aboutsummaryrefslogtreecommitdiff
path: root/NEWS
diff options
context:
space:
mode:
Diffstat (limited to 'NEWS')
-rw-r--r--NEWS35
1 files changed, 34 insertions, 1 deletions
diff --git a/NEWS b/NEWS
index 4d65541..6616d8c 100644
--- a/NEWS
+++ b/NEWS
@@ -1,10 +1,43 @@
-*- org -*-
#+TITLE: guile-email NEWS – History of user-visible changes
-Copyright © 2019 Arun Isaac <arunisaac@systemreboot.net>
+Copyright © 2019, 2023 Arun Isaac <arunisaac@systemreboot.net>
Please send guile-email bug reports to guile-email@systemreboot.net
+* Changes in 0.3.0 (since 0.2.2)
+** Support obsolete syntax
+[[https://tools.ietf.org/html/rfc5322#section-4][Obsolete syntax]] described in RFC5322 is now supported.
+
+** Reimplement (email base64) from scratch
+The (email base64) was originally copied from GNU Guix. Now, it has
+been reimplemented from scratch.
+
+** Quit the autotools build system
+We move away from the autotools build system to a hand-written
+Makefiles. It's simpler and good enough for a small project like
+guile-email.
+
+** Noteworthy bug fixes
+*** Do not rely on bytevector->string throwing an expection with empty bytevector
+Passing an empty bytevector and an invalid charset to
+bytevector->string no longer raises an exception since Guile commit
+5ea8c69e9153a970952bf6f0b32c4fad6a28e839. This caused guile-email
+tests to fail on Guile 3.0.7. We no longer rely on this behavior.
+https://lists.systemreboot.net/guile-email/87mtnv1r2p.fsf@gnu.org/
+*** Support CR LF sequences in quoted-printable encoding
+Earlier, only LF characters where accepted in quoted-printable encoding.
+https://lists.systemreboot.net/guile-email/20230103121942.10497-1-whatson@tailcall.au/
+*** Support Date fields with missing seconds
+The seconds field is optional. Respect that.
+https://lists.systemreboot.net/guile-email/20230105103324.4396-1-whatson@tailcall.au/
+*** Assume application/octet-stream if Content-Transfer-Encoding is unrecognized
+§6.4 of RFC2045 specifies that any entity with an unrecognized
+Content-Transfer-Encoding must be treated as if it has a Content-Type
+of "application/octet-stream", regardless of what the Content-Type
+header field actually says.
+*** Handle Received headers with two tokens but no timestamp
+
* Changes in 0.2.2 (since 0.2.1)
** Noteworthy bug fixes
*** Tolerate decoding errors in MIME encoded words