yggdrasil-jumper: add new package

[one-d-wide]

Description:

Yggdrasil Jumper is independent project that aims to transparently reduce latency of a connection over Yggdrasil network, utilizing NAT traversal to bypass intermediary nodes. It periodically probes for active sessions and automatically establishes direct peerings over internet with remote nodes running Yggdrasil Jumper without requiring firewall or port configuration.

Changes:

• Add yggdrasil-jumper package.
• Add logic to start/stop yggdrasil-jumper within netifd interface script for yggdrasil.

Maintainer: me
Compile tested: x86_64 master
Run tested: x86_64 master

[wfleurant]

awesome ⚙️ work @one-d-wide

[stangri]

@one-d-wide -- great work. What's the reason for splitting the Makefiles for yggdrasil and yggdrasil-jumper instead of using the same Makefile to build both?

[one-d-wide]

@one-d-wide -- great work. What's the reason for splitting the Makefiles for yggdrasil and yggdrasil-jumper instead of using the same Makefile to build both?

Yggdrasil Jumper is optional and logically separate part. And mixing different build processes in one Makefile probably wouldn't benefit readability anyway.

[one-d-wide]

@BKPepe

Thank you. It should be fixed now.

[systemcrash]

@BKPepe

Thank you. It should be fixed now.

If so, mark the conversations as resolved.

[BKPepe]

Nope, I did not finish it. SoB also requires real name, etc Dne po 8. 1. 2024 14:46 uživatel Paul Donald ***@***.***> napsal:

…

@BKPepe <https://github.com/BKPepe> Thank you. It should be fixed now. If so, mark the conversations as resolved. — Reply to this email directly, view it on GitHub <#23072 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA7IDVEV7Q6RYFSJFNG5LK3YNP2CRAVCNFSM6AAAAABBPUNDC6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQOBRGAZTQNBWGM> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[one-d-wide]

Nope, I did not finish it. SoB also requires real name, etc

I replaced it (both in the Makefile and commit details) so CI formalities check should pass.

[stangri]

@one-d-wide -- great work. What's the reason for splitting the Makefiles for yggdrasil and yggdrasil-jumper instead of using the same Makefile to build both?

Yggdrasil Jumper is optional and logically separate part. And mixing different build processes in one Makefile probably wouldn't benefit readability anyway.

You're changing the yggdrasil proper code to accommodate yggdrasil-jumper, so they are evidently connected.

My suggestion would be to merge the makefiles, like other packages do, for example acme: https://github.com/openwrt/packages/blob/master/net/acme-acmesh/Makefile, nebula: https://github.com/openwrt/packages/blob/master/net/nebula/Makefile, etc.

[wfleurant]

You're changing the yggdrasil proper code to accommodate yggdrasil-jumper, so they are evidently connected.

hi @stangri -- the bits of netifd related shell code used by yggdrasil-jumper may be the only option, no? um..

i'm only finding one PKG_SOURCE_URL per Makefile. its late, may be i am missing something... sorry im missing a solution, ttyl 💤

[wfleurant]

@one-d-wide do we fastforward to recheck with ci changes: #23084

[wfleurant]

@BKPepe on your review, please. identity / signature OK in CI #23084

[systemcrash]

Is this ready?

[systemcrash]

@BKPepe Please update your review

[systemcrash]

ping @ynezz @981213

[systemcrash]

@BKPepe Please update your review

ping @ynezz @981213

[ynezz]

@systemcrash FYI openwrt/actions-shared-workflows@12d9551

[BKPepe]

Interesting. Should we revert the recent CI changes here as well and again require real name? I have not any idea that it was reverted, there. :-( Dne po 25. 3. 2024 16:52 uživatel Petr Štetiar ***@***.***> napsal:

…

@systemcrash <https://github.com/systemcrash> FYI ***@***.*** <openwrt/actions-shared-workflows@12d9551> — Reply to this email directly, view it on GitHub <#23072 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AA7IDVGXCJZGYTYG5ZYXSOTY2BB37AVCNFSM6AAAAABBPUNDC6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDAMJYGMZTCOBYG4> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[1715173329]

+1 and this time even no PR was made before committing.

[systemcrash]

The weasel word there is "should". No discussion or vote raised.

[ynezz]

and this time even no PR was made before committing.

@1715173329 can you expand a bit on that? There was a PR openwrt/openwrt#14380 and there was a public discussion (1/2024), two times already (2/2023) , see the last thread https://lists.openwrt.org/pipermail/openwrt-devel/2024-January/042063.html leading to this series of pull requests.

[BKPepe]

I think, he means that revert in the actions-shared-workflows repository.

[systemcrash]

and there was a public discussion (1/2024)

Then evidently the following is erroneous and it should be reverted (revert the revert):

@systemcrash FYI openwrt/actions-shared-workflows@12d9551

[systemcrash]

And this PR can be merged

[BKPepe]

evidently the following is erroneou

I think, @nbd168 wants to do vote, which is done on openwrt-adm, when they are voting about adding new people to the core team.

[systemcrash]

I think, @nbd168 wants to do vote, which is done on openwrt-adm, when they are voting about adding new people to the core team.

I see no desire. Writing "should" two months ago is inconclusive.

[systemcrash]

So @one-d-wide I propose that while everyone is really busy trying to get a vote off the ground, in the meantime you adopt some official looking pseudonym, like One D Wide, or John D Wade and force push. Done.

[one-d-wide]

Hi, thank you for the advice. I also resolved a merge conflict so this (along with luci PR) should be ready to be merged.

@BKPepe
Please update your review, it is displayed as "changes requested" at the moment.

[1715173329]

oh please also increase PKG_RELEASE for yggdrasil.

[one-d-wide]

oh please also increase PKG_RELEASE for yggdrasil.

Oops. Done.

[1715173329]

Merged, thank you!