pkgsrc not updating #287
Assignees
Comments
|
Yes, unfortunately we rely on Joerg's repository conversion as our upstream, and it appears he is on vacation. I'd love to set up our own conversion but it's a complicated process to do correctly (going via fossil as an intermediatory). I'll ask Joerg about getting a copy of his current setup when he returns. Sorry about that. |
|
Thanks for the verification. I'll keep working from my CVS checkout for now. Did Joerg ever document why fossil was actually required as an intermediary? I know he was trying to get a bit-exact import but I never saw the reason for using cvs2fossil instead of git-cvsimport. |
|
I believe it's the only way to get an accurate conversion, but I'll leave it to Joerg to confirm when he returns. |
jperkin
pushed a commit
that referenced
this issue
Aug 24, 2015
Upstream changes:
1.6.2 (2015-02-15)
Changes:
Added check for breaking around a binary operator. (Issue #197, Pull #305)
Bugs:
Restored config_file parameter in process_options(). (Issue #380)
1.6.1 (2015-02-08)
Changes:
Assign variables before referenced. (Issue #287)
Bugs:
Exception thrown due to unassigned local_dir variable. (Issue #377)
1.6.0 (2015-02-06)
News:
Ian Lee <ianlee1521@gmail.com> joined the project as a maintainer.
Changes:
Report E731 for lambda assignment. (Issue #277)
Report E704 for one-liner def instead of E701. Do not report this error in the default configuration. (Issue #277)
Replace codes E111, E112 and E113 with codes E114, E115 and E116 for bad indentation of comments. (Issue #274)
Report E266 instead of E265 when the block comment starts with multiple #. (Issue #270)
Report E402 for import statements not at the top of the file. (Issue #264)
Do not enforce whitespaces around ** operator. (Issue #292)
Strip whitespace from around paths during normalization. (Issue #339 / #343)
Update --format documentation. (Issue #198 / Pull Request #310)
Add .tox/ to default excludes. (Issue #335)
Do not report E121 or E126 in the default configuration. (Issues #256 / #316)
Allow spaces around the equals sign in an annotated function. (Issue #357)
Allow trailing backslash if in an inline comment. (Issue #374)
If --config is used, only that configuration is processed. Otherwise, merge the user and local configurations are merged. (Issue #368 / #369)
Bug fixes:
Don’t crash if Checker.build_tokens_line() returns None. (Issue #306)
Don’t crash if os.path.expanduser() throws an ImportError. (Issue #297)
Missing space around keyword parameter equal not always reported, E251. (Issue #323)
Fix false positive E711/E712/E713. (Issues #330 and #336)
Do not skip physical checks if the newline is escaped. (Issue #319)
Flush sys.stdout to avoid race conditions with printing. See flake8 bug: https://gitlab.com/pycqa/flake8/issues/17 for more details. (Issue #363)
wiedi
pushed a commit
to wiedi/pkgsrc-legacy
that referenced
this issue
Dec 1, 2016
-------------------------------
Added U+25B6 (black right-pointing triangle) and U+25C0
(black left-pointing triangle) (TritonDataCenter#289)
Changed look of Markdown headers ## ### #### to make them easier to tell apart (TritonDataCenter#287)
Fixed BBEdit incorrectly applying ligatures after tab (TritonDataCenter#274)
Returned Nim pragmas {. .} (TritonDataCenter#279)
Added Unicode increment U+2206 (TritonDataCenter#174, TritonDataCenter#298)
Added fish operators >-> <-< (TritonDataCenter#297)
Added safe navigation operators ?. .? ?: (TritonDataCenter#215)
Added <~> (TritonDataCenter#179, used in IntelliJ for collapsed methods)
Added F# piping operators ||> |||> <|| <||| (TritonDataCenter#184)
Added shebang #! (TritonDataCenter#169, TritonDataCenter#193)
jperkin
pushed a commit
that referenced
this issue
Apr 11, 2017
Some of the more important changes: - Fix incorrect truncation in Bcrypt. Passwords in length between 56 and 72 characters were truncated at 56 characters. Found and reported by Solar Designer. (CVE-2017-7252) (GH #938) - Fix a bug in X509 DN string comparisons that could result in out of bound reads. This could result in information leakage, denial of service, or potentially incorrect certificate validation results. Found independently by Cisco Talos team and OSS-Fuzz. (CVE-2017-2801) - Correct minimum work factor for Bcrypt password hashes. All other implementations require the work factor be at least 4. Previously Botan simply required it be greater than zero. (GH #938) - Converge on a single side channel silent EC blinded multiply algorithm. Uses Montgomery ladder with order/2 bits scalar blinding and point randomization now by default. (GH #893) - Add ability to search for certificates using the SHA-256 of the distinguished name. (GH #900) - Support a 0-length IV in ChaCha stream cipher. Such an IV is treated identically to an 8-byte IV of all zeros. - Previously Botan forbid any use of times past 2037 to avoid Y2038 issues. Now this restriction is only in place on systems which have a 32-bit time_t. (GH #933 fixing #917) - Fix a longstanding bug in modular exponentiation which caused most exponentiations modulo an even number to have an incorrect result; such moduli occur only rarely in cryptographic contexts. (GH #754) - Fix a bug in BigInt multiply operation, introduced in 1.11.30, which could cause incorrect results. Found by OSS-Fuzz fuzzing the ressol function, where the bug manifested as an incorrect modular exponentiation. OSS-Fuzz bug #287 - Fix a bug that meant the “ietf/modp/6144” and “ietf/modp/8192” discrete log groups used an incorrect value for the generator, specifically the value (p-1)/2 was used instead of the correct value of 2. - DL_Group strong generation previously set the generator to 2. However sometimes 2 generates the entire group mod p, rather than the subgroup mod q. This is invalid by X9.42 standard, and exposes incautious applications to small subgroup attacks. Now DL_Group uses the smallest g which is a quadratic residue. (GH #818) - The default TLS policy now requires 2048 or larger DH groups by default. - The default Path_Validation_Restrictions constructor has changed to require at least 110 bit signature strength. This means 1024 bit RSA certificates and also SHA-1 certificates are rejected by default. Both settings were already the default for certificate validation in TLS handshake, but this changes it for applications also. - Fix integer overflow during BER decoding, found by Falko Strenzke. This bug is not thought to be directly exploitable but upgrading ASAP is advised. (CVE-2016-9132) - Add post-quantum signature scheme XMSS. Provides either 128 or 256 bit (post-quantum) security, with small public and private keys, fast verification, and reasonably small signatures (2500 bytes for 128-bit security). Signature generation is very slow, on the order of seconds. And very importantly the signature scheme is stateful: each leaf index must only be used once, or all security is lost. In the appropriate system where signatures are rarely generated (such as code signing) XMSS makes an excellent choice. (GH #717 #736) - Add support for client-side OCSP stapling to TLS. (GH #738) - Previously both public and private keys performed automatic self testing after generation or loading. However this often caused unexpected application performance problems, and so has been removed. Instead applications must call check_key explicitly. (GH #704) - Fix TLS session resumption bugs which caused resumption failures if an application used a single session cache for both TLS and DTLS. (GH #688) - The default TLS policy now disables static RSA ciphersuites, all DSA ciphersuites, and the AES CCM-8 ciphersuites. Disabling static RSA by default protects servers from oracle attacks, as well as enforcing a forward secure ciphersuite. Some applications may be forced to re-enable RSA for interop reasons. DSA and CCM-8 are rarely used, and likely should not be negotiated outside of special circumstances. - The default TLS policy now prefers ChaCha20Poly1305 cipher over any AES mode. - The default TLS policy now orders ECC curve preferences in order by performance, with x25519 first, then P-256, then P-521, then the rest.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
It appears that this repo is no longer automatically updating an commits are now 8 days behind CVS. I assume this is related to jsonn/pkgsrc not updating. I've contacted jsonn but have not yet received a response.
Is there anything you can do to fix this?
The text was updated successfully, but these errors were encountered: