The previous/next keys were already in the code I just moved it to the , & . keys... I could see using it if I'm building small rocks or something that I don't have an object for
ahh right I thought there might be a problem with the ` key with some other peoples keyboards... I'll add the tilde and escape key to the same function and see if that helps
Just quickly added some changes to the code and uploaded the newest version... I'm at work over the weekend so wont get chance to do much else
Changes:
Added "J" as shortcut for Jump
Pressing Enter should now OK the selection when entering the coordinates in the jump form
Shift + and - will fully zoom in /out
` None -> should now also be mapped to the ~ and Escape key... (note pressing the ~ on my keyboard doesn't work but I suspect it could be the key next to the numbers on other keyboards... it actually seems to be the '@ key on my keyboard) ... Having it on the Escape key seems good, if the ` or ~ keys cause problems we can just have it on the Escape key only
That feature Far once mentioned working in the editor, keys Z&X selecting previous and next tile respectively, was that really in the Xiyue editor, or maybe in iJam's? Really just selecting the next tile would be sufficient, not sure when you'd want to select previous one (maybe to list through tiles?). Since tiles are in sequences, simply jumping to the next one by keyboard key would be very handy for building objects from libraries. But that is probably not easy to implement.
I have another idea: some servers require you to fulfill some condition in order to unlock something. How about if on the standard euro map, you replace Stone temple in MW with a rock heap and put back limit on it so pple can't walk on it. Set some criteria for unlocking access to new temple and when it is fulfilled, map's Front layer automatically switches to Top layer on which Stone temple magically replaces the previous rock heap (the rock heap had back limit on it since you don't want to trap unsuspecting folks on it when the layer switches and temple appears on that place. But I think switching layers on the fly is probably just a pipe dream.
He did that in code I believe, the temple maze would be done by switching Front & Top layers which could be tied to any criteria besides day & night, as I outline here with ST example. Another example might be a Christmas event, if players collectively kill set number of mobs, the province map now with green grass will turn into snow covered one and with gift trees on it dropping boxes.
This is already achievable through the NPC functions, mainly use of Save/Load values & flags. I used it for a few things, my player housing endeavors, faction progression, etc. With a bit of forethought, I can use an item/click a signpost/complete a quest, etc and have a house appear with NPC's hanging out in-front, remove a blocked passage, change an area of trees into a town, you get the idea.
For instances in which a map change is necessary,
e.g making the whole map snow-covered or modifying interactable areas like rock pile vs ST entrance (This is not just changing the image, but move limits aswell). the principle is the same.
It has downsides of course, you have to choose between global changes or instancing for anything other than NPC's. But it's certainly do-able.
I guess I had a weak moment there, with the previous/next tile choice. Don't really know why I didn't test it on the keys I myself have suggested. Somehow I assumed it wasn't mapped.
the ' is the leftmost on standard keyboard, followed by 1, 2 (shift 2 writes @)
If the ` ~ key will work, I'd leave Escape to its normal function (closes Open map dialog window, probably also the Save dialog, if you change your mind)
That's why input from pple who use the editor in various ways is needed. What editor functions syou use always depends on the type of job. And this was not my case, although I know about the problem when one dark tile looks like another (typically when they make up dark periphery of a larger picture).
A minor issue concerning "F10 Show/Hide Front tag"
Pressing F10 to show front tag places focus on "File" (menu name) and a mouse click someplace in program is needed to deselect that menu name and put the focus back, so other hotkeys can work.
If you show front tag by pressing F10, it focuses 'File', pressing it again to hide the tag removes the focus from 'File'.
But if you take focus off it by mouse click and press F10 to hide tags, it selects 'File' again.
So, basically either action by F10 puts the focus on 'File' menu. It is the only F key that does that.
It would be handy if the editor didn't have to be restarted to show new object I dropped from windows directory into the Object folder.
I found in Object tab, if I use the button Delete to delete the selected object, it refreshes the tree view and the objects I dropped in from outside the program show up in the tree view.
It would be handy to refresh the Objects list some other way without having to delete objects, now that we know the editor can re-read the list to refresh it?
In the right click menu, the "Delete Selected Cell Data" is a handy but very dangerous command. It sits just next to "Save Selection As Object (.X)" in that menu and the deletion cannot be undone.
If you wanted to save something as object but by mistake clicked that delete command, whatever you had selected would just go puff and unless you could return to saved file, the new unsaved work you've just done would not be recoverable.
Before, the 'Save to Objects' button would save whatever was on the map, be it an object or even the whole map, it was ignoring the selection.
Saving as object whatever is selected by grasp rectangle is improvement (especially if you will give us back the ability to save the whole map by some command which would place the selection on the whole map).
Either that Delete Selection should have confirmation before deleting, or the command "Save Selection As Object (.X)" should be moved up in the menu, just below 'rerun' (which is misspelled as 'rerurn' - both 'undo' and 'rerurn' could be capitalized while you are in that menu)
Best would be doing both, the confirmation for Delete and moving the 'Save as object' up in that menu.
Exploring the middle layer issue on mir2, I opened mir3 Bichon 0.map in the editor and found some 95+% of objects are on middle layer.
Front layer seems to be used just to put some flower plants around the buildings so there is no tag conflict from close quarters, you see tree parts here and there on the front layer, when there was conflict like with a building, the parts of the tree were put on front layer.
Evidently, on mir3 the middle layer is the main object layer with front used only if the tags would conflict.
And the weird thing is, the map looks just fine loaded in the game client.
No tile strips shifted out of place, unless they are all shifted by the same amount (but I don't see buildings displaced relative to streets or pathways, which are back tiles). Also, we have these different tile strips that shift different amount from the rest and I don't see anything wrong with the map looks.
I added an object to the mir3 map, left one is on Front layer, right on Middle layer, and it is perfect in the MapEditor.
This is what it looks like loaded in the client, the right object is shifted down and in various amounts.
This is MapEditor with Front layer switched off, only the Middle and Back layers showing here.
Hiding the Middle layer and showing only the Back & Front layers, makes some 95% of objects on go away. Mir3 map is using the Middle layer as a main one.
For mir2, the Middle layer sits on top of the Back layer and serves to put ground details on the Back layer.
For mir3, the Middle layer is the equivalent of the Front layer of mir2 and its (mir3's) Front layer does not exist in mir2. I suggested above to call it Top layer.
Mir3 therefore doesn't have the Middle layer as mir2 has it, and it is up to anyone's guess why it is like that. Likely it goes back to when mir3 came out originally (circa 2002?).
Hope you all like this update... I'm happy I managed to get it working so quickly
Note: I had a crash when I first added the TileCutter and tried to save the .lib ... but after ~20 more test saves I can not replicate the crash so I'm not sure what caused that problem... let me know any details if you find the problem
Hope you all like this update... I'm happy I managed to get it working so quickly
Note: I had a crash when I first added the TileCutter and tried to save the .lib ... but after ~20 more test saves I can not replicate the crash so I'm not sure what caused that problem... let me know any details if you find the problem
I've been looking at the code for drawing the Middle and Front Images
What are the index numbers for the images you used in the example? ... or better question I just need the cell info for one of the cells I can work out the rest once I know what file to look in hehe
EDIT: Nevermind I found the object... its in the Mir3 images (I think I understand what is going on now)
Isn't re-saving the mir3 map after you add those middle objects just changing the type to crystal?
So it's no longer Mir3 (which doesn't use 'middle'.)
it is using the index number to define the tile image as coming from mir3 (so the map type shouldn't make a difference) and then running different draw code for the middle layer (I'm just trying to work out why it is set up that way by looking at the code)
I suspect that any tile strips from mir2 files load fine on the middle layer? (with an index of less than 200) and I'm guessing custom tile strips would break too as they have an index greater than 200... but all images that match the cell size exactly should work as it only runs the other code if it is a tile strip different size than the cell size
EDIT: I'll have another look at this tomorrow... my head is hurting now I can't think why if that is a mir3 map that the trees don't mess up as they are tile strips and if they are from mir3 should have index of 200 or more... I'll have to get a mir3 map and a crystal client to test it on then I should be able to see why it is offsetting the tile strips and not others (I do think it is something to do with the Draw and DrawUp methods)
The code in the map editor doesn't have the mir3 middle layer draw code (and DrawUp) so that is probably why it is viewed ok in the editor as it just draws as middle layer normally
I would really love to make a way to auto generate maps but to stitch all the tiles needed for the objects or by using pre created objects find a way of them not overlapping or causing problems blocking player movement and trapping players would be quite a challenge and require very specific code for our files (I doubt I am good enough at coding currently to do this)
I do however have an idea I want to explore that is to set up some maps in the random style like in Diablo3... I originally wanted to do this so it creates maps on the fly in game but getting the maps populated with mobs and sending the same maps to every player seemed too complicated for me to do so I have now decided it would be better to do the map creation in the map editor and create normal maps that way
Thats one of the reasons I am updating the map editor... so I can make the objects I need and add some data so the objects know how to fit together correctly... then create the code to randomize a map
It is actually WeMade-Mir2 Objects16.Lib, #s 4291-4326
I picked that library at random while testing building objects this weekend, and found it handy to use the built object for this example, just redone this one on middle layer so I have comparison examples.
Isn't re-saving the mir3 map after you add those middle objects just changing the type to crystal?
So it's no longer Mir3 (which doesn't use 'middle'.)
Not sure I understand what you are saying here. I reopened the saved map in editor and it still has object mainly on the middle layer as before. The original objects on mid layer stay untouched, only that one I add on mid layer is misplaced. Maybe the mapeditor is saving only the changes to the map and it uses different saving method from that used originally when the map was made?
Edit: the (so called) 'middle' layer on mir3 maps is actually different layer from the middle layer on mir2 maps. Somehow, the map editor treats that mir3 layer as mir2 middle layer in that it shows middle tags for its tiles, objects, shows and hides its objects by the 'show/hide middle layer' command, but when loaded in the game client, it displays it as front layer. If you ignore the miss-placement on map, you can hide behind the house and all objects on that mir3 middle layer which you can't do with mir2's middle layer (it being ground layer, it always shows you on top of it).
I will test later today if I can actually go behind the two testing objects I placed on the map. I expect to be able to go behind the left one (it being on front layer), but appear on top the right hand one, it being on mir2 middle layer.
----------------
The map source is from the Korean Happy server opening Season4, about a year old. BTW the server went offline I think due to crack down by Korean 'GRAC' authorities. Inside its clients map folder are two folders named Chronicles and Mir3 and I took the Bichon map from the Mir3 folder, moved it up to Map root folder and edited it, saved it, loaded in client, when reopened in the editor it still uses the same layers, the process of saving the map did not alter that. That's the whole story. I don't get why they had nested folders inside Map folder in the client. I believe the client can only load maps in root map folder location. It seems they foisted on pple downloading their client all maps they could put their hands on, while only a fraction of them was actually in use.
BTW not related but I can't open any libraries from Happy Data. I get the message: "Unable to read beyond the end of the stream."
Code:
************** Exception Text **************
System.IO.EndOfStreamException: Unable to read beyond the end of the stream.
at System.IO.BinaryReader.FillBuffer(Int32 numBytes)
at System.IO.BinaryReader.ReadInt16()
at LibraryEditor.MLibraryV2.MImage..ctor(BinaryReader reader) in C:\Users\support_2\Desktop\clients\mir2\source\Test.Environment\LibraryEditor\MLibraryV2.cs:line 279
at LibraryEditor.MLibraryV2.CheckImage(Int32 index) in C:\Users\support_2\Desktop\clients\mir2\source\Test.Environment\LibraryEditor\MLibraryV2.cs:line 120
at LibraryEditor.MLibraryV2.Initialize() in C:\Users\support_2\Desktop\clients\mir2\source\Test.Environment\LibraryEditor\MLibraryV2.cs:line 62
at LibraryEditor.MLibraryV2..ctor(String filename) in C:\Users\support_2\Desktop\clients\mir2\source\Test.Environment\LibraryEditor\MLibraryV2.cs:line 31
at LibraryEditor.LMain.openToolStripMenuItem_Click(Object sender, EventArgs e) in C:\Users\support_2\Desktop\clients\mir2\source\Test.Environment\LibraryEditor\LMain.cs:line 240
at System.Windows.Forms.ToolStripItem.RaiseEvent(Object key, EventArgs e)
at System.Windows.Forms.ToolStripMenuItem.OnClick(EventArgs e)
at System.Windows.Forms.ToolStripItem.HandleClick(EventArgs e)
at System.Windows.Forms.ToolStripItem.HandleMouseUp(MouseEventArgs e)
at System.Windows.Forms.ToolStrip.OnMouseUp(MouseEventArgs mea)
at System.Windows.Forms.ToolStripDropDown.OnMouseUp(MouseEventArgs mea)
at System.Windows.Forms.Control.WmMouseUp(Message& m, MouseButtons button, Int32 clicks)
at System.Windows.Forms.Control.WndProc(Message& m)
at System.Windows.Forms.ToolStrip.WndProc(Message& m)
at System.Windows.Forms.ToolStripDropDown.WndProc(Message& m)
at System.Windows.Forms.NativeWindow.Callback(IntPtr hWnd, Int32 msg, IntPtr wparam, IntPtr lparam)
************** Loaded Assemblies **************
mscorlib
Assembly Version: 4.0.0.0
Win32 Version: 4.6.1038.0 built by: NETFXREL2
CodeBase: file:///C:/Windows/Microsoft.NET/Framework64/v4.0.30319/mscorlib.dll
----------------------------------------
LibraryEditor
Assembly Version: 2017.6.23.1821
Win32 Version: 2017.6.23.1821
CodeBase: file:///D:/Documents/Downloads/Mir2-Crystal%20files%20%20-%20Server%20%26%20Client%2010-08-2017/client-040817/LibraryEditor.exe
----------------------------------------
System.Windows.Forms
Assembly Version: 4.0.0.0
Win32 Version: 4.6.1038.0 built by: NETFXREL2
CodeBase: file:///C:/Windows/Microsoft.Net/assembly/GAC_MSIL/System.Windows.Forms/v4.0_4.0.0.0__b77a5c561934e089/System.Windows.Forms.dll
----------------------------------------
System
Assembly Version: 4.0.0.0
Win32 Version: 4.6.1038.0 built by: NETFXREL2
CodeBase: file:///C:/Windows/Microsoft.Net/assembly/GAC_MSIL/System/v4.0_4.0.0.0__b77a5c561934e089/System.dll
----------------------------------------
System.Drawing
Assembly Version: 4.0.0.0
Win32 Version: 4.6.1038.0 built by: NETFXREL2
CodeBase: file:///C:/Windows/Microsoft.Net/assembly/GAC_MSIL/System.Drawing/v4.0_4.0.0.0__b03f5f7f11d50a3a/System.Drawing.dll
----------------------------------------
System.Configuration
Assembly Version: 4.0.0.0
Win32 Version: 4.6.1038.0 built by: NETFXREL2
CodeBase: file:///C:/Windows/Microsoft.Net/assembly/GAC_MSIL/System.Configuration/v4.0_4.0.0.0__b03f5f7f11d50a3a/System.Configuration.dll
----------------------------------------
System.Core
Assembly Version: 4.0.0.0
Win32 Version: 4.6.1038.0 built by: NETFXREL2
CodeBase: file:///C:/Windows/Microsoft.Net/assembly/GAC_MSIL/System.Core/v4.0_4.0.0.0__b77a5c561934e089/System.Core.dll
----------------------------------------
System.Xml
Assembly Version: 4.0.0.0
Win32 Version: 4.6.1038.0 built by: NETFXREL2
CodeBase: file:///C:/Windows/Microsoft.Net/assembly/GAC_MSIL/System.Xml/v4.0_4.0.0.0__b77a5c561934e089/System.Xml.dll
----------------------------------------
Accessibility
Assembly Version: 4.0.0.0
Win32 Version: 4.6.1038.0 built by: NETFXREL2
CodeBase: file:///C:/Windows/Microsoft.Net/assembly/GAC_MSIL/Accessibility/v4.0_4.0.0.0__b03f5f7f11d50a3a/Accessibility.dll
----------------------------------------
Also if I put Happy Data folder into my client (while renaming the existing Data), the mapeditor doesn't see its libraries, there are just three empty folders in WeMade-Mir2 (Tiles, Smtiles, Objects), Shanda-Mir2 has in addition empty AniTiles1 folder, while both Wemade & Shanda Mir3 don't have even that, those are completely empty.
Reason I tried to open libraries in the first place is because its Wemade Mir2 libraries have additional 24, 25 & 26 libraries and the 24&25 have 36 MB, quite large. I was curious what might be in them.
Also Crystal client Data has nothing in Shanda-Mir3 lib folder while Happy Data does have it full, however it seems to be just a copy of Wemade-Mir3 folder.
But here I am going beyond my know how, maybe the Happy libraries are encrypted or they are v1 and we now use v2? But I don't know anything about this version change apart from noting some talk here on forum about these versions. Thought LibEditor has a conversion for it but is is only for Will and Wix or whatever formats but nothing to convert v1 to v2, I might be missing a program to do that?
ahh right... erm I'll have to check the images and see why it is doing that then (I thought I was close to understanding what was going on if it was mir3 images lol)
EDIT: had a quick look at the code now I have the correct info for the images that were messing up... I know what is causing the problem now and can fix it... the middle layer doesn't do the same size check as the front layer in the client code... I will take another look tomorrow and see if I can make changes
I'm not sure if it is ment to work how it is in the client or map editor... is all normal (mir2) middle layers images the same size as one cell and not larger images like tilestrips? except mir3 maps which do have tilestrips on the middle layer? ... just wondering if there is a reason it could draw the image from the top (it may be that it is ment to be behind the player and this is how that is achieved?) wouldn't want paths that use tilestrips to suddenly be in front of the player lol
The code in the map editor doesn't have the mir3 middle layer draw code (and DrawUp) so that is probably why it is viewed ok in the editor as it just draws as middle layer normally
I will try to restate my closing remarks I made at the end of that post above with screenshots, regarding the use of layers in mir2 vs mir3.
Mir3 doesn't have the equivalent of mir2's middle layer (where it serves to make ground features using single tiles only). Mir3 thus only has back layer for the ground and that's it.
Mir2 places objects on front layer whereas mir3 places them on middle layer and uses the front layer only to deal with the situation when tags are conflicting due to objects being too close to each other.
I am not sure that mir2 mid layer should be 'fixed' to properly place tile strips so they don't get displaced in the game client (Edit:in light of subsequent posts, I changed my mind). That role should be played by a layer alternative to the front one. I will make a table of layer equivalence in mir2 & mir3 and mark the middle layer 2&3 where the numbers refer to mir2 & mir3 respectively (to distinguish different use of this layer by the two mir versions)
- Back layer (same in both Mir2 & Mir3)
- Middle layer2 (exists only in Mir2 as and alternative layer to the Back one - it serves to put various features on (over) back layer. You could call it a 'cosmetic layer' for the back layer).
- Middle layer3 (exists only in Mir3 as a main layer for all objects except where there is a tag conflict with other objects)
- Front layer (exists in both Mir2 & Mir3 but while it is the main layer for Mir2 objects, Mir3 uses it only to resolve the tile conflicts. Mir2 doesn't have this option, can't place a tree to be partly cover a house in front of it as mir3 can - as you can see it in the screens I posted above, with the pink tree planted right behind a house but covering its roof by its middle section (because it is on Front, the topmost layer to resolve its midsection tag conflicts with those of the house) while its left and right ends on both sides are hidden by the roof since they are on middle layer with the house and don't have tag conflict.
In effect, to use Mir2 terminology as a base, it is as if Mir3 has elevated the Mir2's Middle layer to its main object layer and it has an extra layer (which it calls 'Front' layer) which sits above its main object Middle layer3 to resolve tag conflicts (so Mir3's Front layer is the Topmost layer, second of the two object layers, and it has no Middle layer as we know it from Mir2 maps where it is used to place various features on (over) the back layer).
I think that if something would be done about the layers, it might be better to add an entirely new, additional layer, the topmost layer, to be equivalent of mir3 'Front' layer and leave the middle layer as it is now. Elevating the middle layer function, to make it serve as a layer below the mir2's front layer (making it into a second front layer) would likely not be easy but mainly it would be confusing. If raised to serve as object layer, it should be elevated to sit on top of the current front layer, not below it.
Ideal case would be to add a new topmost 'Top' layer in the mapeditor and if the middle layer was fixed to allow placing tile strips on it, that would be just a bonus. Then the middle layer would be an extra layer over what mir3 maps have, adding an extra functionality that mir3 doesn't have. But if the middle layer was fixed to allow proper placing of tile strips, it would at least partly obviate the need for the second, topmost object layer.
The fixed Middle layer and the topmost Top layer would not have identical function since the Middle layer is really a ground 'back' layer, still some difficulties building maps could be solved with either layer.
I will upload the version they called Season4 later (Client from August 18 2018 but here is the link to the Season1 I uploaded April 30 2017 for us English players to join in on the fun. Season2 client I am using right now may have different lib files in it, didn't check the files in the Season1 (that was when the server first opened for us here on LOMCN). But there is big difference in client size, while the season1 (but it may have actually been season3 when we were invited on it, I just call it season1 because it was the first for us here) is only 3GB, the season4 is 8.82 GB.
I was browsing maps in it last night and there are repeated multiple copies of maps all over, which is what makes the map folder itself quite huge - 2.63 GB vs 1.17 GB in the default crystal client.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
I'll see what else needs doing then maybe upload the source code next week sometime
