Date: Tue, 23 Jan 2024 15:59:35 +0100 From: Michael Gmelin <grembo@freebsd.org> To: Daniel Engberg <daniel.engberg.lists@pyret.net> Cc: Hiroki Sato <hrs@FreeBSD.org>, diizzy@FreeBSD.org, ports-committers@FreeBSD.org, dev-commits-ports-all@FreeBSD.org, dev-commits-ports-main@FreeBSD.org Subject: Re: git: 0c895b819c3f - main - net/http_ping: Deprecate and set expiration date to 2024-02-29 Message-ID: <20240123155935.3e7d6a98.grembo@freebsd.org> In-Reply-To: <edac21b21fd99da04c7f84b9c5367b86@mail.infomaniak.com> References: <202401221958.40MJwjFC061115@gitrepo.freebsd.org> <20240123.055130.1936217963367393840.hrs@FreeBSD.org> <edac21b21fd99da04c7f84b9c5367b86@mail.infomaniak.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 23 Jan 2024 07:01:44 +0100 Daniel Engberg <daniel.engberg.lists@pyret.net> wrote: > >=C2=A0 > >=C2=A0I=C2=A0do=C2=A0not=C2=A0understand=C2=A0why=C2=A0a=C2=A0tool=C2=A0= with=C2=A0no=C2=A0build=C2=A0or=C2=A0functionality=C2=A0issue > >=C2=A0 > >=C2=A0should=C2=A0be=C2=A0removed=C2=A0because=C2=A0it=C2=A0only=C2=A0su= pports=C2=A0HTTP/1.0.=C2=A0=C2=A0Do=C2=A0we=C2=A0have=C2=A0a > >=C2=A0 > >=C2=A0policy=C2=A0to=C2=A0remove=C2=A0ports=C2=A0based=C2=A0on=C2=A0a=C2= =A0lack=C2=A0of=C2=A0modern=C2=A0protocol=C2=A0support=C2=A0or > >=C2=A0 > >=C2=A0fewer=C2=A0users? > >=C2=A0 > >=C2=A0--=C2=A0Hiroki =20 >=20 > Hi, >=20 > There are functionality issues when HTTP 1.0 isn't supported? >=20 > Quick example: >=20 > http_ping https://chat.openai.com/ >=20 > http_ping: SSL connection failed - -1 >=20 > 002041AE453D0000:error:0A000410:SSL routines:ssl3_read_bytes:sslv3 > alert handshake > failure:/usr/src/crypto/openssl/ssl/record/rec_layer_s3.c:1586:SSL > alert number 40 >=20 > While on Windows for comparison and what I would call expected results > using http-ping. >=20 > HTTP-PING https://chat.openai.com/ GET >=20 > =C2=A0=C2=A0 =E2=94=80=E2=86=92=C2=A0=C2=A0=C2=A0=C2=A0 server advertised= HTTP/3 endpoint, using HTTP/3 >=20 > =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 1: HTTP/3.0, 104.18.37.228:443, code= =3D307, size=3D11 bytes, > time=3D183.2 ms >=20 > =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 2: HTTP/3.0, 104.18.37.228:443, code= =3D307, size=3D11 bytes, > time=3D186.2 ms >=20 > =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 3: HTTP/3.0, 104.18.37.228:443, code= =3D307, size=3D11 bytes, > time=3D193.5 ms >=20 > =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 4: HTTP/3.0, 104.18.37.228:443, code= =3D307, size=3D11 bytes, > time=3D187.0 ms >=20 > --- https://chat.openai.com/ ping statistics --- >=20 > 4 requests sent, 4 answers received, 0.0% loss >=20 > round-trip min/avg/max/stddev =3D 183.215/187.493/193.535/3.764 ms >=20 Well, yes, you demonstrated what it means if only HTTP/1.0 is supported by trying to measure the latency to a popular website. The question was if this is reason enough to deprecate a port that has a maintainer who is generally active in the project. A less intrusive alternative would have been to modify the COMMENT field to say "Measure HTTP/1.0 Latency". Cheers Michael --=20 Michael Gmelin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20240123155935.3e7d6a98.grembo>