MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

New build announcements plus site news and changes.

Moderators: dorpond, trevor, Azhrei, Craig

User avatar
Azhrei
Site Admin
Posts: 12086
Joined: Mon Jun 12, 2006 1:20 pm
Location: Tampa, FL

MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by Azhrei »

This is the "oooo, not another one?!" build.

Otherwise known as, Release Candidate 5.

I know that we still have two issues that are biggies to fix: a problem with the GM drawing global fog-of-war and it being shared incorrectly amongst tokens, and a problem with token movement distances not being displayed.

I have identified a process that recreates the first problem so it's just a matter of tracking it down. The second one hasn't been clearly identified yet as to exactly what configuration causes the problem, but we'll find it.

Please check the ChangeLog from the bottom right corner of the Main Page of the wiki -- there have been a huge number of bug fixes in this build. In the "new stuff" category we're limited to translations -- we can add new translations and do localization work without getting complicated in the code. We have a new translation for Japanese and our long-time supporters have updated the German, French, and Spanish translations. (I can't speak any of those languages but I think it's soooo cool to work on a product that's been translated to so many languages!) The localization work will continue -- albeit a little slower -- even once 1.3 goes Final. If you're interested in doing translation work, there's a Sticky thread in the MapTool forum with instructions.

Note that this build requires Java 6! I made an easy fix and am looking for code to replace the easy fix with something that is Java 5-compatible. The ChangeLog specifies which bug this is related to.

Our primary reason for this build: to give everyone a chance to check out those fixes while we (the dev team) hash out the last couple of issues. We might as well both be working on stuff in parallel, right?

This version should be just as stable as b81 (I've noticed a lot of people using b81 in the past couple of weeks -- thank you!) with lots of additional fixes, so I expect it to be very solid.

User avatar
aliasmask
RPTools Team
Posts: 9023
Joined: Tue Nov 10, 2009 6:11 pm
Location: Bay Area

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by aliasmask »

Removed small dot drawn at (0,0) by the RadiusTemplate (this bug goes back to at least 1.3b34!)
Heh, I've seen this but I just though it was me mistakenly drawing it. Occasionally, I'd say, "hey, when did I do that."

I've been getting a mystery NPE error (b81, about 3 times now), but I couldn't reliably duplicate it. I have a feeling it's java memory related because it always happened when I used another application while MT was open. First time was when I was upgrading my window media player and the second time was when I used Screen2Exe for the first time. Both work fine now with MT with no crashes, but it's strange that MT would crash rather than other programs if it was a memory error. I don't recall the circumstance for 3rd crash, but I know I was doing something else outside of MT. I don't think we can address this problem until it happens again, so if it does, then I'll post the details.

User avatar
silversonic
Cave Troll
Posts: 59
Joined: Thu Oct 15, 2009 6:42 pm
Location: Cleveland, OH

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by silversonic »

I've played around with it a little bit, and so far 82 works nicely on my macbook. I'll report back when I run a session.
I don't want the world, I just want your half.

User avatar
aliasmask
RPTools Team
Posts: 9023
Joined: Tue Nov 10, 2009 6:11 pm
Location: Bay Area

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by aliasmask »

Just tested Wiki: currentToken() and Apply to Selected Tokens with duplicate token names... works like charm, thanks.

User avatar
Azhrei
Site Admin
Posts: 12086
Joined: Mon Jun 12, 2006 1:20 pm
Location: Tampa, FL

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by Azhrei »

aliasmask wrote:I've been getting a mystery NPE error (b81, about 3 times now), but I couldn't reliably duplicate it.
No information in the log.txt? That's weird.
aliasmask wrote:Just tested Wiki: currentToken() and Apply to Selected Tokens with duplicate token names... works like charm, thanks.
Excellent. :)
silversonic wrote:I've played around with it a little bit, and so far 82 works nicely on my macbook. I'll report back when I run a session.
Yes, please do!

User avatar
francogp
Cave Troll
Posts: 66
Joined: Thu Feb 02, 2006 1:06 pm
Location: Mordor

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by francogp »

hi,
I can not see any difference in translation between RC4 and RC5 version (spanish).
The options and campaign preferences remain fully in English.

User avatar
aliasmask
RPTools Team
Posts: 9023
Joined: Tue Nov 10, 2009 6:11 pm
Location: Bay Area

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by aliasmask »

Azhrei wrote:
aliasmask wrote:I've been getting a mystery NPE error (b81, about 3 times now), but I couldn't reliably duplicate it.
No information in the log.txt? That's weird.
Actually I posted the log in the previous b81 thread: http://forums.rptools.net/viewtopic.php ... 45#p186417

User avatar
Azhrei
Site Admin
Posts: 12086
Joined: Mon Jun 12, 2006 1:20 pm
Location: Tampa, FL

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by Azhrei »

francogp wrote:hi,
I can not see any difference in translation between RC4 and RC5 version (spanish).
The options and campaign preferences remain fully in English.
Yep. That's why I said:
Azhrei wrote:The localization work will continue -- albeit a little slower -- even once 1.3 goes Final.
The implication being that we're not done yet.

Only one dialog has been localized. They're a pain in the butt to do currently. :( Every string that appears in the dialog has to be looked up in the translation file and then replaced with whatever is found. There is only one dialog that does this currently and the code is pretty ugly, so my plan is to create a separate function that knows how to do this work in some kind of automated way (I have some ideas for that) and then add a call to that function in every piece of code that opens a dialog.

Needless to say, this is slow going. And I've been working on bug in the actual operation of the program and leaving the cosmetic stuff for later.

If anyone with some Java experience wants to take a shot at it, I can outline my plan in more detail. Otherwise, I'll get to it eventually...
aliasmask wrote:Actually I posted the log in the previous b81 thread:
Ah, okay. That's not a NullPointerException (NPE), that's an IllegalStateException (ISE?). I commented a few posts down that this is a Java/OS bug, not a MapTool bug.

You hypothesized that it could be memory related, and I agree. It could be application memory, but more likely it's low GDI memory in Windows. This article is from a developer who hit this limit and how he diagnosed it. He refers to a free diagnostic tool, GDIUsage, and says that even Task Manager can provide some of the information needed to determine if this is a problem.

His application was .Net-based but it could happen to any application that talks to the OS using the Win32 APIs... which includes the Java runtime.

So if you run those tools and find that you're having the GDI problem, let me know which resource it is and we can report the bug more specifically to Oracle and perhaps find a way to work around it in MapTool...

The obvious solution is to use an OS that doesn't statically allocate memory or array sizes. Linux is probably the best choice as just about everything is done dynamically. OSX might be a good second choice, but I'm not familiar enough with Aqua/Carbon to know how they manage memory internally.

osman
Giant
Posts: 100
Joined: Sat Apr 12, 2008 12:23 pm
Location: Latham, NY

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by osman »

Hmmm, while testing some of my new framework out, I stumbled onto a fairly basic (and easy to reproduce) bug with multiple-selection token movement. The oldest build I have on my PC is b63, and it's present at least that far back. To reproduce (on Windows XP, 32-bit, at least):

Open a fresh copy of MT. Drag any two tokens onto the default map from the resource library. Select them both. Try two experiments:

1) Move them such that they both end their movement a good distance away from where they both started. Everything works. In particular, they both remain selected.

2) Again, with both tokens selected, move them such that one token ends the movement in the same space where the other token started. That token will become deselected. You can also drag them as if you're moving them, and then position them back where you started. In this case, one of the tokens may become deselected.

Known issue? Not incredibly important, but I could see it being sort of annoying in certain situations.

User avatar
aliasmask
RPTools Team
Posts: 9023
Joined: Tue Nov 10, 2009 6:11 pm
Location: Bay Area

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by aliasmask »

osman wrote:Hmmm, while testing some of my new framework out, I stumbled onto a fairly basic (and easy to reproduce) bug with multiple-selection token movement. The oldest build I have on my PC is b63, and it's present at least that far back. To reproduce (on Windows XP, 32-bit, at least):

Open a fresh copy of MT. Drag any two tokens onto the default map from the resource library. Select them both. Try two experiments:

1) Move them such that they both end their movement a good distance away from where they both started. Everything works. In particular, they both remain selected.

2) Again, with both tokens selected, move them such that one token ends the movement in the same space where the other token started. That token will become deselected. You can also drag them as if you're moving them, and then position them back where you started. In this case, one of the tokens may become deselected.

Known issue? Not incredibly important, but I could see it being sort of annoying in certain situations.
Confirmed. Wow, that's a random bug. Nice catch. I also noticed the time after release seems to take longer when that happens. Also note that it will depend on the token z-order too. If dragging with a token with a lower z-order that ends over the start position of a higher z-order you see the bug, but if the z-order is higher, then it doesn't happen.

osman
Giant
Posts: 100
Joined: Sat Apr 12, 2008 12:23 pm
Location: Latham, NY

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by osman »

Yeah, I can't believe I just sort of stumbled on to it... I'm working on a board-game implementation that supports automatically stacking and splitting of token 'stacks' via onTokenMove, and I was just trying to stress test my macro code with a bunch of edge cases. I initially assumed I had done something screwy (one of my macros previously had some selection altering logic in it), but then I tested on fresh runs with no campaign loaded, so realized I had hit an actual bug.

Edit: The bug is even more basic than I previously stated. Create two tokens, then disable snap to grid (this makes it easier to test/see). Drag one token so that it overlaps the other (generally such that the mouse cursor ends over the 2nd token). Based on Z-order, the selection will actually switch to the 2nd token when you release the token to complete the move.

User avatar
Azhrei
Site Admin
Posts: 12086
Joined: Mon Jun 12, 2006 1:20 pm
Location: Tampa, FL

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by Azhrei »

osman wrote:Edit: The bug is even more basic than I previously stated. Create two tokens, then disable snap to grid (this makes it easier to test/see). Drag one token so that it overlaps the other (generally such that the mouse cursor ends over the 2nd token). Based on Z-order, the selection will actually switch to the 2nd token when you release the token to complete the move.
All I can do is echo aliasmask: "Wow." That's amazing that someone actually found this and then was able to isolate the conditions needed to reproduce it! This kind of bug has the potential to go unsolved for a long time...

Also, Phergus has reported elsewhere that the MapExplorer is showing the wrong information. It's already been fixed and is in SVN as revision 5605.

User avatar
Natha
Dragon
Posts: 733
Joined: Sat Oct 11, 2008 3:37 am
Location: Limoges/Guéret, France
Contact:

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by Natha »

There's a problem with the latest french translation. (note : I'm credited for it though I didn't make it in time to send it to Azhrei for b82 ???).

Character enconding is wrong : accents and other special character are not displayed correctly.

Example :
"CoordonnÈes" instead of "Coordonnées"
or
"Forcer les joueurs ‡ se placer" instead of "Forcer les joueurs à se placer"

I'll send another file as soon as possible.

EDIT : can't remember : do we have to always double single quote ?
ImageImageImageImage

User avatar
Azhrei
Site Admin
Posts: 12086
Joined: Mon Jun 12, 2006 1:20 pm
Location: Tampa, FL

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by Azhrei »

Natha wrote:There's a problem with the latest french translation. (note : I'm credited for it though I didn't make it in time to send it to Azhrei for b82 ???).
Oops. Well, you'll get credit for the new one you're sending in. I'll have to find out who did the last one. Thanks for the note.
Character enconding is wrong : accents and other special character are not displayed correctly.
Hm. I'll check that. I might need to run the file through native2ascii. What character encoding is the author likely to be using? ISO-8859-1? Or UTF-8? I'll try both and see how it comes out...
EDIT : can't remember : do we have to always double single quote ?
Yes. There are some weird rules in some cases; there's a URL reference in the translation file that points to the Java documentation on the web. Not the easiest thing in the world to read, but you can skip the programming stuff and just concentrate on the contents of the translation file.

Edit: Okay, it appears that changing the (assumed) incoming encoding changes the resulting Unicode characters. I probably did the conversion assuming UTF-8 because ISO-8859-1 produces a different result. Here's an example of the differences (a minus sign prefix means UTF8 and a plus sign prefix means iso8859):

Code: Select all

--- i18n_fr_utf8.txt        2011-01-21 23:23:11.000000000 -0500
+++ i18n_fr_iso8859.txt     2011-01-23 11:16:52.000000000 -0500
@@ -4,8 +4,8 @@
 menu.tools=&Outils
 menu.help=&Aide
 menu.zoom=&Zoom
-menu.window=&Fen\u00cdtre
-menu.recent=&Campagnes R\u00c8centes
+menu.window=&Fen\u00eatre
+menu.recent=&Campagnes R\u00e9centes
 menu.map=&Carte
 menu.export=Exporter

@@ -39,17 +39,17 @@
 action.saveCampaignAs=Enregistrer Campagne &Sous...
 action.saveCampaignAs.accel=A
 action.saveMessageHistory=Enregistrer l'&Historique des Messages
-action.exportScreenShot=Capture d'\u00c8cran
+action.exportScreenShot=Capture d'\u00e9cran
 action.exportScreenShot.accel=Shift S
-action.exportScreenShotAs=Capture d'\u00c8cran Sous...
+action.exportScreenShotAs=Capture d'\u00e9cran Sous...
 
-action.addIconSelector=Ajouter \u2021 la bib&lioth\u00cbque de Ressources...
-action.addDefaultTables=Ajouter des tables par d\u00c8faut
+action.addIconSelector=Ajouter \u00e0 la bib&lioth\u00e8que de Ressources...
+action.addDefaultTables=Ajouter des tables par d\u00e9faut
 #action.addIconSelector.accel=ctrl shift I
 
-action.serverStart=D\u00c8marrer un Serveur
-action.clientConnect=Se connecter \u2021 un Serveur
-action.clientDisconnect=Se d\u00c8connecter du Serveur
+action.serverStart=D\u00e9marrer un Serveur
+action.clientConnect=Se connecter \u00e0 un Serveur
+action.clientDisconnect=Se d\u00e9connecter du Serveur
 action.showServerInfo=Information de &Connexion
Not shown in the above list is the following change, which is the one you referred to in your post:

Code: Select all

-action.showCoordinates=&Coordonn\u00c8es
+action.showCoordinates=&Coordonn\u00e9es
Notice how the \u00c8 becomes \u00e9? The first assumed a UTF-8 encoding, the second assumed ISO-8859-1. If the second one is correct, then I can fix it on my end. :) And I'll add a note to the English translation file that submitters should always send their files to me using UTF-8!

osman
Giant
Posts: 100
Joined: Sat Apr 12, 2008 12:23 pm
Location: Latham, NY

Re: MapTool 1.3 Release Candidate Build 82 (1.3-RC5)

Post by osman »

Another one (that I couldn't find in my searches of the forums): Hot Keys for Lib: macros don't appear to work until that Lib token has been selected at least once. This could be known/expected? Easy to reproduce, just make a new campaign, drop in a single token, rename it to 'Lib:Something', add a single macro with a hot key that prints a message. Save the campaign, then re-open it. The hot key won't do anything until the first time you select the token. After that, you can use the hot key whenever you'd like, as expected. Should I only use hot keys on campaign/global macros?

Edit: No, it's worse. It only starts working when you select the token, AND have the 'Selection' window open & visible. Hmmm...

Post Reply

Return to “Announcements”