Time Nick Message 11:46 sfan5 MTG should check minetest.features on startup, people using mtg 5.4.0 with engine 5.3.0 causes errors that don't make sense are hard to debug and generally wasted time 11:46 sfan5 this is probably the fifth time that I look at an user report and have to tell that that MTG/engine versions need to match 11:47 Krock yes I also wondered why there's yet no such check but only ever cared about it in my own mods 11:48 Krock a simple error() would suffice 11:49 pgimeno why minetest.features and not minetest.get_version()? 11:49 Krock forks may change get_version() 11:49 pgimeno the more reason, no? 11:49 Krock also it's about the supported API version, not the strict Minetest version 11:50 Krock also I think it's easier to check against features than caring about version comparisons ;) 11:51 pgimeno I'd rethink what you're saying 11:51 pgimeno minetest.features is specified to hold only flags, what would you do, add a new flag per version? 11:52 Krock you'd check for a specific API function or feature, depending on what are the requirements for default and dependencies 11:52 Krock not limited to features ofc 11:52 Krock if you need client-side translations in your mod, check against minetest.get_translator, for example. 11:53 Krock requirements can be checked per-mod if necessary, depending on what kind of new API is required there 11:58 Krock though your approach would be straight-forward to enforce the version matching 12:06 pgimeno minetest.features has proven to be very prone to being forgotten 12:07 pgimeno for that reason, I question its usefulness in preventing a version mismatch 12:10 sfan5 it literally does the job 12:13 pgimeno if you say so 12:41 MTDiscord The problem with the "features" thing is that it really needs to contain like 2 orders of magnitude more flags to ve useful. One of the major reasons I want to change what features I use in a mod is not because a feature is absent, but because it's broken. 12:42 MTDiscord There'd need to be like attachments, attachments_fixed, attachments_fixed_again, attachments_fixed_for_real_this_time, etc. 12:44 MTDiscord At least with a version number we'd be able to negotiate support for official MT releases, even if it missed everything between. 14:20 MTDiscord yep 14:20 MTDiscord every PR would add a feature flag 14:22 pgimeno pretty much, yes 15:59 sfan5 without checking the github org one might think the development team consist of exactly two people 15:59 sfan5 wonder why everyone else is so quiet 16:04 MTDiscord forum has been deader recently to 16:05 MTDiscord discord/matrix seems to be normal 16:05 Krock well yes. limited time recently. 16:17 MTDiscord Well I'm quiet because I'm usually quiet, I am possibly the most minor contributor, and I don't want to disturb the professional discussion in the dev channels. ? 16:18 MTDiscord But I'm frequently here. ?