For example here: https://mail-archive.freebsd.org/mail/archive/2026/dev-commits-ports-all/20260803.dev-commits-ports-all.html Could anybody please point me to how are these displayed (is it a PHP script?) and I'd try to fix it.
You can decode it using `japanese/nkf`, but is that not what you meant? $ echo '=?UTF-8?B?UmU6IGdpdDogNDVjMTQyOTdiMTkxIC0gbWFpbiAtIHNlY3VyaXR5L3B5LWdvb2dsZS1hdQ==?= =?UTF-8?B?dGg6IHVwZGF0ZSAyLjU1LjIg4oaSIDIuNTYuMg==?='|nkf -m Re: git: 45c14297b191 - main - security/py-google-auth: update 2.55.2 → 2.56.2
We have several interfaces for mailing lists. https://mail-archive.freebsd.org/mail/archive/ is one of them. The getmsg.cgi is written in perl, and probably has problems with multiline UTF-8 subjects. You can find it in the clusteradm svn repo at www/mailarchive.nyi/cgi/getmsg.cgi https://lists.freebsd.org/archives/ is another, more recently developed one. The same message: https://lists.freebsd.org/archives/dev-commits-ports-all/2026-August/219152.html has a readable subject in that archive.
(In reply to Kurt Jaeger from comment #2) Aha, RFC2047 specifies those header line encodings. The pointer to nkf is really helpful 8-)
Somehow when I click on the "archive" link in https://mail-archive.freebsd.org/mail/archive/2026/dev-commits-ports-all/20260803.dev-commits-ports-all.html this archive doesn't show the same messages as the html page. Specifically, the word "folly" appears on the html page but not in the archive itself. Furthermore, wget fails to download this archive and keeps retrying. Something isn't right with this web server. $ wget "https://mail-archive.freebsd.org/cgi/getmsg.cgi?fetch=0+0+archive/2026/dev-commits-ports-all/20260803.dev-commits-ports-all+archive" -O archive.mbox --2026-08-12 11:55:44-- https://mail-archive.freebsd.org/cgi/getmsg.cgi?fetch=0+0+archive/2026/dev-commits-ports-all/20260803.dev-commits-ports-all+archive Resolving mail-archive.freebsd.org (mail-archive.freebsd.org)... 96.47.72.84, 2610:1c1:1:606c::50:15 Connecting to mail-archive.freebsd.org (mail-archive.freebsd.org)|96.47.72.84|:443... connected. HTTP request sent, awaiting response... 200 OK Length: unspecified [text/plain] Saving to: ‘archive.mbox’ archive.mbox [ <=> ] 192.00K 420KB/s in 0.5s 2026-08-12 11:55:45 (420 KB/s) - Read error at byte 196608 (No error: 0).Retrying. --2026-08-12 11:55:46-- (try: 2) https://mail-archive.freebsd.org/cgi/getmsg.cgi?fetch=0+0+archive/2026/dev-commits-ports-all/20260803.dev-commits-ports-all+archive Connecting to mail-archive.freebsd.org (mail-archive.freebsd.org)|96.47.72.84|:443... connected. HTTP request sent, awaiting response... 200 OK Length: unspecified [text/plain] Saving to: ‘archive.mbox’ archive.mbox [ <=> ] 192.00K 483KB/s in 0.4s 2026-08-12 11:55:46 (483 KB/s) - Read error at byte 196608 (No error: 0).Retrying. --2026-08-12 11:55:48-- (try: 3) https://mail-archive.freebsd.org/cgi/getmsg.cgi?fetch=0+0+archive/2026/dev-commits-ports-all/20260803.dev-commits-ports-all+archive Connecting to mail-archive.freebsd.org (mail-archive.freebsd.org)|96.47.72.84|:443... connected. HTTP request sent, awaiting response... 200 OK Length: unspecified [text/plain] Saving to: ‘archive.mbox’ archive.mbox [ <=> ] 192.00K 421KB/s in 0.5s 2026-08-12 11:55:49 (421 KB/s) - Read error at byte 196608 (No error: 0).Retrying. --2026-08-12 11:55:52-- (try: 4) https://mail-archive.freebsd.org/cgi/getmsg.cgi?fetch=0+0+archive/2026/dev-commits-ports-all/20260803.dev-commits-ports-all+archive Connecting to mail-archive.freebsd.org (mail-archive.freebsd.org)|96.47.72.84|:443... connected. HTTP request sent, awaiting response... 200 OK Length: unspecified [text/plain] Saving to: ‘archive.mbox’ archive.mbox [ <=> ] 256.00K 512KB/s in 0.5s 2026-08-12 11:55:53 (512 KB/s) - Read error at byte 262144 (No error: 0).Retrying.
Created attachment 273691 [details] getmsg-fixed.tgz Kurt sent me the archive getmsg.tgz that has the CGI Perl script to display an individual message. I fixed UTF-8 handling there - no garbled content any more. What was fixed: 1. MIME-encoded subject line was decoded and converted to the HTML format 2. Immediate UTF-8 chars in the message body are now properly displayed. The script that shows the list of ML messages is still missing. I am asking Kurt via e-mail about it.
The ML fetching issue: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=297485
Cc: wosch@ who maintains this script.
(In reply to Kousuke Kannagi from comment #1) No, the decoding is done generically for any UTF-8 chars, not just Japanese.
I don't know how to get to clusteradm svn repo at www/mailarchive.nyi/cgi/getmsg.cgi Could somebody please send me, attach, or point to the CGI script that displays the list of ML messages in the first URL that I've posted?
(In reply to Yuri Victorovich from comment #9) It's www/mailarchive.nyi/mailindex/ (some dirs etc) in the clusteradm repo, sent by mail. wosch@, can you work with yuri to get the patches into production ?
About the failing file download: wget fails but the following brute-force style command keeps re-connecting and succeeds eventually: > aria2c --allow-overwrite=true -m 100 "https://mail-archive.freebsd.org/cgi/getmsg.cgi?fetch=0+0+archive/2026/dev-commits-ports-all/20260706.dev-commits-ports-all+archive" -o archive.mbox
(In reply to Yuri Victorovich from comment #8) $ echo '=?UTF-8?B?UmU6IGdpdDogNDVjMTQyOTdiMTkxIC0gbWFpbiAtIHNlY3VyaXR5L3B5LWdvb2dsZS1hdQ==?= =?UTF-8?B?dGg6IHVwZGF0ZSAyLjU1LjIg4oaSIDIuNTYuMg==?='|nkf -m | file - /dev/stdin: Unicode text, UTF-8 text
I added support for MIME mail headers in getmsg.cgi. I will fix mailindex.cgi later - this is more work.
(In reply to Wolfram Schneider from comment #13) Hi Wolfram, I have a fix which I will post here over this weekend (it had wrinkle so I didn't post it yet). Please don't waste your time. Thanks, Yuri
Created attachment 273841 [details] mailindex-fixed.tgz List view is fixed.
(In reply to Yuri Victorovich from comment #15) Thanks, I will look at the patches soon.