Re: netty split
Bastien Roucaries <[email protected]> Thu, 08 Jan 2026 21:24:52 +0100
| Newsgroups | gmane.linux.debian.devel.java |
|---|---|
| Message-ID | <4671822.1IzOArtZ34@debian-ei> |
--nextPart3086118.BddDVKsqQX Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8"; protected-headers="v1" From: Bastien Roucaries <[email protected]> To: [email protected] Subject: Re: netty split Date: Thu, 08 Jan 2026 21:24:23 +0100 Message-ID: <4671822.1IzOArtZ34@debian-ei> MIME-Version: 1.0 Le jeudi 8 janvier 2026, 18:00:48 heure normale d=E2=80=99Europe centrale M= arkus Koschany a =C3=A9crit : > Hi Bastien, >=20 > Am Donnerstag, dem 08.01.2026 um 17:47 +0100 schrieb Bastien Roucaries: > > Le lundi 5 janvier 2026, 17:22:47 heure normale d=E2=80=99Europe centra= le Bastien > > Roucaries a =C3=A9crit : > > > Le lundi 5 janvier 2026, 10:28:30 heure normale d=E2=80=99Europe cent= rale Bastien > > > Roucaries a =C3=A9crit : > > > > Hi, > > > >=20 > > > > in order to compile netty I need brotli4j that need netty buffer > > > >=20 > > > > I think I should split in this case netty, and extract the buffer l= ibrary > > > > out netty main package. > > > >=20 > > > > Because buffer depend on utils (commons jar) I will need to split a= lso > > > > the common lib > > > >=20 > > > > Could you confirm this analysis ? > > > >=20 > > > > And how can I implement this ? > > > I tried this > > > https://salsa.debian.org/rouca/netty > > >=20 > > > But depends of debian package fails to be correct (empty) > > I have manually added the depends and pushed to netty tree > > https://salsa.debian.org/java-team/netty > >=20 > > I do not know if it is correct >=20 > I can't do a deep dive into netty currently but if you discover a circular > dependency I would try to disable those modules which require brotli4j or= netty > buffer first. You can ignore certain modules with debian/libnetty-java.po= ms by > adding the --ignore flag. Try to get a minimum / bare bone version of net= ty > first. Another option is to patch out certain functionality like brotli if > that's too troublesome at the moment.=20 Agreed but longer term getting brotli4j compiled will be nice so for me sp= lit was needed > Splitting the package into multiple > different ones is usually not necessary. >=20 > Regards, >=20 > Markus >=20 >=20 --nextPart3086118.BddDVKsqQX Content-Type: application/pgp-signature; name="signature.asc" Content-Description: This is a digitally signed message part. Content-Transfer-Encoding: 7Bit -----BEGIN PGP SIGNATURE----- iQIzBAABCgAdFiEEXQGHuUCiRbrXsPVqADoaLapBCF8FAmlgEncACgkQADoaLapB CF+1OA//W1VCSwn3JPLwKpwNDutyPlr+wIY1i9pbP6mAC4kLgP7W2FvMgu+AYMSJ aiTmINxEKlmfgUMIbDSY9OmxtiXYK+V51qQXIuUT1keBWUjz2bd3FIWsGd4MuA15 syvby8M8uktboIc5eRmoe7wKKrT6LM8/f8aBt8a4BN7Ovy4C4tHNUfrr4bbeq/gH JbwycBbPbgYGDpkhNsmXp3GRMovWqLPB+f3SuP9rYBsG7GsWEDoDoVUNTM1TEbVB OIJ9GTCfspnfOhkHRN9A00LOUR3SfK32PcfaJQKeHos1A+UYKsXnfJnqSEAInWkc QoNmXJk34GOOhJopqDkQILOY8Ovoi4cyrA0KaJmSNVURBGP7JqdEjN/mNCUMGi2S Mn/UaH3Oe5mnQH9dhATUHYHo6iR4oh+1RKhbWN9MjBGPDzR/J/DfVBWwkZpoqCr7 i0/gkab3mOIQe3vC6pKI1ILZNOQfZHNe2QCFIJ0ijLHw1ll0Mh2+GCuW9lukWxVq G8umZlOz7UzO2H9HVV47dDLpdJ1W5fXugyF09zrDa54jfvGCWolCyGgsjirj/hZF 9b0sZUUjlMVMvhpVJjcmtLDGhcd0BUyTQOetRLxBhR7wGapYbiJ1o9JkV+OrHvKp J27OZ0NNxwyr+ulAfffpXfDIi2X8yXxyFRTNVac+VCupC90iBv0= =UE5J -----END PGP SIGNATURE----- --nextPart3086118.BddDVKsqQX--