TW Map Reader v2.24 [1-Jul-2015: update]

[FireFreak111]

Thanks for that simple change, should help me keep realizing when I accidentally saved layers while not making me relaunch the program.

For the PNG the more I compare I fail to see a difference, so I will drop this one . No reason to continue with it.

And for the more pixels, yeah that's why it isn't worth it. I use layers anyway currently to match up the files. Thank-you for your continual support, I love this program.

[Withwnar]

No dramas. The PNG export is provided by inbuilt java functionality; I don't know much about how or what it is doing except that it is possible that it behaves differently on other PCs, if they have different java libraries installed. I have noticed that they take a long time to open up in Windows' image viewer but Photoshop has no problem with them.

[Withwnar]

v2.8 released. From the Change Log:

• Fixed FE resize bug: non-FE zooms were stretching to fit other map proportions, instead of FE map proportions.
• Fixed "starting regions" bug: was including all fort and watchtower regions as slave starting regions.
• Added state feature: save settings for reloading (reapplying) at a later time.
• Realtime preview for all colour selections: while the colour chooser dialog is open it applies all changes to the map view, instead of only doing it after the dialog is closed.
• If 200 RGB errors are encountered while reading a map then it stops trying. Avoids excessive wait times while it builds thousands of error messages.
• Added HSB option for region highlight, colouring the regions without painting over them. (Painting over them - the "overlay" option, previously named "fill" - can mask their detail, depending on opacity level).
• Added "flash" option for settlements view; they now flash only if you want them to.
• Swapped the order of map views: "settlements" is now after "region highlight".
• HSB is now the default picker in the colour chooser.

Some useful additions here, most of all the "state". I really hated setting it up with map views, opacity adjustments and region highlight colours etc. that I liked, only to lose them when I closed the tool and having to redo it all when I came back later. Those days are over.

The new HSB option is quite effective, especially over the FE map...

Attachment 263159

...but it does suffer a bit from processing time. Not too bad but moving the map around or realtime-colour-adjustment can be a bit laggy.

[FireFreak111]

Great update. One suggestion though. With the 200 RGB errors, after the message it doesn't log the ones it found. Just thinking of a scenario that is sort of likely, a person accidentally replaces all 000000 Wilderness RGBs with some black variant from map_heights. The tool would basically just be saying 'Something went wrong!!! We know but we know too much so sorry!' Thanks for fixing the issue though, and these new features will be very useful when I move onto region work.

[Withwnar]

It still reports those first 200 so you could fix those and reload... it will tell you of the new first 200, and so on.

[FireFreak111]

Can I suggest another error message to show? If the descr_aerial_map_ground_types.txt file has a non-existant texture path pop-up an error when the user ticks either of the texture options that renders the map. It would help others if they come across the issue I did (non-existant paths causing CTD when you go near it). Another good error message you might want to add is if a river doesn't have a white pixel on either end, as the river simply doesn't render, no CTD it just doesn't appear on the map.

This tool helped me alot though when adding the core regions that had custom settlements or were of significant importance.

Also I just saw that it does report the 200 errors, my bad..

[Withwnar]

That is heading down the road of "map validator" which is not its purpose. It is a map viewer; the error messages that it reports are only there to explain why it can't enable some maps/functions, not to tell you what you've done wrong as such.

[FireFreak111]

Well when the Reader tries to read a nonexistant texture, it just shows up pink (and if its on a non-center tile it doesnt show at all). An error simply stating why (nonexistant file) isn't too unreasonable.

[irishron]

Pink tiles? the oldskool way before these tools existed to find pixels with wrong colors and the only thing you have are the coordinates to go through the layers one at a time until you find it or them.
If his tool is supposed to just read the map and get me in the ballpark, I'm happy. I can do the grunt work from there.

[Withwnar]

Well when the Reader tries to read a nonexistant texture, it just shows up pink (and if its on a non-center tile it doesnt show at all). An error simply stating why (nonexistant file) isn't too unreasonable.

To be fair, you would need to know how much effort on my part a request would require before you describe it as "unreasonable".

However, in this case... these errors are already being found and reported but there's a bug which is preventing the Errors tab from being enabled so you can't see them.

I misread your earlier post: I thought you meant show errors if a climate is missing a ground type entry in DAMGT (that it will not do) OR if the texture file can't be found (the bug has existed for so long that I had forgotten that I had already done this).

What do you mean by that non-center comment?

[FireFreak111]

I guess I am being unreasonable for saying it would be unreasonable

Some ground_type tiles (pixels) don't show up in the Reader simply because of their positioning, which isn't the Readers fault in the slightest, when you downscale without interpolation (which is what map_regions, map_features and the reader do), of course some pixels will be lost. The problem is when the pixels not showing in the reader are the ones that have a pink tile .

[Withwnar]

I guess I am being unreasonable for saying it would be unreasonable

Yep. Suggestions are always welcome. Their reasonableness, and whether I choose to do them, is my call.

Some ground_type tiles (pixels) don't show up in the Reader simply because of their positioning, which isn't the Readers fault in the slightest, when you downscale without interpolation (which is what map_regions, map_features and the reader do), of course some pixels will be lost. The problem is when the pixels not showing in the reader are the ones that have a pink tile

Still not following. The only downscaling by this tool is the FE zoom but you shouldn't be using that for map checking, for that very reason: things get lost/blurred in the scaling. All other maps are simply 1:1 with the original image (even pixels discarded in some maps of course), upscaled by a factor of 2, 3, 4 or 5 for the different zooms; no pixels are lost during upscaling, they are just larger.

[FireFreak111]

Ground_types and map_heights and climates downscaling to the resolution of map_regions or features is what I am talking about

[Withwnar]

Those maps aren't downscaled, they just have every second pixel discarded. The game ignores them (so does the tool) so even if you did have a 'bad' pixel in one of those that resulted in a missing texture, it isn't a problem because the game doesn't see it anyway: that tile doesn't exist. Or rather: that pixel does not represent a tile, it is a 1 pixel gap between tiles on those maps. Therefore will never appear in the tool and therefore could not possibly be pink.

[irishron]

@FireFreak, over at the .Org is a full thread in the Medival2 Mods section on how the different layers are stacked and how the pixels of map_regions overlays the map_grountypes for instance that predates the 1.2 patch.

I am not going searching for something that is as good as six years old but it maybe worth your while. Withwnar is right on how those pixels work and not exactly as you and I see them.

[Withwnar]

There is also this, in case you haven't seen it: http://www.twcenter.net/forums/showthread.php?t=177334. What the game and the tool see are only the ones that make it through the gaps.

[irishron]

I forgot about this tool. Thank you for the reminder.

[Gigantus]

The 'Map Strainer' I call it - can be hard on the eyes, mind you.

[FireFreak111]

Thanks for all this information. I use the odd pixel tool already for forests, but now I know the extent it can be used

[Withwnar]

v2.9 released.