Menu
▾
▴
#11 qgo2 sgfparser of nodes in which a delete occurs
Attached is an sgf file which illustrates this bug. Basically, with a node which adds a black stone followed by a node which removes that black stone, qgo2 treats those nodes as if they are one. I'm about to attach my patch for this change as well, but I do not see how to attach multiple attachments at a time through sourceforge.
Discussion
Anonymous
Before I add this, could you explain the bug and your patch to me a little better? I'm not really sure what the erase property is used for and I didn't write the sgfparser so I'd like to make sure this isn't intended somehow. It does seem like a bug but, would you mind?
I'm not sure how much more I can really expound, but I'll give it a shot. I was trying to view a problem file I created through qgo2, but quickly discovered that many problems were missing stones because of this bug. This is because, after creating one problem, I would erase all of the stones on the board to create the next problem. Not sure if this helps, but I also tested this in qgo and the behaviors match up only after my patch.
Well for instance, you can use AE to remove multiple stones and this should be treated as the same node. Does your patch handle this?
Also, there are definitely some bugs in the sgf editor. I think specifically, its not clear how to use the black and white stone buttons to add and delete stones. And it seems like one can't add multiple white stones.
Testing against multiple stones in AE now. Also, I'm trying to work through these sgf editor bugs, as my main use for qgo is to take notes from lectures and games.
Answer is yes, uploading my test file.
Shoot, forgot to log in on the last two comments.
Yeah, there's definitely something wrong with it as is, I'm not sure what though. I'll try to look into it soon.
I looked into this a bit. I don't know why you would use "AE". I could consider using it for making tesuji problems, except there's a bunch of other missing editor features for that. But there's no qgo button that does an AE, so not really sure what to do about that.
Like I said, there's definitely problems with the edit mode and the sgf parser, I was thinking of rewriting those down the line, but they're super low priority right now. If you need a very specific feature, I could look into that. (Like someone a few months ago sent me a patch for an "insert move" feature that lets you toggle an insert move mode instead of adding a new variation.) But besides that I'm going to put other work on this off until a later version.
Well, I was actually looking into rewriting the sgf parser myself, as it's much higher priority for me. If nobody else is working on a rewrite, but it was planned at some point, I'm probably going to dive in and start doing so.
Using AE has it's uses in taking notes during lectures, when the instructor clears the board or removes a stone to demonstrate a concept. In any case, I have found need on several occasions to use AE during note taking.
Also, the only reason there is no button for AE in qgo2 is because of a bug in the sgf editor:
- Start a new game
- Play a stone
- Click on the add stone edit tool
- Right click on the stone you played, to remove it
- Save to file
- Reopen file
- Observe that the remove was not saved
But note that the user may not want to save an AE for each misclick they make in editing an sgf file...
Regardless of if they want to save an AE for each misclick, what is saved to a file should accurately represent what is seen in the interface. If navigating from node A to node B causes stone x to disappear; then after saving, closing, and reopening the saved file, stone x should still disappear when navigating from node A to node B.
Yes, you're right. I also didn't write that stuff, I looked into it tonight and found several bugs so I'll fix that in a few days hopefully.
Okay, I fixed some things up maybe, let me know. Note that erasing the last stone won't save an AE because its the last stone, but otherwise right clicking should save an AE.