"Max KSP version" format needs standardization.

[Prezombie]

There seems to be a lot of different ways mods have max version labelled, and it screws up sorting.

"1.2" and "1.2.99" do the same thing, enabling use on all 1.2.x, but "1.2.0" only works for exactly that version. Mods which only work for 1.2.0 but not 1.2.1 should be labelled that, not "1.2". It should be standardized to "1.2.x"

Some go even further with "1.99.99" or even "99.99.99". This is getting silly.

A blank entry and "all versions" is identical, and should be merged.

Any entry with a plus sign, like "0.25+" should be re-evaluated. and set to All versions, 1.x.x, or the accurate max version.

[politas]

All the unusual choices are there because the mod authors set them that way.

[politas]

The biggest problem we have is that Spacedock only allows a single KSP version , and they use the exact version numbers Squad uses. Squad don't put out "1.2.0", they put out "1.2". It's quite a nuisance.

[NexxusDrako313]

I'll admit, the reason why GrindyScience and LessGrindyScience use 9.99.999 is because there's no standard on how to make it available on all versions... or at least from SpaceDock anyway. 😐

Strangely enough.... I only did that in the .version file, not intending it to be listed in CKAN that way.

[politas]

@ValynEritai , we can override that with "Any" in CKAN if that makes sense.

[HebaruSan]

Any entry with a plus sign, like "0.25+" should be re-evaluated. and set to All versions, 1.x.x, or the accurate max version.

Agreed, and this was done in KSP-CKAN/CKAN#2131. (That format was used when a mod had only a minimum version defined, which isn't really relevant to the maximum version.)

[yalov]

You can add recommendation to wiki about version patterns. So 1.4.x is better than 1.4.9 ?

(along with recommendations about mods versioning — x.y.z, not Vx.y.z or any_version_because_its_a_MM_patch. https://semver.org)

There are still plenty nines, so

// pseudocode

if ( (major+1) % 10 == 0 || (minor+1) % 10 == 0 )      
    str = "Any";

else if ( (patch+1) % 10 == 0 || patch == null ) 
    str = major +"."+ minor +  ".x";

[HebaruSan]

Sorry, are you reporting a problem, suggesting a change, or something else? The wiki is editable to anyone, that's part of the common notion of what a wiki is.

(The link in my previous comment was wrong previously, it's updated now to point to the right issue.)

[yalov]

Sorry, are you reporting a problem, suggesting a change, or something else?

I am participating in open issue called: "'Max KSP version' format needs standardization".
I am asking what do you choose for that, "1.4.x" and 1.x.x and "Any"?

we can override that with "Any"

or more intensive override, code above

[yalov]

How do you want standardize that? this way?

  "KSP_VERSION_MAX": {
       "MAJOR": 1,
       "MINOR": x,
       "PATCH": x
   }

    "KSP_VERSION_MAX": "Any"

[HebaruSan]

This issue is talking about how the code formats the column of the "Max KSP Version" column in the GUI, not metadata.

[yalov]

then I propose more intensive override, pseudocode above

[yalov]

I'll admit, the reason why GrindyScience and LessGrindyScience use 9.99.999 is because there's no standard on how to make it available on all versions

so we need one.

---

Versions

We can recommend you to stick to Semantic Versioning (MAJOR.MINOR.PATCH) of your mod.

For making mod available on all versions you can set
KSP_VERSION_MAX to 1.9.9 or 1.4.9 depending from target interval.
They will be changed to Any and 1.4.x respectively in the GUI "Max KSP Version" column.

[HebaruSan]

What would be the advantage of doing that, rather than showing the value accurately?

[yalov]

Some go even further with "1.99.99" or even "99.99.99". This is getting silly.

So, advantage is ... un-silly-fying

[HebaruSan]

How is that silly, if that's what the author of those mods chose as the max version?

[HebaruSan]

BTW, the "99.99.99" value is just incorrect, attempting to fix it in KSP-CKAN/CKAN-meta#1338 and Dunbaratu/LaserDist#23.

[HebaruSan]

Fixed in KSP-CKAN/CKAN#2437.