lynx

I don't think so, this was just a temporary hack for Angel. It removes checking for confusion.

I didn't check the EEs, but since they merged both datasets, it is not a surprised that you have them.

I've checked all the data files using iesh to get more info. Last research showed it was used in bg1 and not in bg2 and iwd2. Here's the rest and some details. Structure: https://gibberlings3.github.io/iesdp/file_formats/ie_formats/are_v1.htm#formAREAV1_0_Song_entries iwd1, iwd2 and pst have no areas with any of those ambient fields set. bg2 does have a few, mostly inns, probably copies of areas in bg1: None of those sound files are present, but I didn't check if adding them makes a difference. bg1 of course has a long bunch If volume isn't printed, it's 100. There's a few cases where it differs between day and night, usually being louder in the night. Almost a hundred use different ambients for the two fields (eg. two entries for day: : AMB_E19L and AMB_E19N). Don't know when this "extended" second field is supposed to come into play. Might be they're all longer and weren't used in low memory situations / settings. Bit 5 of the flags is supposed to toggle something like this, but I haven't checked. EDIT: with 183 vs 1, the size hypothesis is confirmed. I checked the uncompressed sizes, which is not the same as length, but close enough. The outlier is amb_m14a vs amb_m14. Stuff for the fixpack, I guess.

I've recently gave in, enabled the github bugtracker and we're slowly moving old reports (wiki, sf) there. So from now on, please file any reports, missing features and enhancement requests here: https://github.com/gemrb/gemrb/issues

I'm surprised at the opposite actually.

Modding on ees is harder for the user, but gog is a good option either way.

The download tool could do it for them quite easily. Also for the deltas, I don't see a problem with shared directories. If two mods have the same filenames, there'll be problems much sooner than this. If you're nuking the dir before updating, then it's not a delta update anymore, but a redownload. Dealing with user changes is tricky, but you could git stash beforehand and git stash pop afterwards. The only problem is if there are conflicts, since that requires a manual intervention.

Amber is indeed one of the best, so an update would be exciting!

didn't scan the content, but the effect variant that sets a stat is there. https://gibberlings3.github.io/iesdp/opcodes/bg2tob.htm#op295

Typically code is released under one of the plethora of FLOSS licenses, while blobs and other non-code as CC. However, most mods are small, so there's no real gain in making a distinction and using two. Choosing is a good idea though, not because github will ask you, but as it makes updating abandoned stuff clear as day and sets expectations from the contributors. The authors can always agree to sublicense differently too. A very contrived example: I generally license gemrb mods as GPL, but I could grant them to Saros under a BSD license, in case he wanted to sell them without reciprocity.

But you're literally defining specifications, there's no need to be stuck in the past.

Come on, it's 2019, everyone is using unicode now. And the mods that have a safety against this are exactly the ones hosted in version control, which can be another motivator to use it.

Do you need any help with the url rewrite stuff? It's quite annoying and degrading search engine results as well.

It's simple to do with gemrb, but then you'd be playing something else.

What's the problem with a single file? I don't see how it would cause usability issues for translators.

I don't really care what numbers they resolve to, as long as they do. Our file effectively hides the mod created one at runtime (it's the only hard override we use). For scripts it of course works, while dialogs actions aren't compiled, so we have a problem. But projectiles aren't a problem, since there we ourselves use an extra file to define all the base missing ones. First projectl.ids is read then ours on top of it. Some of the original ids might get overriden, but mod problems would be easy to fix.

We're trying to figure out how to best handle the dynamic stats.ids (re)generation and extension that DS does. Eg. in SCS: https://github.com/Gibberlings3/SwordCoastStratagems/blob/master/stratagems/ds/ds.tph#L460 It would be really handy if we could see all those flexible stats at 400+ that get added and I thought the sole mention of stats_extra.2da might have it. But a user says it doesn't exist, so I'm confused and looking for other shortcuts. We use our own override in gemrb to not have a headache with all ~10 engine variants, so I'd rather patch scs and any other mod relying on random stats over 400 than change that. The easy way out would be to just also install a duplicate of the final stats.ids, which we can then use to extend ours, but maybe a smaller interim file already exists and would do.

So what kind of deltas are you talking about, just the fact that the mod changed, not actually any format? I don't see a problem there either; things under SCM will update nicely, while for normal archives you can wipe everything but inis and it should also be fine.

Too bad, this means it would also complicate distribution if you wanted to do it in advance. It would basically have to be a mod for us, each user doing it themselves.

Very nice results indeed! Hopefully someone will integrate it into SDL, so we geet more upscaling options for free.

1. Unless SR and IR (iirc) got new releases, they still use both mechanisms for two different release streams (3 and 4). 2. and 3. You can do this later if there is a need. 4. Are you talking about diffs or binary deltas here? The first don't have this problem.

What's the point of a repo if you're just uploading exes though?

I'm sure you're familiar with regexes enough, this is a simple case. The thing to realize is that reading from the disk several times is a bad idea, so read info should be cached. If I remember correctly, he's using powershell, which is OOP, so it can be done cleanly through the use of custom objects.

Why change any checks, it's just one regex to remove the setup- prefix if it's present.

the big ones are used also in bgs for character generation, not just iwd.