[meta] remove duplicate crate versions

[metajack]

We currently have 49 duplicate crate versions in Libra Core. This meta bug documents them and tracks progress eliminating them.

To get a full list you can run cargo lockfile dups.

Duplicate Crate Versions

• ansi-term (0.11.0, 0.9.0)
  • 0.9.0 comes from prettydiff, which needs to upgrade
• arrayvec (0.5.1, 0.4.12)
• autocfg (0.1.7, 1.0.0)
• base64 (0.11.0, 0.9.3, 0.10.1)
• bit-vec (0.5.1, 0.6.1)
  • Upgrade to bit-vec 0.6.1 contain-rs/bit-set#18
• bytes (0.4.12, 0.5.4)
• codespan (0.2.1, 0.5.0)
• codespan-reporting (0.2.1, 0.5.0)
• crossbeam-channel (0.3.9, 0.4.2)
• crossbeam-queue (0.1.2, 0.2.1)
• crossbeam-utils (0.6.6, 0.7.2)
• curve25519-dalek (1.2.3, 2.0.0)
  • will be resolved when our private forks collapse see [meta] remove duplicate crate versions #1050 (comment)
• dirs (1.0.5, 2.0.2)
• ed25519-dalek (1.0.0-pre.1, 1.0.0-pre.3)
  • will be resolved when our private forks collapse see [meta] remove duplicate crate versions #1050 (comment)
• futures (0.1.29, 0.3.4)
• h2 (0.2.1, 0.1.26)
• http (0.1.21, 2.0.0)
• http-body (0.1.0, 0.3.1)
• hyper (0.13.2, 0.12.35)
• hyper-rustls (0.19.1, 0.17.1)
• idna (0.1.5, 0.2.0)
  • url
• log (0.4.8, 0.3.9)
• mime (0.2.6, 0.3.16)
• mime_guess (1.8.7, 2.0.1)
• miow (0.2.1, 0.3.3)
• nix (0.14.1, 0.17.0)
• parking_lot (0.4.8, 0.9.0, 0.10.0)
  • chashmap
• parking_lot_core (0.2.14, 0.7.0, 0.6.2)
• chashmap
• percent-encoding (1.0.1, 2.1.0)
  • Upgrade url and percent-encoding deps to 2.1.0 libp2p/rust-libp2p#1267
• proc-macro2 (0.4.30, 1.0.9)
  • update quote and syn to 1.0, rust to 2018 and run cargo fmt mcarton/rust-derivative#50
• quote (0.6.13, 1.0.2)
  • failure
• rand (0.4.6, 0.5.6, 0.6.5, 0.7.3)
  • Consider upgrading the rand crate to a version >= 0.5  zkcrypto/pairing#105
• rand_chacha (0.1.1, 0.2.1)
• rand_core (0.3.1, 0.4.2, 0.5.1)
  • [language, network] Upgrade dep criterion to 0.3.0 #1202
• rand_hc (0.1.0, 0.2.0)
  • rand (>= 0.7.1)
• rand_pcg (0.2.1, 0.1.2)
• smallvec (1.2.0, 0.6.13)
• structopt (0.2.18, 0.3.9)
  • 0.2.18 comes from prettydiff which needs to upgrade
• structopt-derive (0.2.18, 0.3.9)
• subtle (1.0.0, 2.2.2)
  • parity-multiaddr
• syn (0.15.44, 1.0.16)
• term (0.6.1, 0.5.2)
• time (0.2.7, 0.1.42)
• tokio (0.1.22, 0.2.11)
• tokio-rustls (0.10.3, 0.12.2)
• unicase (2.6.0, 1.4.2)
• unicode-xid (0.1.0, 0.2.0)
  • failure
• url (1.7.2, 2.1.1)
• version_check (0.9.1, 0.1.5)
• want (0.2.0, 0.3.0)
• webpki-roots (0.18.0, 0.17.0)
• winapi (0.2.8, 0.3.8)
  • reqwest
• x25519-dalek (0.5.2, 0.6.2)
  • will be resolved when our private forks collapse see [meta] remove duplicate crate versions #1050 (comment)

[metajack]

The old scopeguard seems to come in via yamux and tokio-reactor. Tokio issue is tokio-rs/tokio#1348. yamux needs a dependency update.

[metajack]

sha2 0.7.1 comes via rusoto, which needs a dependency upgrade.

[metajack]

rusoto/rusoto#1510 submitted upstream to fix sha2 dupe.

[huitseeker]

One count of rand 0.4.x probably comes from crypto/crypto's import of zkcrypto/pairing, upstream issue zkcrypto/pairing#105

[huitseeker]

The (curve|x|ed25519)-dalek disjunction comes from us: client uses the upstream repo, crypto uses this series of forks:
https://github.com/calibra/ed25519-dalek
https://github.com/calibra/x25519-dalek
https://github.com/calibra/curve25519-dalek
This is a temporary situation necessary to help us offer the option of a formalized crypto backend for those, work is ongoing to no longer need those forks in the first place.

[metajack]

@huitseeker Thanks for those. I've updated the issue. Is there a tracking issue for unforking?

[metajack]

Updated the list and removed all the fixed ones. 36 dupes now.

[metajack]

Updated the list and removed fixed. 54 dupes now.