[LY] Hurnscald storage room building sign

[Wellvin]

Esteria, who is playtesting, pointed that Hurnscald storage room & bank has no building sign. So I added one.



A copy of the map file is here (with an altered file name due to the upload site): https://rpgmaker.net/media/content/user ... r/0091.tmx

I would like to learn the proper way to submit this change because I want to work on maps.

[jesusalva]

First create an account at https://git.themanaworld.org and we'll be glad to help

[Wellvin]

Wellvin account created !

[jesusalva]

I'll be busy 'til Monday

[Wellvin]

I now have access to git.themanaworld.org.

How to submit / propose the map change ?

[Wellvin]

I've just submitted a merge request. It is ready to get approved.

[Hello=)]

Hello! Wellvin says he submitted this MR and I can see it's been in august. Now its october - and I'm also dead sure I've got at least one resource update with new monsters and somesuch (as HoraK asked to help with testing that). But I'm dead sure there's still no sign near Hurns banking building. Has something went wrong or something?

[WildX]

I don't know if anything went wrong with this specific MR but in general it's hard to get approval for Legacy MRs because:
1. People don't ask enough. Now that you posted I know to poke people. A lot of contributors are too quiet about requesitng action.
2. Admins aren't willing to merge. You need to ask the admins why. All I know is that when you tell an admin something several times and he always changes subject it usually means "no". I think they should consider solutions like appointing a maintainer.

[Hello=)]

I've digged deeper and found some more hilarious story...
1) Ppl went complaining their tiled saved it as map 1.0 format while original been map 1.4. Its the only objection I've seen in the wild.
2) Technically, Wellvin's map works just fine - I've tested on my M+ using local override. There's building sign after all.
3) There was no data or feature loss. I've did very thorough map data comparison. Actually, the only difference I can see (by carefully grinding on map and its format using git diff <original> <wellvin's tmx>) would be few extra harmless fields ADDED by old tiled. These appear to be quite harmless.

So basically ppl moaned on wrong map version, and adding sign to building took 2 months. Despite map being technically okay on nearly all aspects I can think of - and I've even tested it by local override. That is: git diff only lists legit layer changes, numeric "ID" added for layers (newer formats dont have numeric ID, just names, 1.0 had both ID and name) and ... yes, header mentions map version 1.4 instead of 1.0. So basically re-saving it as 1.4 doesn't changes anything at all. Except maybe getting rid of few unused fields and changing header (we talk about ~50 bytes of data per 160K map file). So much for welcoming new contributors

Upd: if ppl really so anal on map version, I've built tiled 1.4.1 and resaved map as 1.4.1 (attached in zip), I've tested this map locally and it works. Actually, it makes no any visible differences in any regards so its been some really pointless ritual around numbers in map header.

p.s. .tmx extension not allowed? Oh, really?

[jesusalva]

Hello! Wellvin says he submitted this MR and I can see it's been in august. Now its october - and I'm also dead sure I've got at least one resource update with new monsters and somesuch (as HoraK asked to help with testing that). But I'm dead sure there's still no sign near Hurns banking building. Has something went wrong or something?

I've digged deeper and found some more hilarious story...
1) Ppl went complaining their tiled saved it as map 1.0 format while original been map 1.4. Its the only objection I've seen in the wild.
2) Technically, Wellvin's map works just fine - I've tested on my M+ using local override. There's building sign after all.
3) There was no data or feature loss. I've did very thorough map data comparison. Actually, the only difference I can see (by carefully grinding on map and its format using git diff <original> <wellvin's tmx>) would be few extra harmless fields ADDED by old tiled. These appear to be quite harmless.

So basically ppl moaned on wrong map version, and adding sign to building took 2 months. Despite map being technically okay on nearly all aspects I can think of - and I've even tested it by local override. That is: git diff only lists legit layer changes, numeric "ID" added for layers (newer formats dont have numeric ID, just names, 1.0 had both ID and name) and ... yes, header mentions map version 1.4 instead of 1.0. So basically re-saving it as 1.4 doesn't changes anything at all. Except maybe getting rid of few unused fields and changing header (we talk about ~50 bytes of data per 160K map file). So much for welcoming new contributors

Upd: if ppl really so anal on map version, I've built tiled 1.4.1 and resaved map as 1.4.1 (attached in zip), I've tested this map locally and it works. Actually, it makes no any visible differences in any regards so its been some really pointless ritual around numbers in map header.

p.s. .tmx extension not allowed? Oh, really?

1) Yes, wrong tiled version is the sole and only reason it wasn't merged
2) As it should, in ManaPlus. M+ should run from Tiled 0.5 or so without issues.
3) Old tiled does not add fields, new tiled does. And XML is a picky format, so other tools (like converters, parsers, shiny silver bullets, whatever) will notice the missing fields and moan.

Remember that Wellvin is downgrading the map version. So the meaning is inverted, unless you use a colored markup:
https://git.themanaworld.org/legacy/cli ... /473/diffs

To be honest, if you have an up-to-date Tiled, all you need to do is resave it and it should hopefully fix everything. Anyone could do it in theory, provided they have enough time, interest and willpower to do so
(I don't have any of these three right now, but may have later)

Spoiler:

So much for welcoming new contributors

“Please use the latest software and tools, or at least the same everyone here uses, so we at least are on the same page if things goes downhill” is hardly an unwelcome message.

C'mon, Tiled is on 1.9 already, it was five years ago since Tiled 1.0 release and I can't recall a stable supported distro which still has it except for Ubuntu 18.04 Extended Support which exists to support servers. No excuses!

Not only that, but in the long run, not pointing out to update the software is slightly worse. Remember that all FLOSS devs are butes. When a dev start screaming about disrespecting the world map layout which did not exist prior to Tiled 1.4, then it would be unwelcome.

Having a standard minimum version on development team allows to ensure everyone have minimum features e.g. World map, and aids aligning expectations between everyone involved. This also reduces the amount of "garbage" generated by diffs, allowing to see more important changes (e.g. if you save the map with different visible/hidden layers, this messes up thumbnails for other people working with map files ─ tmxrasterizer's normal behavior ─ so unless it is intentional, that should be enough to prevent a merge. If you are using the same version, you can spot such stupid mistake in 5 seconds and fix in less. But if you are changing versions all the time, that gets quickly hidden in a lot of garbage.)

Whatever. I'm rambling.

EDIT: TMX whitelisted.

[jesusalva]

By the way, there was also a warning that they used the master branch (which is always a bad idea). Dang, I really need someone who can walk others through using git, where is Micksha and gumi when you need them?!

[Hello=)]

1) Yes, wrong tiled version is the sole and only reason it wasn't merged
2) As it should, in ManaPlus. M+ should run from Tiled 0.5 or so without issues.
3) Old tiled does not add fields, new tiled does. And XML is a picky format, so other tools (like converters, parsers, shiny silver bullets, whatever) will notice the missing fields and moan.

Sry but you've stumbled on mage belonging to some funny school of magic. I'm quite sure I can stand my points on technical level.
1) IMO its ourtght stupid reason, inconvenient if not hostile to new contributor who wanted to make our world better.
2) If this comes to it, okay, I'll waste a bit more of my time. I've investigated further. TMW stll uses plenty of maps in 1.0 format. So I guess whole pipeline can swallow it.
3) TBH I never doubted of code quality of silver bullets and something, it always matched my wildest expectations. But even then, all this stuff somehow lives with "2)" to the date. I dont get how +/- 1 map in 1.0 format changes that, there're plenty others. So if tools moaned you should have ton of that long before Wellvin came with his map...

Remember that Wellvin is downgrading the map version. So the meaning is inverted, unless you use a colored markup:

Yes he does and most unusual thing is fact it mostly ADDS few (relatively useless and harmlsee) extra fields. Most sane XML parsers would just ignore it if they dont need it. Not to mention TMW still haves plenty of 1.0 maps to the date. Tiled started quick version bumping only few years ago. And even then it doesn't seems to be something really major down to format level.

https://git.themanaworld.org/legacy/cli ... /473/diffs

One of beauty of using local tools is that its me who decides what sides are convenient to me so no "inversion" and I can haz quite sane view of what's going on the way it maps my brain better.

Another interesting thing I can do project-widea analisys of map version and so on to verify some claims. I just did. Guess what.

To be honest, if you have an up-to-date Tiled, all you need to do is resave it and it should hopefully fix everything. Anyone could do it in theory, provided they have enough time, interest and willpower to do so

To be honest, to me it looks like if in this particular case ppl were eating brain of new contributor for NOTHING. It took 2 months. And all that to add damn sign to building. And when it comes to map versions TMW still got some 1.0 format maps. So everything lives just fine with it, eh?

So if we're about being honest, I have impression Wellvin faced ton of inconvencies out of nowhere. Furthermore, same stands true for most Linux contributors using tiled from their distro repos.

(I don't have any of these three right now, but may have later)

TBH to me it looks like like ppl created problem out of nothing in first place. There was no such problem Jes, you've invented it yourself. But I've solved it anyway, building tiled 1.4.1 and resaving map. So zip in previous message holds updated TMX. But realistically its been very pointless tantrum to bump few numbers. It doesn't changes anything.

“Please use the latest software and tools, or at least the same everyone here uses, so we at least are on the same page if things goes downhill” is hardly an unwelcome message.

1) It dooms Linux contributors to a lot of unnecessary tantrums out of nowhere instead of installing tiled from repos and happily using it. So you break contributor experience here.
2) I've investigated map versions a bit and there're different ones, down to 1.0 (it didnt had exact tiled version by then it seems). So whole thing had to live with 1.0 maps anyway, Wellvin haven't made it any worse.
3) TMW uses assorti of map versions, there're plenty of 1.0 as well. At which point I wonder how "everyone" defined in this case given assorti of maps. I'd say 1.0 could fit this bill just as well, no?

C'mon, Tiled is on 1.9 already, it was five years ago since Tiled 1.0 release and I can't recall a stable supported distro which still has it except for Ubuntu 18.04 Extended Support which exists to support servers. No excuses!

I don't see use of tiled from repos of Lubuntu 18.04 or so as major crime given there's plenty of 1.0 maps in TMW - this map isnt any worse than any and there was no data loss, I've investigated that. So WTF, seriously?

Not only that, but in the long run, not pointing out to update the software is slightly worse. Remember that all FLOSS devs are butes. When a dev start screaming about disrespecting the world map layout which did not exist prior to Tiled 1.4, then it would be unwelcome.

There's such thing as compatibility and interop. Ruining it without MAJOR gain to justify that is silly and causes a lot of trouble and aggravation out of nowhere. Then ppl lecture me about welcoming new contributors on IRC... I think there's something doesn't matches, eh?

Having a standard minimum version on development team allows to ensure everyone have minimum features e.g. World map, and aids aligning expectations between everyone involved.

And given I see 1.0 maps in TMW, what this "standard" could be?

Whatever. I'm rambling.

And valid technical justification could be like what? How about exact fields you needed, exact versions with sane justification of features they got vs older ones or something? I think these are right questions to ask after seeing someone turned down for 2 months? Doubly so after I investigated the matter. Esp after being lectured in IRC about welcoming new contributors.

[jesusalva]

Actually, I did mistook TMW mapping guildelines with ML mapping guidelines

And you wont find maps below 1.4 in ML client-data. This being said, any valid XML parser will ignore extra fields, but missing fields always cause errors in most languages.

(And no, I'm not having a good day in overall, but that should be quite obvious)

[Hello=)]

Alright it happens. But I still think on contributor's side "draw map -> submit" is more appealing than getting turned down for 2 months, getting asked to install other SW versions, etc. Esp in Linux where one can just install tiled from repos finding it by description. Demanding to go some web site, download some thing, etc makes contributor experience worse and makes me wonder "why?". I've not really found good answer on this here.

Technically looking on fields I can deduce while map is 1.0, Wellvin likely used version 1.2.x as (apparently used by Debian 10 and Ubuntu 18 LTS, both still supported IIRC) where few new fields appeared. Interestingly layer ID fields apparently gone away in 1.4.x yet Tiled changelog doesn't mentions it so I suspect Tiled's changelog isn't 100% perfect either, but it happens and when in doubt I prefer to trust my eyes.

IMO pipeline compat with wider range of maps is feature, not bug. Either way hope next time it works a bit better than that for ppl who are trying to improve my homeworld.

[WildX]

But I still think on contributor's side "draw map -> submit" is more appealing than getting turned down for 2 months, getting asked to install other SW versions, etc. Esp in Linux where one can just install tiled from repos finding it by description. Demanding to go some web site, download some thing, etc makes contributor experience worse and makes me wonder "why?"

I could not agree more with this statement.