Date: Sun, 21 Jan 2024 19:35:40 +0100 From: Kurt Jaeger <pi@freebsd.org> To: "Mikhail T." <mi+t@virtual-estates.net> Cc: Daniel Engberg <daniel.engberg.lists@pyret.net>, ports-committers@freebsd.org, dev-commits-ports-all@freebsd.org, dev-commits-ports-main@freebsd.org Subject: Re: git: b430a140c818 - main - net-im/purple-gowhatsapp: add WhatsApp plugin for libpurple Message-ID: <Za1j_ENNptehLVre@fc.opsec.eu> In-Reply-To: <df410abd-dd00-4e7e-83eb-f8a95ec15b9f@virtual-estates.net> References: <f22ba5fb94cbee57ef6dbac2bdb3db87@mail.infomaniak.com> <f74f9837-121e-47ac-819b-27a40d3b4891@virtual-estates.net> <ewlhhqgckfoo4nj2jmryynhh2admdz6wy3lwkyav7nvhok565l@liht5kzkacig> <7e07375b-32bc-4778-8977-d87d6e135679@virtual-estates.net> <ptq5dukjavkfpoe2gic6mssywdxh7f77xsounaxmy5llxinz2j@w6243b7unduq> <8ab62f5f-bb62-4633-9d1c-d7a8a8e1fc8a@virtual-estates.net> <Zaz0W9odSaxyDrCj@fc.opsec.eu> <13433172-a8cb-494c-a435-fd4d8418a2e6@virtual-estates.net> <21dcca053d36f9bec4005ffb18897f51@mail.infomaniak.com> <df410abd-dd00-4e7e-83eb-f8a95ec15b9f@virtual-estates.net>
next in thread | previous in thread | raw e-mail | index | archive | help
Hi! > I don't see the implication at all. I appreciate the fallout e-mails, but > I don't see, why that makes it mandatory for me to use the same tool(s) > locally. For example, the cluster builds every port on multiple hardware > platforms -- for different OS-releases. Does that imply the committers > also must have such multitude of different hardware/release combinations > locally too? Well, it helps to test-build at least on our Tier 1 arch (amd64) and on all production releases (13.2, 14.0), and CURRENT would be helpful as well. This post https://lists.freebsd.org/pipermail/freebsd-announce/2021-April/002030.html lists FreeBSD/arm64 (also known as AArch64) as Tier 1 as well. The official list of plattforms is here: https://www.freebsd.org/platforms/ > If the 37 seconds it took the cluster to fail the port is really such a > drain on the resources, marking the port BROKEN would be a thing to do -- > that's a one-line change, that still keeps the code available for sharing. The cluster builds on many OS versions, so it's probably 37seconds times approx. 20 variants. So yes, it taxes the cluster. > Anyway, I think, I hacked the port into pre-fetching the additional > modules using go.mk's facilities, and will be committing that shortly. It > still is not perfect, because the port is a mixture of C and Go-code, but > it should build fine now. Thank you for the feedback. Yours, Thanks! That sounds reasonable! -- pi@FreeBSD.org +49 171 3101372 Now what ?
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Za1j_ENNptehLVre>