IRC channel logs

<damo22>youpi: i noticed my recent commits have author email set to bug-hurd@gnu.org, is there something i can do on my end to fix that?

<damo22>eg, my git send email command

<ZhaoM>morning

<youpi>damo22: I don't know, for a start maybe check the mails produced by your git commands

<damo22> https://paste.debian.net/plain/1341566

<damo22>i think its related to how you are applying the patches direct from the list?

<damo22>i seem to get my email from the list as an email from the list email address

<damo22>The email i get from the list says:

<damo22>Reply-To: "Damien Zammit" <damien AT zamaudio.com>

<damo22>From: "Damien Zammit via Bug reports for the GNU Hurd" <bug-hurd AT gnu.org>

<damo22>i replaced @ with AT

<damo22>maybe its the list configuration

<damo22>ok i just disabled getting a copy of my own list emails to the list, maybe that will fix it?

<ZhaoM>when compiling vim with '-DDEBUG' and run `make test_conceal` I can see the log saying `libvterm: Unhandled CSI t`

<ZhaoM>not sure if this is the cause of failing test_conceal

<ZhaoM>still investigating

<damo22>youpi: (12:49:09 PM) rwp: Meanwhile... If the FSF exim configuration is setting the mailing list from address to the mailing list then that indicates that your email service provider has implemented a strict DMARC setting which prohibits using email through mailing lists.

<damo22>somehow my email service provider uses strict DMARC that is incompatible with mailing lists, but fsf has set the from address to the mailing list address to stop my messages being lost

<youpi>damo22: ah, yes, DMARC can be a problem

<youpi>I don't know if git people have considered a fix for that in send-email

<youpi>like putting the From: to be used in another header

<damo22>it has envelopeSender

<damo22>but thats not the problem i think, mailman doesnt re-sign the email with DKIM, it detects the dmarc is strict and then changes the from address to the list address

<youpi>sneek: later tell ZhaoM maybe re-run the vim testsuite with the latest hurd debian package, the testsuite result has changed a lot apparently

<sneek>Will do.

<damo22>any changes that rewrite the email need to be re signed

<damo22>or wrapped with ARC

<youpi>sneek: later tell ZhaoM (also use vim version 2:9.1.0967-1), the testsuite used to take a very very long time, now it's much faster

<sneek>Okay.

<youpi>damo22: it's rather "any changes needs to be signed again, or else the from needs to be changed and the signature dropped"

<damo22>or the email can be wrapped into another message that is parsed by the mailing list

<damo22>with ARC that preserves the original chain of signatures

<damo22>apparently there is a "Sender: " email header but its not widely used

<damo22>for relaying emails on someone elses behalf

<damo22>but if the mailing list adds that in, then the signature will not verify

<damo22>it gets complicated pretty quickly

<damo22>if there was a header prefix that was excluded from being part of the signed payload, that would make it easier to inject headers for things like mailing lists

<damo22>and it would be obvious that it was added later because it would be called something like X-Additional-Header-*

<ZhaoM>a

<sneek>Welcome back ZhaoM, you have 2 messages!

<sneek>ZhaoM, youpi says: maybe re-run the vim testsuite with the latest hurd debian package, the testsuite result has changed a lot apparently

<sneek>ZhaoM, youpi says: (also use vim version 2:9.1.0967-1), the testsuite used to take a very very long time, now it's much faster

<ZhaoM>OK

<azert>a

<sneek>Welcome back azert, you have 1 message!

<sneek>azert, youpi says: grub also supports http. It can very probably also read/write memory

<azert>maybe they are equivalent, but I just prefer u-boot documentation that is written for developers, and the marketing that comes with a friendly feeling

<azert>the embedded segment is very friendly to developers in general , with top one being microcontrollers

<azert>there are some fields of computer engineering that are very friendly to developers: embedded, web, data bases. Others that are pure cancer, 3d graphics and artificial intelligence for instance

<azert>the latter being presented as anti-humanity

<azert>GNU and Linux was a big opportunity for the people of my generation back in the days. I’m too young to find a correspondence with the perspective that are being presented in the artificial intelligence field to young people. Probably nukes?

<azert>Something that gets pushed like cray and can profit only a few. An economic irrationality of sort

<azert>crazy*

<azert>even reading the feuillettons of the XIX century, it feels that TNT was more profitable to young people back then than AI nowadays

<azert>You want to train that model? 200k euro just for trying!

<azert>Ok let’s stop this rant, bye

IRC channel logs

2024-12-29.log