Bug Reports (snapshot builds)

[Chris H.]

Then, same in the deck editor, when you resize a column of one of the two tables, all other columns used to resize in percentage, but now only the closest column resizes; the previous way was far way better, I think.

Hrm, I had just changed this because I found it much easier to size the columns accurately when the other columns stayed put. Does anyone else have opinions on this behavior?

 
I found that I was able to adjust to the new way fairly quickly. My monitor is 1920 x 1200 and I narrowed most of the columns and used the extra pixels to widen the name and type columns.

[Chris H.]

I could not reproduce this (see screenshot) -- I presume you're entering a word that the gatecrash cards contain, right?

 
The Gatecrash cards do not yet have the Oracle info added to the card scripts. I guess that this will be added in a few days when the mtg-data.txt file is updated.

At this time the new Gatecrash cards will only match on name and type. The oracle script will add in the missing text field data.

[Hellfish]

EDIT: Sejiri Refuge works fine for me and Realmwright interacts correctly with both evolve and Soul Warden when I convert it to an ETBReplacement keyword. We should really be moving away from static triggers as they are not actually rules-compliant (most of them should be ETBReplacements and *maybe* a few are useful for scripting trickery)

Did you try it for the Human or the AI? I was reporting the issue with the AI playing it. It may have to do with some of Max's code reorg changes, since it seems like each one of them lately causes some issues somewhere else in the code. The same problem didn't occur when I played the Refuge.

http://imgur.com/rl9VvRP

And in case you are curious, as I stated below, the Log shows the Land was played first (since thats what the AI does) and then activated Level Up.

Okay, yeah, I see it now. Almost looks more like the AI keeps priority after the trigger goes on *or* has a chance to activate abilities before the trigger goes on, since Sejiri Refuge (among others) work fine if the AI has nothing else to do.

[friarsol]

Okay, yeah, I see it now. Almost looks more like the AI keeps priority after the trigger goes on *or* has a chance to activate abilities before the trigger goes on, since Sejiri Refuge (among others) work fine if the AI has nothing else to do.

Right, and I know this used to work, which is why I thought this part might have been due to Max's recent checkins, because I saw him in that area of the codebase.

I think originally the two back to back just felt like the same bug even though they were slightly different.

[friarsol]

r19227

Is Draft or Sealed working for anyone else? I thought it had to do with missing Oracle text, but since I just added those I'm still getting this NPE.

Quickest repro:

1. Goto Sealed Mode
2. New Sealed Mode Game
3. Choose Block/Set
4. Choose Gatecrash
5. Choose 6 packs
6. 1 Opponent (although I'm not sure it matters)
7. Any name

NPE | Open

java.lang.NullPointerException
at forge.gui.deckeditor.tables.EditorTableView.updateView(EditorTableView.java:371)
at forge.gui.deckeditor.tables.EditorTableView.setFilter(EditorTableView.java:316)
at forge.gui.deckeditor.controllers.CFilters.buildFilter(CFilters.java:179)
at forge.gui.deckeditor.CDeckEditorUI.setCurrentEditorController(CDeckEditorUI.java:87)
at forge.gui.home.sanctioned.CSubmenuSealed.setupSealed(CSubmenuSealed.java:228)
at forge.gui.home.sanctioned.CSubmenuSealed.access$2(CSubmenuSealed.java:128)
at forge.gui.home.sanctioned.CSubmenuSealed$2.mousePressed(CSubmenuSealed.java:66)

And a very similar one when opening up the deck editor after finishing a draft.

[FabioFLX]

Thanks for the replies, I understand Gatecrash's Oracle is still missing, so no problem at all
About the columns resize topic, I'm using a 1440 pixels monitor and I think it was better before. Is it too difficult to put an option to choose the behaviour in the settings?
Thanks again, Fabio

[RumbleBBU]

r19227

Is Draft or Sealed working for anyone else? I thought it had to do with missing Oracle text, but since I just added those I'm still getting this NPE.

Not just you, Sol, I'm getting the same problem. And I have an easier/quicker way to reproduce the bug - just try to open the deck editor for an existing (any existing) Sealed mode game.

It seems to be fairly recent, too, but other than that, haven't tracked it down to a specific revision yet.

[RumbleBBU]

Cast Lich made me lose game

That's not a bug!
(Well, yes, I know what you mean.)

[RumbleBBU]

Then, same in the deck editor, when you resize a column of one of the two tables, all other columns used to resize in percentage, but now only the closest column resizes; the previous way was far way better, I think.

Hrm, I had just changed this because I found it much easier to size the columns accurately when the other columns stayed put. Does anyone else have opinions on this behavior?

 
I found that I was able to adjust to the new way fairly quickly. My monitor is 1920 x 1200 and I narrowed most of the columns and used the extra pixels to widen the name and type columns.

I play Forge on laptops, widescreen laptops, docked laptops, and sometimes (very rarely) even on desktop computers. Consequently, I use a plethora of resolutions and switch between them, and have to resize the columns regularly and I'd say...I dunno. Both methods have their pros and cons. Adjusting 'accurately' is not something I would want to attempt (or even be able to do) with my laptop touchpads. Easy, fast approximations (the old method) would be preferable. But I can also see the benefits of the new method.

I think it might be a good idea to add a user preference to control the resizing behavior.

[myk]

I dunno. Both methods have their pros and cons. Adjusting 'accurately' is not something I would want to attempt (or even be able to do) with my laptop touchpads. Easy, fast approximations (the old method) would be preferable. But I can also see the benefits of the new method.

I think it might be a good idea to add a user preference to control the resizing behavior.

can do -- but which should be the default?

[RumbleBBU]

I dunno. Both methods have their pros and cons. Adjusting 'accurately' is not something I would want to attempt (or even be able to do) with my laptop touchpads. Easy, fast approximations (the old method) would be preferable. But I can also see the benefits of the new method.

I think it might be a good idea to add a user preference to control the resizing behavior.

can do -- but which should be the default?

Personal opinion - the new behavior should be the default. Then, if someone grossly dislikes it, they can be directed to enable the old behavior by setting the preference.
If the old behavior were the default, virtually nobody would discover the new one. And its benefits would be lost on the majority of players.

[FabioFLX]

I dunno. Both methods have their pros and cons. Adjusting 'accurately' is not something I would want to attempt (or even be able to do) with my laptop touchpads. Easy, fast approximations (the old method) would be preferable. But I can also see the benefits of the new method.

I think it might be a good idea to add a user preference to control the resizing behavior.

can do -- but which should be the default?

Personal opinion - the new behavior should be the default. Then, if someone grossly dislikes it, they can be directed to enable the old behavior by setting the preference.
If the old behavior were the default, virtually nobody would discover the new one. And its benefits would be lost on the majority of players.

Well, I think this is the right choice too.

[myk]

done (r19242). The preference is now settable in the deck editor preferences pane. I also took the opportunity to set the default column widths, so people who have never resized the columns before should get something a little more sane now. It seems very strange that active user preferences are written to files that are checked into svn. Doesn't this cause spurious change lines in the svn status/diff output? Are there any plans to separate user pref files from default pref files?

[Chris H.]

When adjusting the column widths the mouse pointer used to change when it was hovered over the titles. This feature has apparently dissapeared.

[myk]

When adjusting the column widths the mouse pointer used to change when it was hovered over the titles. This feature has apparently dissapeared.

did this happen recently? when did you notice it? What JVM are you running (I see a resolved bug that mentions this for the JDK)?