Newest at the top
| 2025-01-08 17:02:09 +0100 | <sm> | yup, thanks |
| 2025-01-08 17:02:02 +0100 | <sm> | it became supported in https://github.com/haskell/cabal/blob/master/Cabal/ChangeLog.md#2200-mikhail-glushenkov-march-2018 I think |
| 2025-01-08 17:02:00 +0100 | <tomsmeding> | looks like 2.2 and onwards? |
| 2025-01-08 17:01:26 +0100 | <hseg> | https://0x0.st/8-sC.txt is the list of tags containing that commit |
| 2025-01-08 17:00:59 +0100 | mari-estel | (~mari-este@user/mari-estel) mari-estel |
| 2025-01-08 17:00:56 +0100 | <tomsmeding> | https://cabal.readthedocs.io/en/stable/file-format-changelog.html#spec-history |
| 2025-01-08 17:00:54 +0100 | <hseg> | The earliest commit I see in the cabal repo mentioning SPDX is 3dfc0ea466d254fcefb3826d6f15cda30d95cc0a |
| 2025-01-08 17:00:54 +0100 | <tomsmeding> | > The sequence of specification version numbers is not contiguous because it’s synchronised with the version of the Cabal library. |
| 2025-01-08 17:00:50 +0100 | Smiles | (uid551636@id-551636.lymington.irccloud.com) (Quit: Connection closed for inactivity) |
| 2025-01-08 17:00:38 +0100 | <tomsmeding> | naturally :) |
| 2025-01-08 16:59:56 +0100 | <sm> | nobody puts those in their docs :) |
| 2025-01-08 16:59:40 +0100 | <sm> | next I hunt for cabal-version release dates |
| 2025-01-08 16:58:52 +0100 | rachelambda8 | (~rachelamb@cust-95-80-25-71.csbnet.se) (Quit: β reduced) |
| 2025-01-08 16:58:36 +0100 | <tomsmeding> | in any case, long |
| 2025-01-08 16:58:25 +0100 | <sm> | 🙏🏻 |
| 2025-01-08 16:58:19 +0100 | <tomsmeding> | it's with reference to a `cabal-version` version, not cabal-install, but perhaps those are synchronised? |
| 2025-01-08 16:57:30 +0100 | <tomsmeding> | https://cabal.readthedocs.io/en/stable/cabal-package-description-file.html#pkg-field-license |
| 2025-01-08 16:57:03 +0100 | <sm> | do you know the earliest version of cabal-install that supports that ? |
| 2025-01-08 16:55:49 +0100 | <sm> | hseg: good to know, thanks! |
| 2025-01-08 16:55:23 +0100 | <yushyin> | glad to see that cabal adopted spdx ids |
| 2025-01-08 16:52:03 +0100 | <hseg> | it can! license: GPL-3.0-or-later |
| 2025-01-08 16:51:23 +0100 | <hseg> | sm: that's true of hledger, and makes my life easier |
| 2025-01-08 16:51:08 +0100 | <hseg> | 2) as mentioned above -- some licenses require distributing their copyright notices, it is insufficient to have a report saying "I depend on the BSD-licensed foo package" |
| 2025-01-08 16:50:58 +0100 | <sm> | and for the record it's GPLv3+ (cabal format can't express that IIRC) |
| 2025-01-08 16:50:39 +0100 | <sm> | hledger itself is the only GPL thing (!) |
| 2025-01-08 16:50:33 +0100 | tromp | (~textual@92-110-219-57.cable.dynamic.v4.ziggo.nl) (Quit: My iMac has gone to sleep. ZZZzzz…) |
| 2025-01-08 16:49:55 +0100 | <hseg> | two problems. 1) non-canonical license names -- am seeing BSD2, BSD-2. That's OK, but with GPL there's a significant difference between GPL3 and GPL3+ which needs to be reflected |
| 2025-01-08 16:49:46 +0100 | <sm> | stack ls dependencies --license | awk '{print $2 " " $1}' | sed -E 's/(BSD|GPL)-/\1/' | sort |
| 2025-01-08 16:47:45 +0100 | <sm> | group them by license, then add the texts ? |
| 2025-01-08 16:47:29 +0100 | <sm> | stack ls dependencies --license looks excellent, isn't that a good starting point ? |
| 2025-01-08 16:46:29 +0100 | <hseg> | problem is that that doesn't really satisfy obligations |
| 2025-01-08 16:46:17 +0100 | <hseg> | yeah. Or I was thinking of installing something like https://hackage.haskell.org/package/src/example/cabal-plan.html next to hledger's LICENSE file |
| 2025-01-08 16:45:45 +0100 | <sm> | that sounds nice |
| 2025-01-08 16:45:08 +0100 | <merijn> | hseg: having a --about or --licenses that prints out all licenses seems the most correct way to comply |
| 2025-01-08 16:44:01 +0100 | <hseg> | s/a top-level dep/the package I'm building/ |
| 2025-01-08 16:43:42 +0100 | <hseg> | at least it's a top-level dep |
| 2025-01-08 16:43:35 +0100 | <hseg> | Ah, dammit -- hledger *is* GPL |
| 2025-01-08 16:43:16 +0100 | <hseg> | binary |
| 2025-01-08 16:43:09 +0100 | <merijn> | Do you have a library or a binary? |
| 2025-01-08 16:42:54 +0100 | <merijn> | hseg: I mean, that's pretty much the way to go, tbh |
| 2025-01-08 16:42:34 +0100 | <hseg> | Still, that's better than stack ls dependencies --license which just gives the author's spelling of their licenses |
| 2025-01-08 16:42:06 +0100 | <hseg> | Which looks official and nice, but given BSD, MIT, ISC all require you to distribute their copyright notices, is still slightly off what's technically needed |
| 2025-01-08 16:41:01 +0100 | <hseg> | Yup -- cabal plan license-report |
| 2025-01-08 16:40:44 +0100 | <merijn> | hseg: btw, I vaguelly recall cabal-plan having a command for collecting all transitive dependencies |
| 2025-01-08 16:40:08 +0100 | <merijn> | hseg: Realistically, if you ship GPL in a binary you're already kinda hosed :p |
| 2025-01-08 16:37:04 +0100 | saulosilva | (~saulosilv@181.216.220.21) (Quit: Client closed) |
| 2025-01-08 16:29:43 +0100 | <hseg> | Fair enough. The topic came to mind because Arch just started demanding packaging files be licensed, and the train of thought got me there |
| 2025-01-08 16:28:33 +0100 | <haskellbridge> | <magic_rb> Especially if your project is also free software |
| 2025-01-08 16:28:17 +0100 | <haskellbridge> | <magic_rb> Realistically no one does this, so unless youve a very good reason to do it, i wouldnt worry about it |
| 2025-01-08 16:22:06 +0100 | <hseg> | (at least for those dependencies that require distribution of their license, eg MIT, BSD-n with n>=2, GPL, ...) |