libre package [pacman] marked out-of-date Message-ID: <20200611070548.1189.23107@winston.parabola.nu> grizzlyuser at protonmail.com wants to notify you that the following packages may be out-of-date: * pacman 5.2.1-4.parabola3 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/pacman/ * pacman 5.2.1-4.parabola3 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/pacman/ The user provided the following additional text: Thank you for maintaining this package! A new pkgrel is available upstream: https://git.archlinux.org/svntogit/packages.git/log/trunk?h=packages/pacman From nobody at parabola.nu Fri Jun 12 12:39:43 2020 From: nobody at parabola.nu (Parabola Website Notification) Date: Fri, 12 Jun 2020 12:39:43 -0000 Subject: [Dev] Orphan Pcr package [toot] marked out-of-date Message-ID: <20200612123943.1190.21105@winston.parabola.nu> eomc20a2q99n at opayq.com wants to notify you that the following packages may be out-of-date: * toot 0.25.2-1 [pcr] (armv7h): https://parabolagnulinux.org/packages/pcr/armv7h/toot/ * toot 0.25.2-1 [pcr] (i686): https://parabolagnulinux.org/packages/pcr/i686/toot/ * toot 0.25.2-1 [pcr] (x86_64): https://parabolagnulinux.org/packages/pcr/x86_64/toot/ The user provided the following additional text: toot 0.26.0 released on 2020-04-15: https://github.com/ihabunek/toot/releases From dennis at dennislee.xyz Fri Jun 12 03:20:39 2020 From: dennis at dennislee.xyz (Dennis Lee) Date: Thu, 11 Jun 2020 20:20:39 -0700 Subject: [Dev] blacklist correction Message-ID: <06931E4E-C151-4EF4-9A06-F7DCE5A9C9D1@getmailspring.com> browserpass-firefox is on the blacklist (https://git.parabola.nu/blacklist.git/tree/blacklist.txt) for firefox dependency, but the firefox dependency is not listed in Arch (https://www.archlinux.org/packages/community/any/browserpass-firefox/). I have the package installed and working with icecat, so I would assume that it works for any Firefox-based browser, not just Firefox itself. -- Dennis -------------- next part -------------- An HTML attachment was scrubbed... URL: From bill-auger at peers.community Fri Jun 12 18:20:55 2020 From: bill-auger at peers.community (bill-auger) Date: Fri, 12 Jun 2020 14:20:55 -0400 Subject: [Dev] blacklist correction In-Reply-To: <06931E4E-C151-4EF4-9A06-F7DCE5A9C9D1@getmailspring.com> References: <06931E4E-C151-4EF4-9A06-F7DCE5A9C9D1@getmailspring.com> Message-ID: <20200612142055.14704719@parabola> there is an open ticket on the bug tracker, with a long discussion, about how to handle mozilla add-ons generally - it is not clear if we can simply allow them, as is - historically, we have been re-packaging the popular ones for icecat and iceweasel, simply replacing '-firefox' with '*icecat' and '-iceweasel' https://labs.parabola.nu/issues/2164 From GNUtoo at cyberdimension.org Fri Jun 12 22:15:39 2020 From: GNUtoo at cyberdimension.org (Denis 'GNUtoo' Carikli) Date: Sat, 13 Jun 2020 00:15:39 +0200 Subject: [Dev] [PATCH] README.maintenance: clarify how to become redmine admin Message-ID: <20200612221539.3419-1-GNUtoo@cyberdimension.org> Signed-off-by: Denis 'GNUtoo' Carikli --- README.maintenance | 27 ++++++++++++++++++++------- 1 file changed, 20 insertions(+), 7 deletions(-) diff --git a/README.maintenance b/README.maintenance index 3f856d16d..f7b768b64 100644 --- a/README.maintenance +++ b/README.maintenance @@ -1,13 +1,26 @@ # Redmine Web Admin -To set yourself as web admin temporarily: - $ sudo su labs - $ cd /srv/http/labs.parabola.nu/ - $ ./enter-rails-console - > HACKERS['me'].admin = true - > HACKERS['me'].save - > exit +To set yourself as web admin, you need to use Redmine's rail console. +After logging in winston, run the Redmine's rail rail console. This +will give you a full ruby interpreter and access to some Redmine +data: + $ sudo su labs + [labs at winston labs.parabola.nu]$ cd /srv/http/labs.parabola.nu/ + [labs at winston labs.parabola.nu]$ ./enter-rails-console + irb(main):001:0> + +Then you need to find your user. To do that you can use your +username. Here my exact Redmine username is GNUtoo, so I'll +use that in the example below: + irb(main):002:0> me = (User.all.where :login => 'GNUtoo')[0] + +Then we set that username to admin, save and exit: + irb(main):003:0> me.admin = true + irb(main):004:0> me.save + irb(main):001:0> exit + [labs at winston labs.parabola.nu]$ +Now an "Administration" link should appear when you are logged in Redmine. # Maintenance -- 2.27.0 From megver83 at hyperbola.info Fri Jun 12 23:51:19 2020 From: megver83 at hyperbola.info (Megver83) Date: Fri, 12 Jun 2020 19:51:19 -0400 Subject: [Dev] [PATCH] README.maintenance: clarify how to become redmine admin In-Reply-To: <20200612221539.3419-1-GNUtoo@cyberdimension.org> References: <20200612221539.3419-1-GNUtoo@cyberdimension.org> Message-ID: <38096768-cc32-b7e7-2f78-e88bc4d46eb1@hyperbola.info> El 12-06-20 a las 18:15, Denis 'GNUtoo' Carikli escribi?: > +After logging in winston, run the Redmine's rail rail console. did you write "rails" twice purposely or by mistake? > + $ sudo su labs > + [labs at winston labs.parabola.nu]$ cd /srv/http/labs.parabola.nu/ > + [labs at winston labs.parabola.nu]$ ./enter-rails-console > + irb(main):001:0> wouldn't be a better option to run that script with `sudo -u labs ./enter-rails-console`? -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From megver83 at hyperbola.info Sat Jun 13 04:36:19 2020 From: megver83 at hyperbola.info (Megver83) Date: Sat, 13 Jun 2020 00:36:19 -0400 Subject: [Dev] [RFC] Let's solve the Chromium freedom issues, definitively In-Reply-To: <20200603162957.6cd5511b@parabola> References: <4dcf3ee7-c081-94d5-0a0d-e07475b97d04@hyperbola.info> <5827880e567ead1d967d5336010972de@cock.email> <20200603162957.6cd5511b@parabola> Message-ID: <6289ad56-9c67-0c4a-530b-a8716271f695@hyperbola.info> I opened an issue in ungoogled-chromium's GitHub, asking if it completely cleans Chromium from non-free software. It certainly doesn't, but I got very interesting and useful info. For further reading: https://github.com/Eloston/ungoogled-chromium/issues/1054 in fact it looks like the ungoogle-chromium guys are open to make a fork to clean Chromium completely from non-free software: https://github.com/Eloston/ungoogled-chromium/issues/1054#issuecomment-643565898 Let them know if you want that (I think we all want it) to incentivize them. I'll give a thumb up to that comment :) -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 488 bytes Desc: OpenPGP digital signature URL: From GNUtoo at cyberdimension.org Sat Jun 13 13:26:11 2020 From: GNUtoo at cyberdimension.org (Denis 'GNUtoo' Carikli) Date: Sat, 13 Jun 2020 15:26:11 +0200 Subject: [Dev] [PATCH] README.maintenance: clarify how to become redmine admin In-Reply-To: <38096768-cc32-b7e7-2f78-e88bc4d46eb1@hyperbola.info> References: <38096768-cc32-b7e7-2f78-e88bc4d46eb1@hyperbola.info> Message-ID: <20200613132612.9080-1-GNUtoo@cyberdimension.org> On Fri, 12 Jun 2020 19:51:19 -0400 Megver83 wrote: > El 12-06-20 a las 18:15, Denis 'GNUtoo' Carikli escribi?: > > +After logging in winston, run the Redmine's rail rail console. > did you write "rails" twice purposely or by mistake? I did it by mistake. > > + $ sudo su labs > > + [labs at winston labs.parabola.nu]$ cd /srv/http/labs.parabola.nu/ > > + [labs at winston labs.parabola.nu]$ ./enter-rails-console > > + irb(main):001:0> > wouldn't be a better option to run that script with `sudo -u labs > ./enter-rails-console`? Indeed, I'll resend a patch with that improvement. Denis. From GNUtoo at cyberdimension.org Sat Jun 13 13:26:12 2020 From: GNUtoo at cyberdimension.org (Denis 'GNUtoo' Carikli) Date: Sat, 13 Jun 2020 15:26:12 +0200 Subject: [Dev] [PATCH v2] README.maintenance: clarify how to become redmine admin In-Reply-To: <20200613132612.9080-1-GNUtoo@cyberdimension.org> References: <38096768-cc32-b7e7-2f78-e88bc4d46eb1@hyperbola.info> <20200613132612.9080-1-GNUtoo@cyberdimension.org> Message-ID: <20200613132612.9080-2-GNUtoo@cyberdimension.org> Signed-off-by: Denis 'GNUtoo' Carikli --- README.maintenance | 25 ++++++++++++++++++------- 1 file changed, 18 insertions(+), 7 deletions(-) diff --git a/README.maintenance b/README.maintenance index 3f856d16d..507253287 100644 --- a/README.maintenance +++ b/README.maintenance @@ -1,13 +1,24 @@ # Redmine Web Admin -To set yourself as web admin temporarily: - $ sudo su labs - $ cd /srv/http/labs.parabola.nu/ - $ ./enter-rails-console - > HACKERS['me'].admin = true - > HACKERS['me'].save - > exit +To set yourself as web admin, you need to use Redmine's rail console. +After logging in winston, run the Redmine's rail console. This +will give you a full ruby interpreter and access to some Redmine +data: + $ cd /srv/http/labs.parabola.nu/ + $ sudo -u labs ./enter-rails-console + irb(main):001:0> + +Then you need to find your user. To do that you can use your +username. Here my exact Redmine username is GNUtoo, so I'll +use that in the example below: + irb(main):002:0> me = (User.all.where :login => 'GNUtoo')[0] + +Then we set that username to admin, save and exit: + irb(main):003:0> me.admin = true + irb(main):004:0> me.save + irb(main):001:0> exit +Now an "Administration" link should appear when you are logged in Redmine. # Maintenance -- 2.27.0 From bill-auger at peers.community Sat Jun 13 19:28:33 2020 From: bill-auger at peers.community (bill-auger) Date: Sat, 13 Jun 2020 15:28:33 -0400 Subject: [Dev] [PATCH] README.maintenance: clarify how to become redmine admin In-Reply-To: <20200613132612.9080-1-GNUtoo@cyberdimension.org> References: <38096768-cc32-b7e7-2f78-e88bc4d46eb1@hyperbola.info> <20200613132612.9080-1-GNUtoo@cyberdimension.org> Message-ID: <20200613152833.5268c1fd@parabola> On Sat, 13 Jun 2020 15:26:11 +0200 Denis wrote: > On Fri, 12 Jun 2020 19:51:19 -0400 > Megver83 wrote: > > wouldn't be a better option to run that script with `sudo -u > > labs ./enter-rails-console`? > Indeed, I'll resend a patch with that improvement. i dont consider the patch to be an improvement; just because it is making the process more complex - i could have put the verbose commands in that README originally - as a matter of abstraction, i added the HACKERS dict to make the process simpler and easier to remember that is defined here: https://git.parabola.nu/server/redmine.git/tree/config/initializers/constants.rb?h=parabola#n127 if HACKERS['gnutoo'] returned nil, that indicates a bug in the HACKERS code, or in libretools, where it gets the data from a real improvement here, would be to simplify the process even more, with a script which does the entire job - instead of ./enter-rails-console plus the manual typing, one could simply run ./make-me-web-admin 'gnutoo'; which abstracts the entire procedure it was just a matter of the usual trade-off of "how much code to write" vs "how many caveats to document", and "how much time should i spend on either" From bill-auger at peers.community Sat Jun 13 20:33:04 2020 From: bill-auger at peers.community (bill-auger) Date: Sat, 13 Jun 2020 16:33:04 -0400 Subject: [Dev] [PATCH] README.maintenance: clarify how to become redmine admin In-Reply-To: <20200613152833.5268c1fd@parabola> References: <38096768-cc32-b7e7-2f78-e88bc4d46eb1@hyperbola.info> <20200613132612.9080-1-GNUtoo@cyberdimension.org> <20200613152833.5268c1fd@parabola> Message-ID: <20200613163304.693253a0@parabola> i suppose that as this is being discussed on the mailing list now, i should add for clarification, the reason why any of that is there the simplest thing to do would be to make every team member a web admin permanently - it is mainly just a security concern; because the redmine admins controls are very brittle and difficult to understand - it is a very rare occasion when modifications to the redmine config are required; so its just best that no one has permanent admin access via the web From nobody at parabola.nu Fri Jun 19 23:43:56 2020 From: nobody at parabola.nu (Parabola Website Notification) Date: Fri, 19 Jun 2020 23:43:56 -0000 Subject: [Dev] Orphan Libre package [linux-libre] marked out-of-date Message-ID: <20200619234356.1188.23524@winston.parabola.nu> eliotreyna at disroot.org wants to notify you that the following packages may be out-of-date: * linux-libre 5.7.2-1 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/linux-libre/ * linux-libre 5.7.2-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/linux-libre/ * linux-libre 5.7.2-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/linux-libre/ * linux-libre-chromebook 5.7.2-1 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/linux-libre-chromebook/ * linux-libre-docs 5.7.2-1 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/linux-libre-docs/ * linux-libre-docs 5.7.2-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/linux-libre-docs/ * linux-libre-docs 5.7.2-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/linux-libre-docs/ * linux-libre-headers 5.7.2-1 [libre] (armv7h): https://parabolagnulinux.org/packages/libre/armv7h/linux-libre-headers/ * linux-libre-headers 5.7.2-1 [libre] (i686): https://parabolagnulinux.org/packages/libre/i686/linux-libre-headers/ * linux-libre-headers 5.7.2-1 [libre] (x86_64): https://parabolagnulinux.org/packages/libre/x86_64/linux-libre-headers/ The user provided the following additional text: Linux-libre kernel 5.7.4 has been released in June 18th. https://pod.libreplanetbr.org/posts/1071270 From bill-auger at peers.community Thu Jun 25 04:09:23 2020 From: bill-auger at peers.community (bill-auger) Date: Thu, 25 Jun 2020 00:09:23 -0400 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update In-Reply-To: References: Message-ID: <20200625000923.5efcf001@parabola> great work again grizzlyuser - sry it took so long to get to this - oaken-source has not been around lately; and the build server has been offline for a few weeks - i am going to try getting this packaged today On Wed, 03 Jun 2020 21:04:29 +0000 grizzlyuser wrote: > const release = { > -- title: "What?s New in DevTools (Firefox 76)", > -+ title: "What?s New in DevTools (Iceweasel 76)", > +- title: "What?s New in DevTools (Firefox 77)", > ++ title: "What?s New in DevTools (Iceweasel 77)", > linkText: "Read more", > - linkUrl: `https://developer.mozilla.org/docs/Mozilla/Firefox/Releases/76?${utmParams}`, > + linkUrl: `https://developer.mozilla.org/docs/Mozilla/Firefox/Releases/77?${utmParams}`, > features: [ presumably, those literal version numbers are going to require maintenance for every release - it may be better to process that file (devtools/client/whats-new/src/main.js) with sed after patching, like: sed 's|Firefox |Iceweasel |' On Wed, 03 Jun 2020 21:04:29 +0000 grizzlyuser wrote: > const dev = { > title: "Experimental Features in Firefox Developer Edition", > linkUrl: `https://www.mozilla.org/firefox/developer/?${utmParams}`, > - linkText: "Get DevEdition", > - features: [ > - { > -- header: `Full Asynchronous Stacks in Debugger`, > -- description: `Step through event, timeout and ....`, > -- href: `https://developer.mozilla.org/ ....`, > +- header: `Cross-browser CSS compatibility audit`, > +- description: `Analyze your CSS for cross-browser ....`, > +- href: `https://discourse.mozilla.org/ ....`, > - }, likewise with that 'dev' struct - to minimize future maintenance, maybe its better to leave it mainly in tact, but to clobber its definition, something like: - const dev = { + const dev = {} ; const dsfg_dev = { or if that causes problems, just clobber the important data, like: const dev = { const dev = { .... - }; + }.merge( { linkUrl: '', linkText: '' , features: [] } ) ; On Wed, 03 Jun 2020 21:04:29 +0000 grizzlyuser wrote: > Removed them for now. Could also change the URLs to open the Lockwise > website, but although libre licensed, it still has large links to > nonfree app stores and Firefox on the front page, so that wouldn't be a > huge improvement. i dunno what Lockwise is - is it a built-in or default feature; and is it so difficult to use that it needs documentation? - we could add an article on the parabola wiki explaining it; or maybe its fine without any URL in that slot From bill-auger at peers.community Thu Jun 25 11:28:11 2020 From: bill-auger at peers.community (bill-auger) Date: Thu, 25 Jun 2020 07:28:11 -0400 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update In-Reply-To: <20200625000923.5efcf001@parabola> References: <20200625000923.5efcf001@parabola> Message-ID: <20200625072811.1eefdbb1@parabola> at least one of the patches do not apply onto firefox 77.0.1 | 1 out of 1 hunk FAILED -- saving rejects to file browser/locales/en-US/browser/protections.ftl.rej the patches in this thread are applied onto the master branches of abslibre and the iceweasel branding git repo now, and the branding package is in place on the repp server; but it appears that at least one of the patches in abslibre will need re-working to allow the build start From grizzlyuser at protonmail.com Fri Jun 26 14:25:28 2020 From: grizzlyuser at protonmail.com (grizzlyuser) Date: Fri, 26 Jun 2020 14:25:28 +0000 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update In-Reply-To: <20200625072811.1eefdbb1@parabola> References: <20200625000923.5efcf001@parabola> <20200625072811.1eefdbb1@parabola> Message-ID: On Thursday, June 25, 2020 7:09 AM, bill-auger wrote: > On Wed, 03 Jun 2020 21:04:29 +0000 grizzlyuser wrote: > > > > const release = { > > -- title: "What?s New in DevTools (Firefox 76)", > > -+ title: "What?s New in DevTools (Iceweasel 76)", > > +- title: "What?s New in DevTools (Firefox 77)", > > ++ title: "What?s New in DevTools (Iceweasel 77)", > > linkText: "Read more", > > > > - linkUrl: `https://developer.mozilla.org/docs/Mozilla/Firefox/Releases/76?${utmParams}`, > > > > - linkUrl: `https://developer.mozilla.org/docs/Mozilla/Firefox/Releases/77?${utmParams}`, > > features: [ > > > > presumably, those literal version numbers are going to require > maintenance for every release - it may be better to process that > file (devtools/client/whats-new/src/main.js) with sed after > patching, like: sed 's|Firefox |Iceweasel |' > > On Wed, 03 Jun 2020 21:04:29 +0000 grizzlyuser wrote: > > > const dev = { > > title: "Experimental Features in Firefox Developer Edition", > > linkUrl: `https://www.mozilla.org/firefox/developer/?${utmParams}`, > > > > - linkText: "Get DevEdition", > > - features: [ > > - { > > -- header: `Full Asynchronous Stacks in Debugger`, > > -- description: `Step through event, timeout and ....`, > > -- href: `https://developer.mozilla.org/ ....`, > > +- header: `Cross-browser CSS compatibility audit`, > > +- description: `Analyze your CSS for cross-browser ....`, > > +- href: `https://discourse.mozilla.org/ ....`, > > > > - }, > > likewise with that 'dev' struct - to minimize future maintenance, > maybe its better to leave it mainly in tact, but to clobber its > definition, something like: > > - const dev = { > > - const dev = {} ; const dsfg_dev = { > > or if that causes problems, just clobber the important data, like: > > const dev = { > const dev = { > .... > > > - }; > > - }.merge( { linkUrl: '', linkText: '' , features: [] } ) ; Good points. I'm not happy either with this manual patch updates each new release. Will try to rework that after a new release, that is expected in Tuesday. Meanwhile, if anybody wants, feel free to change it even before that. > On Wed, 03 Jun 2020 21:04:29 +0000 grizzlyuser wrote: > > > > Removed them for now. Could also change the URLs to open the Lockwise > > website, but although libre licensed, it still has large links to > > nonfree app stores and Firefox on the front page, so that wouldn't be a > > huge improvement. > > i dunno what Lockwise is - is it a built-in or default feature; and > is it so difficult to use that it needs documentation? - we could > add an article on the parabola wiki explaining it; or maybe its fine > without any URL in that slot Lockwise is a new password manager. It's built into Firefox, and also has companion apps for iOS and Android. The apps appear to be free software, but they are hosted on Google and Apple app stores and on GitHub. It looks like all these websites serve nonfree JS, so being unable to find good replacement URLs, I just removed links and menu items that open these websites. Whoever wants the apps is still free to search the web to find them. On Thursday, June 25, 2020 2:28 PM, bill-auger wrote: > at least one of the patches do not apply onto firefox 77.0.1 > > | 1 out of 1 hunk FAILED -- saving rejects to file > browser/locales/en-US/browser/protections.ftl.rej > > the patches in this thread are applied onto the master > branches of abslibre and the iceweasel branding git repo now, > and the branding package is in place on the repp server; but it > appears that at least one of the patches in abslibre will need > re-working to allow the build start I checked out the commit before these changes, then downloaded my patch from Dev ML archives and applied it. prepare() ran without errors. Then ran a diff between patched version and the one from the origin/HEAD. It looks like whitespace has been munged at some point before applying it to the master branch. I guess either whitespace has to be fixed, or libre.patch has to be applied in PKGBUILD with -l option, like this: patch -lNp1 -i "$srcdir/libre.patch" From bill-auger at peers.community Fri Jun 26 20:39:31 2020 From: bill-auger at peers.community (bill-auger) Date: Fri, 26 Jun 2020 16:39:31 -0400 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update In-Reply-To: References: <20200625000923.5efcf001@parabola> <20200625072811.1eefdbb1@parabola> Message-ID: <20200626163931.28ff21e5@parabola> On Fri, 26 Jun 2020 14:25:28 +0000 grizzlyuser wrote: > Lockwise is a new password manager. It's built into Firefox, > and also has companion apps for iOS and Android. The apps > appear to be free software, but they are hosted on Google and > Apple app stores and on GitHub. It looks like all these > websites serve nonfree JS, so being unable to find good > replacement URLs the javascript is really not the issue in this case - the "app stores" themselves conflict directly with the FSDG; because those websites offer downloads of non-free programs - also, if the only purpose of those URLs, is to suggest to the user to install some phone "app", then that suggestion itself would conflict directly with the FSDG; because that would be condoning the use of a non-free OS, merely because there are no free OS for 99% of those phones - IMHO, even f-droid should not be suggested formally, despite that it hosts only free software, and probably is acceptable by the FSDG; because its mere existence suggests the use of a non-free OS, and hopelessly non-libre hardware On Fri, 26 Jun 2020 14:25:28 +0000 grizzlyuser wrote: > It looks like whitespace has been > munged at some point before applying it to the master branch. yea that is likely the problem - i got many whitespace errors while applying the patches from the mailing list using `git am` - it would not accept them otherwise - i fixed all the ones that git was complaining about; but the patch itself maybe introduced some more into the abslibe patches oaken-source is looking into it now From dllud at riseup.net Fri Jun 26 21:45:32 2020 From: dllud at riseup.net (dllud) Date: Fri, 26 Jun 2020 22:45:32 +0100 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update In-Reply-To: <20200626163931.28ff21e5@parabola> References: <20200625000923.5efcf001@parabola> <20200625072811.1eefdbb1@parabola> <20200626163931.28ff21e5@parabola> Message-ID: <15a57559-2d88-b792-1281-bd131d9a08b0@riseup.net> Slightly off-topic, but I decided to reply as I can add a bit more info about the subjects that bill-auger touched. On 26/06/2020 21:39, bill-auger wrote: > to suggest to the user to > install some phone "app", then that suggestion itself would > conflict directly with the FSDG; because that would be condoning > the use of a non-free OS, merely because there are no free OS > for 99% of those phones - It seems that is stretching the FSDG wording a bit too far. As long as there is at least 1 phone in the world that runs fully free software and can run that app, then I do not see how the suggestion could conflict FSDG. Practically speaking, if you're suggesting an iOS app, I can see a conflict. On the other hand, if you're suggesting an Android app that runs on Replicant[1], I would deem it ok. > IMHO, even f-droid should not be > suggested formally, despite that it hosts only free software, > and probably is acceptable by the FSDG; because its mere > existence suggests the use of a non-free OS, and hopelessly > non-libre hardware Actually, in light of the FSDG, F-Droid is not acceptable[2]. It distributes some apps that: - promote non-free software; - depend on non-free software; F-Droid has this clearly marked as anti-features for those apps, but that is not deemed to be enough. Replicant had to refrain from including F-Droid while the proper fix is still in the making[3][4]. References: ----------- [1] https://www.replicant.us/ [2] https://blog.replicant.us/2016/08/replicant-6-early-work-upstream-work-and-f-droid-issue/ [3] https://gitlab.com/fdroid/fdroidclient/-/issues/564 [4] https://redmine.replicant.us/issues/1629 From bill-auger at peers.community Fri Jun 26 23:33:20 2020 From: bill-auger at peers.community (bill-auger) Date: Fri, 26 Jun 2020 19:33:20 -0400 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update In-Reply-To: <15a57559-2d88-b792-1281-bd131d9a08b0@riseup.net> References: <20200625000923.5efcf001@parabola> <20200625072811.1eefdbb1@parabola> <20200626163931.28ff21e5@parabola> <15a57559-2d88-b792-1281-bd131d9a08b0@riseup.net> Message-ID: <20200626193320.3a702559@parabola> On Fri, 26 Jun 2020 22:45:32 +0100 dllud wrote: > It seems that is stretching the FSDG wording a bit too far. > Practically speaking, if you're suggesting an iOS app, I can > see a conflict. On the other hand, if you're suggesting an > Android app that runs on Replicant[1], I would deem it ok. yes, that was a bit of an exaggeration; but not much - "99%" was being generous, in reality - if there is any legitimate use-case in which some software can be used in 100% freedom, then it would not conflict with the FSDG, strictly-speaking On Fri, 26 Jun 2020 22:45:32 +0100 dllud wrote: > Actually, in light of the FSDG, F-Droid is not acceptable[2]. > It distributes some apps that: > - promote non-free software; > - depend on non-free software; ok, that i did not know - but now that concern reduces to: f-droid, apple, and google app-stores are all prohibited by the FSDG, and those are the only ones that exist for phone software, yes? - so we can not recommend getting any phone software from any service, even if the phone OS were 100% free like replicant From grizzlyuser at protonmail.com Sat Jun 27 08:22:50 2020 From: grizzlyuser at protonmail.com (grizzlyuser) Date: Sat, 27 Jun 2020 08:22:50 +0000 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update In-Reply-To: <20200626163931.28ff21e5@parabola> References: <20200625000923.5efcf001@parabola> <20200625072811.1eefdbb1@parabola> <20200626163931.28ff21e5@parabola> Message-ID: On Friday, June 26, 2020 11:39 PM, bill-auger wrote: > the javascript is really not the issue in this case - the "app > stores" themselves conflict directly with the FSDG; because > those websites offer downloads of non-free programs - also, if > the only purpose of those URLs, is to suggest to the user to > install some phone "app", then that suggestion itself would > conflict directly with the FSDG; because that would be condoning > the use of a non-free OS, merely because there are no free OS > for 99% of those phones - IMHO, even f-droid should not be > suggested formally, despite that it hosts only free software, > and probably is acceptable by the FSDG; because its mere > existence suggests the use of a non-free OS, and hopelessly > non-libre hardware How come nonfree JS is not the issue? It's one of the issues that's enough to disqualify these websites. Other issues you mentioned are valid too. Iceweasel doesn't come with any functionality to block nonfree JS out of the box (LibreJS, NoScript, etc.). So I think it shouldn't contain any links to websites that serve nonfree JS. From dllud at riseup.net Sat Jun 27 09:03:45 2020 From: dllud at riseup.net (dllud) Date: Sat, 27 Jun 2020 10:03:45 +0100 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update In-Reply-To: <20200626193320.3a702559@parabola> References: <20200625000923.5efcf001@parabola> <20200625072811.1eefdbb1@parabola> <20200626163931.28ff21e5@parabola> <15a57559-2d88-b792-1281-bd131d9a08b0@riseup.net> <20200626193320.3a702559@parabola> Message-ID: On 27/06/2020 00:33, bill-auger wrote: > and those are the only ones that exist for phone software, > yes? There are a few more software stores for Android, for instance Aptoide (F-Droid forked from it a long time ago), but all those are even worst, they allow proprietary software. F-Droid only contains free software, albeit some with anti-features that are not acceptable according to the FSDG. > so we can not recommend getting any phone software from > any service Yup. I guess you can still point directly to apk download links (e.g. at F-Droid's website, or the app's own site). From bill-auger at peers.community Sat Jun 27 10:58:00 2020 From: bill-auger at peers.community (bill-auger) Date: Sat, 27 Jun 2020 06:58:00 -0400 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update In-Reply-To: References: <20200625000923.5efcf001@parabola> <20200625072811.1eefdbb1@parabola> <20200626163931.28ff21e5@parabola> Message-ID: <20200627065737.086e2809@parabola> i just meant that its not the primary issue, in this case - the FSDG does not really specify what to do about remote scripts; but the app stores are explicitly forbidden - one could argue that it is showing it's age in that respect; but AFAIK, the only friction i have seen regarding javascript, is when the websites of the distro itself serves non-free javascript From grizzlyuser at protonmail.com Sat Jun 27 18:05:26 2020 From: grizzlyuser at protonmail.com (grizzlyuser) Date: Sat, 27 Jun 2020 18:05:26 +0000 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update In-Reply-To: <20200627065737.086e2809@parabola> References: <20200625000923.5efcf001@parabola> <20200625072811.1eefdbb1@parabola> <20200626163931.28ff21e5@parabola> <20200627065737.086e2809@parabola> Message-ID: In my understanding, the following two sentences of the FSDG pretty much cover the JS issue: > A free system distribution must not steer users towards obtaining any nonfree information for practical use, or encourage them to do so. > Programs in the system should not suggest installing nonfree plugins, documentation, and so on. So for example, a program has a link in the UI that suggests going to some website with nonfree JS. When the user opens that link, this nonfree software will by default be obtained, installed in the web browser (doesn't matter if temporarily), and automatically executed. I guess even the markup, stylesheets, SVG (given the feature set) can count as ?Information for practical use? in terms of FSDG. From bill-auger at peers.community Sun Jun 28 03:32:33 2020 From: bill-auger at peers.community (bill-auger) Date: Sat, 27 Jun 2020 23:32:33 -0400 Subject: [Dev] [PATCH] Iceweasl 77.0.1 update In-Reply-To: References: <20200625000923.5efcf001@parabola> <20200625072811.1eefdbb1@parabola> <20200626163931.28ff21e5@parabola> <20200627065737.086e2809@parabola> Message-ID: <20200627233233.688a286a@parabola> On Sat, 27 Jun 2020 18:05:26 +0000 grizzlyuser wrote: > In my understanding, the following two sentences of the FSDG > pretty much cover the JS issue: > > > A free system distribution must not steer users towards > > obtaining any nonfree information for practical use, or > > encourage them to do so. > > > Programs in the system should not suggest installing nonfree > > plugins, documentation, and so on. i think "obtaining" and "installing" are the key words there - you really need to squint your eyes just so, in order to place those terms regarding javascript on a web page - it is never installed, and you can not even access the copy you have of it, other than very crudely, within the sandbox - if you simply downloaded it, and replaced it with a customized version, it most likely would not function - its really a very different beast than "real software", that which the FSDG was written to address - in order for any software to be "real software" by my definition, i need to be able to practically modify it, then practically use my modified version in the original context for the same original purpose - therefore, the current web does not permit "real software" and precludes the four freedoms, usually all four if you take into account that most every website has strict TOS; because any TOS precludes freedom #0 it could be interpreted that way; but i do not believe that javascript was ever in mind at the time of its writing, and i have never seen anyone put that interpretation into practice - also, i think the intention of the recommendation is most important - the intention in this case is not to obtain javascript, it is to obtain the phone app - thats the main reason i see that as the over-riding concern that was not to argue against the interpretation - we (the parabola devs) generally think the FSDG is long overdue for an update, to be more specific about the grey areas that it had not the context or foresight to prescribe for explicitly On Sat, 27 Jun 2020 18:05:26 +0000 grizzlyuser wrote: > I guess even the markup, stylesheets, SVG (given the feature > set) can count as ?Information for practical use? in terms of > FSDG. again at the time the FSDG was written, no web pages were for practical use really - JS and CSS existed for not much more than to make web pages prettier - i do not consider reading a pretty web page to be any kind of "use", which would be different from reading a magazine perhaps today its various uses would qualify as more practical; but that was almost surely not the original intention of those words - again because the software is not really "obtained" - imagine that instead of client-side javascript, the very same website had all of the same functionality, but everything was running on the server and accessed only through a VNC client - the "practical use" would be exactly the same, and so would the users ability to modify its behavior; because the software is not ever "obtained for practical use" any software licenses for website javascripts are pretty much useless, unless the entire web service code is freely licensed and can be self-hosted IMHO all javascript which can not be self-hosted is inherently non-free; because it is impossible to exercise all four freedoms when relying on someone else computer to feed code into your blindly, even if some license grants the four freedoms theoretically - so AFAIC, we should not direct users to any third-party website for any reason, and we generally do not