The Validator - Find errors quickly and with minimal pain!

[Jamie550]

This is not a good suggestion. I don't care about existing revoke_allowed and I'm not going to waste time removing them. Now instead of a few errors in decisions, I have 648. That doesn't give a good overview and hides real issues.
What suggestion should've been was "ignore whether revoke stuff exists". They are not "invalid nodes" they just don't do anything.

I intentionally and with full awareness suggested for it to be a warning rather than an error (warning as in, non-fatal and silenceable with some application flag or setting). And I hope there's some configurable flags to disable such warnings as desired, but novice modders should certainly be made aware of such things as these, lest we have them wasting time on coding useless ai_will_revoke blocks.

I purposely ignored @ngppgn's suggestion for it to be a warning and made it a full-fledged error.

Unless there is some compelling reason to keep such blocks ("it was there before" I do not consider by itself a compelling reason), it will stay a non-configurable error.

In general the Validator cares about code quality, which is one reason it doesn't accept straNGe_miXEd_casE_teXT. For similar reasons, it does not accept useless code existing simply because it was there before (again, unless there is some compelling reason). Removing the dead code, however annoying it is, is a one-time thing and would not need to be done in the future.

 

[vukica]

My point wasn't that it should be a warning, my point was that it's a waste of time for everyone.

 

[Voy]

What kind of problems arise with having more than 9 characters for character ID?

 

[Rydelfox]

I'm pretty sure the same stores IDs as a signed 32-bit integer. A little past 2 billion, it will treat them as negative numbers. A little past 4 billion, it will loop over.

 

[ngppgn]

I purposely ignored @ngppgn's suggestion for it to be a warning and made it a full-fledged error.

Unless there is some compelling reason to keep such blocks ("it was there before" I do not consider by itself a compelling reason), it will stay a non-configurable error.

In general the Validator cares about code quality, which is one reason it doesn't accept straNGe_miXEd_casE_teXT. For similar reasons, it does not accept useless code existing simply because it was there before (again, unless there is some compelling reason). Removing the dead code, however annoying it is, is a one-time thing and would not need to be done in the future.

A compelling reason IMHO: let's say you are making a mod that override some and relies onother vanilla files (e.g. extendin ga vanilla event.chain fired from a decision that sets event targets. but with the mod not touching the dexision file itself, just the events), you need to validate the decision for the validator to recognised the event target as being properly set. But you don't modify that decidion file, so you're stucl with those harmless errors.

In general, I think it would be wise for flexibility of use to let the user globally toggle on/off error reports that are not harmful to the code. Somethinf like fatalErrorsOnly = no that an user could change to yes to make the Validator focus only on the properly code-breaking errors.

 

[Jamie550]

My point wasn't that it should be a warning, my point was that it's a waste of time for everyone.

I guess we'll have to agree to disagree here.

A compelling reason IMHO: let's say you are making a mod that override some and relies onother vanilla files (e.g. extendin ga vanilla event.chain fired from a decision that sets event targets. but with the mod not touching the dexision file itself, just the events), you need to validate the decision for the validator to recognised the event target as being properly set. But you don't modify that decidion file, so you're stucl with those harmless errors.

In general, I think it would be wise for flexibility of use to let the user globally toggle on/off error reports that are not harmful to the code. Somethinf like fatalErrorsOnly = no that an user could change to yes to make the Validator focus only on the properly code-breaking errors.

If the file is still in vanilla, then you can happily say to show only mod errors. If you had to copy it to your mod, then well ideally lobby Paradox to clean their files up and stop confusing people.

The Validator does lots of non-fatal errors and I don't see it worthwhile to generally allow them with a flag as after all it's hard to know if something's nonfatal or not. For this specific case because it exists right now in quite a few places maybe a flag would be valid. However really Paradox and everyone else should clean up dead code that's of no use.

 

[Jamie550]

v1.22.8: https://www.dropbox.com/s/9lpzhg7ypa8xury/Audax.Validator v1.22.8.zip?dl=0

 

[IlikeTrains]

Is there any more detailed changelog than the one in Changelog.rtf?

 

[Abominus]

--- Error 1 of 1 ---
Parse Failure
Path: <mod>\events\harem_event.txt
Approximate location: Line 1, column 1
Error: Detected a byte order mark on the file, which should not be present.

And now?

 

[jursamaj]

--- Error 1 of 1 ---
Parse Failure
Path: <mod>\events\harem_event.txt
Approximate location: Line 1, column 1
Error: Detected a byte order mark on the file, which should not be present.

And now?

Now remove the byte order mark from your file. Don't save CK2 files as unicode, only as Windows-1252.

 

[Okawoa]

hardest working man in showbiz

 

[Jamie550]

Is there any more detailed changelog than the one in Changelog.rtf?

Nope. If you want you can diff two versions and at least see what changed within the text files. Generally it consists of stuff people posted since the last release that I didn't reply to or that were eventually resolved.

hardest working man in showbiz

I don't understand?

 

[Okawoa]

Its a James Brown reference and you've been on this project for like 5 years so credits to you!

 

[Jamie550]

Its a James Brown reference and you've been on this project for like 5 years so credits to you!

Ah thanks! I'll just amend it to "hardest working woman in showbiz"

 

[Mart9203]

I can’t get the program to work. I keep getting this error message:
“Mod not found: MMM
Mods in system: My Modifications Mod
Problem with My Modifications Mod: Missing the 'path' attribute. Even if you have set 'archive', the Validator cannot read zip files. You must set the 'path' attribute and unzip the files to where the 'path' attribute specifies.”

I don’t know what “'path' attribute” is supposed to mean.
It isn’t a .zip, so I don’t know why it writes that.
In the .mod file, I have tried archive with and without .zip.
It’s the only mod I have in …\Documents\Paradox Interactive\Crusader Kings II\mod\

Please help, I’m stuck in both my mod and this program.

 

[IlikeTrains]

I can’t get the program to work. I keep getting this error message:
“Mod not found: MMM
Mods in system: My Modifications Mod
Problem with My Modifications Mod: Missing the 'path' attribute. Even if you have set 'archive', the Validator cannot read zip files. You must set the 'path' attribute and unzip the files to where the 'path' attribute specifies.”

I don’t know what “'path' attribute” is supposed to mean.
It isn’t a .zip, so I don’t know why it writes that.
In the .mod file, I have tried archive with and without .zip.
It’s the only mod I have in …\Documents\Paradox Interactive\Crusader Kings II\mod\

Please help, I’m stuck in both my mod and this program.

It will be easier to help if you send the .mod file.

 

[Mart9203]

Thanks, it turns out that I had used archive instead of path.
But now as I validate, I get a long list, of what I presume is the vanilla files being checked.
How do I get it to do my mod instead.

 

[IlikeTrains]

Thanks, it turns out that I had used archive instead of path.
But now as I validate, I get a long list, of what I presume is the vanilla files being checked.
How do I get it to do my mod instead.

Select "Show mod errors".

 

[Mart9203]

Thanks for the help and swift responses. It turns out I had systematically forgotten brackets.

 

[theStormWeaver]