On Sat 14 Feb 2026 at 11:19:27 (+0700), Max Nikulin wrote: > On 26/01/2026 2:42 pm, David wrote: > > On Mon, 26 Jan 2026 at 03:18, Max Nikulin wrote: > > > On 26/01/2026 5:01 am, David wrote: > > > > On Sun, 25 Jan 2026 at 18:01, D. R. Evans wrote: > > > > > > > (see my e-mail > > > > > <[email protected]> in another > > > > > sub-thread), > > > > > > > > If you want to reference other list messages, can you please do that by > > > > providing links into the list archive which can be found at for > > > > example: https://lists.debian.org/debian-user/2026/01/threads.html > > > > > > David, if you wish to open that message in a browser then > > > you may easily do it: ↑↑↑↑↑↑ ?
> > > https://lists.debian.org/msgid-search/?m=MESSAGE_ID > > Notice "?m=", it is important, "a bit broken" below is related to it. > > > > Actually every message has a bit broken header with a link to the list > > > archive, see List-Archive. > > > > > > Having Message-ID, it is more convenient to open that message inside > > > mailer I think that way depends on the target message being in the mailer's currently open mailbox. > > > or in another web mailing list archive. Do you mean some site that carries debian-user, or are you broadening this discussion to other mailing lists besides Debian's? If so, are you saying that other mailing lists use …/msgid-search/?m=MESSAGE_ID URLs? I don't find that to be so. > > > Instead, your are > > > suggesting to use a link that is local to the lists.debian.org web site. Isn't lists.debian.org the official archive? Ironic that you call the link "local" when your mailer suggestion above appears to be limited to one mailbox. > > Here's an example link that I wrote recently in another message: > > https://lists.debian.org/debian-user/2026/01/msg00411.html > > I believe, both links: Message-ID and HTTPS with internal message > number, have some disadvantages. On the other hand, I consider both > are acceptable. What I did not like, is that you requested to use > another style of links. Message-ID itself is not a link, and I'll hazard a guess that many people reading this list may not know how to turn a Message-ID into a URL. Whatever writes the List-Archive URLs doesn't get this right either. > > If I click on that link, the Message-ID you prefer is provided in > > line 7 of the served page. So the Message-ID is available there for anyone > > who needs it. > > I am aware of it, but it works only if HTTP server with mail archive > is available. I have seen enough broken links on the web. Come again? If: https://lists.debian.org/debian-user/2026/02/msg00276.html is not available, what likelihood is there that: https://lists.debian.org/msgid-search/[email protected] will be? > If a mail > list archive moved to other site or removed completely then "msg00411" > is not a helpful identifier. It can not be used for obtaining the same > message from another archive. I wasn't aware that the list archive would be moved. I see that: https://lists.debian.org/debian-user/1994/01/msg00000.html is still where it always was. > I admit, Message-ID's usually have no > hints concerning date and mailing list name. Actually I prefer > redundancy and references with sender, destination, and timestamp: > > David to debian-user. Re: Referencing mail messages (was: Use > grub-rescue on a non-bootable RAID-formatted drive) Mon, 26 Jan 2026 > 07:42:56 +0000. > <mid:CAMPXz=q3itpmvy2nz8n0j5epusnuo0mxy2icyhj3qk9o1ra...@mail.gmail.com> > (Or https://...) > > So generic search may be used to obtain the message. I am realizing > that almost nobody will use detailed links. Wow, it's hard enough to get some people to attribute there quotes, let alone persuade them to write references like that. Again, I think for "mid:" to work, the target message has to be in the mailer's currently open mailbox. RFC2392: "The following message points to another message (_hopefully_ still in the recipient's message store)." (my emphasis) > > Here are the reasons why I am inclined to continue preferring the direct > > link to the archive: > > 1) It is human-readable, so it is clear and safe to click on. > > 2) Its format and use is consistent with links to other web sources. > > I believe, > <https://lists.debian.org/msgid-search/?m=CAMPXz=q3itpmvy2nz8n0j5epusnuo0mxy2icyhj3qk9o1ra...@mail.gmail.com> > is not really worse It has to be whenever it doesn't work. > > 3) It does not require special knowledge for any (eg newbie) reader to find > > the information, because everyone knows how to follow https links. > > 4) It works for readers who are not using a dedicated mail client. > > 5) It works for readers who are reading the archive in a web browser. > > I agree that some kind of trick (e.g. browser extension, boookmark > with search substitution) is necessary if Message-ID links are > actively used. > > > 5) Sometimes the Message-ID search of the Debian mail archives fails to > > find messages, so I prefer to use a method that seems to always work. > > See "?m=" above. The ?m= construction seems unnecessary for most Message-IDs, yet it's insufficient make those like the ones at the end of this post actually work without correcting them. > Some effort is required to fix HTTP server > configuration or the script parsing URL (or both) and smartlist (to > use Archived-At and safer URL). My knowledge of Perl is not enough to > be confident that changes I might suggest will be safe from security > point of view. AFAICT, the debian-user server can handle …/msgid-search/?m=MESSAGE_ID URLs perfectly correctly for all the Message-IDs that I've seen used on the list, but you have to correct some of them manually yourself. > On 26/01/2026 12:17 pm, David Wright wrote: > > > > Some Message-ID references are very long, anything up to about > > 160 characters, and are not easy to check when cut and pasted > > into a browser due to their random nature. > It is positive feedback loop. Nobody cares that some soft generates > excessively long or peculiar Message-ID's since they are rarely used > in links. More active usage of Message-ID will create some pressure to > developers. Some of the URLs I am sent in emails have hundreds of "random" characters in them. I've just turned up a 1478-character URL just for unsubscribing from PlutoTV marketing emails. Knowing that Message-IDs are going to be used in URLs will not feel like much of a constraint on their length. > > https://lists.debian.org/msgid-search/Y/+tJdQluFDMC4Ci@use > > "+" is more tricky, it is not enough to just insert "?m=". In my > opinion, mailers should use URL- and shell-safe Message-ID's even if > the price is a bit longer identifiers. I anticipate that some > developers may be upset because they follow RFC's, but I prefer to > avoid exercising if all Message-ID handlers are written in safe way. I think RFC3986 implies that safe URLs should be generated from such Message-IDs by the list remailer, because the remailer knows what the safe set is for performing the msgid-search with that syntax. Message-IDs' generators don't necessarily know all the uses to which they might be put. It's tedious to have to make the ID safe yourself, assuming you know what the rules are. > David Wright, I have noticed that you sometimes skips explicit > "https://"; prefix. At least Thunderbird does not make links active > ones. That's right. And sometimes, that's the difference between a message posting, or being immediately returned. You might also notice that I remove ? and "funny bits" from Subject lines. They can increase my spam score enough for posting to fail. Last year the mere mention of Eben King's American gmx address in the body of an email ensured that it would fail to post here. (This has nothing to do with the reason that the debian.org domain was rejecting /any/ emails from me for a couple of months late last summer.) I think most browsers will add http: or https: for you when you paste one of my URLs lacking the protocol. > I do not see a point in insisting on HTTPS link when Message-ID is > provided. Any variant is better than mentioning messages without any > identifier. Everybody has its own preferences concerning style of > linking. As a Message-ID itself is not a URL, I did wonder whether the OP has been misled by the appearance of the links at the top of webpages in the archive. Links are rendered there with the format of a Message-ID, but the link's underlying URL is in the …/msg00000.html format that we two Davids are preferring. To see a Message-ID style of _URL_, you have to hover over the magnifying glass at the left. BTW, the web archiver also appears unable to form these URLs when the Message-IDs look like Y/+tJdQluFDMC4Ci@use or ZPJmz/[email protected] As a result, the Message-ID is rendered as dead text, rather than as an active link. Cheers, David.
