1155 replies to this topic

[Sam.]

Okay, another crack at this.
This build fixes a bug in the effect index checker. Link.
NearInfinity.jar contains just the bugfix.
NearInfinityCI.jar also contains Bigg's case patch and a fix for the issue reported by phordicus

What kind of bug? If you don't mind, I'll wait till a few people have tried it, and if no one reports any problems, I'll link to them in the first post .

[drxw]

I found out why Near Infinity didn´t work correctly. The installation of BG2 is located in the Programs(x64) folder which is write protected. Installing the game to C:\ solved the problem.

Edited by drxw, 18 April 2012 - 08:45 PM.

[aVENGER]

If you don't mind, I'll wait till a few people have tried it, and if no one reports any problems, I'll link to them in the first post .

Works fine for me.

BTW, there's one small feature that crept in a couple of versions ago which kind of annoys me. Every time I try to compile a script with NI, it starts "Gathering creature and area names..." which seems to thoroughly thrash my hard disk, and sometimes lasts up to two or three minutes. What exactly does that do and can we make it optional please?

Edited by aVENGER, 18 April 2012 - 09:51 PM.

[Wisp]

What kind of bug?

It doesn't work. Specifically, attempting to use the function gives you an unhandled exception to the effect of "there is no attribute called 'Effects index'" (because the attribute is really called 'First effect index').

BTW, there's one small feature that crept in a couple of versions ago which kind of annoys me. Every time I try to compile a script with NI, it starts "Gathering creature and area names..." which seems to thoroughly thrash my hard disk, and sometimes lasts up to two or three minutes. What exactly does that do and can we make it optional please?

It looks like something that was added in v1.33. It warns you if you are attempting to compile a script that contains a non-existent script name.

Edited by Wisp, 19 April 2012 - 05:45 AM.

[aVENGER]

It looks like something that was added in v1.33. It warns you if you are attempting to compile a script that contains a non-existent script name.

Hmm, I gather the amount of script names it has to index rises drastically in a heavily modded game, hence the hdd thrashing.

Is there any way that this can be made toggleable via the Options menu? I'd rather keep track of my death variables manually and avoid the constant re-indexing.

Edited by aVENGER, 19 April 2012 - 06:31 AM.

[Wisp]

Is there any way that this can be made toggleable via the Options menu? I'd rather keep track of my death variables manually and avoid the constant re-indexing.

Yeah, I imagine the option of doing compilation without a safety net could be added.
However, I am having trouble reproducing your problem. Everything I try results in the script being compiled instantly without any appreciable hard drive activity. Could you describe your typical set of circumstances when you experience this?

[aVENGER]

I am having trouble reproducing your problem. Everything I try results in the script being compiled instantly without any appreciable hard drive activity. Could you describe your typical set of circumstances when you experience this?

It seems to occur most often when I first start NI (when it was closed before), open a script (i.e. AR1600.BCS), alter a thing or two and hit "Compile".

On subsequent "Compile" attempts (while NI remains open) the re-indexing doesn't seem to occur that frequently. However, if I close NI and repeat the procedure the re-indexing kicks in again. It doesn't always occur, but then it does, it's very annoying. Also, it seems to happen very often when I open a different game installation (i.e. switch from BG2 to Tutu) and open a script for the first time.

Edited by aVENGER, 19 April 2012 - 07:53 AM.

[Wisp]

Also, it seems to happen very often when I open a different game installation (i.e. switch from BG2 to Tutu) and open a script for the first time.

That did it. Okay, give me a while and I'll see if I can come up with something.

[aVENGER]

That did it. Okay, give me a while and I'll see if I can come up with something.

No rush, and thanks for looking into this.

[Wisp]

I have uploaded a version w1.0.2b (to break with this silliness of making changes to the source without altering the version number) to github. As before, the regular version is regular and the "ci" version has extras for working under e.g., Linux.

This version adds the option "Interactive script names" to the menu bar. It is enabled by default and disabling it will keep NI from caching script names, as well as disable all features that require cached script names.

It might be considered a good idea to treat this as a beta version for the time being. I have tested it, but maybe whoever wrote the initial code, or I, overlooked some feature that require cached script names (I don't think I have, fwiw).

Edited by Wisp, 20 April 2012 - 05:47 AM.

[aVENGER]

This version adds the option "Interactive script names" to the menu bar. It is enabled by default and disabling it will keep NI from caching script names, as well as disable all features that require cached script names.

Nice work!

I'll be using this version from now on, so you'll have at least one active tester.

[Wisp]

I have built a version w1.0.3, available here.

In lieu of a change log I give you this commit (bugfix) and this commit (feature).

FredSRichardson, if you are reading this, my master branch now contains everything I have done to date (there is the "ci" branch as well, but I don't think it is fit for inclusion in a mainline version of NI, due to the [necessary, afaik] clunkiness of the the case patch).

[Balquo]

Excellent work, Wisp!

[Sam.]

1^st post updated with links to newest versions of NearInfinity, including both Wisp's and FredSRichardson's branches as they contain different updates that have not yet been merged.

[Sam.]

FredSRichardson's branch updated. See 1^st post.

[Wisp]

As a heads up, I've built w1.0.4, which merges FredSRichardson's master branch. I also took the liberty of fixing a very small, harmless bug introduced in one of the recent commits (details).

Downloads

[Wisp]

New build. Version w1.1.0 this time. Usual download location.

Rough change log:

• The results from a search for unused strings can be saved.
  
• acmtool from libacm is now supported for decoding acm files, in addition to acm2wav.exe. acmtool is supported on all operating systems and NI expects to find it on the system path. acm2wav.exe is only supported on Windows and NI expects to find it in the usual places (in the game directory). You choose which decoder to use from the options menu (acm2wav.exe is the default).
  
• Assorted bugfixes.

And an additional item exclusive to the CI build (not relevant to the non-CI build):

• Various problems with having the tolowered game on a path that is not in all lower-case have been sorted out. But there are probably a few I haven't yet found.

A note about using NI to play the decoded audio files on operating systems other than Windows: it may not work. Specifically, stereo files produce no sound on my system. The files play fine elsewhere, but something in the Java innards of NI won't play nicely with the relevant parts of my OS. Mono sounds play in NI. I'll be looking into it.

Also, I've attached a 64-bit linux build of acmtool (built from Debian stable), in case it's of use to anyone.

Attached Files

•  acmtool.zip   27.32K   316 downloads

[Vlad]

Sam, there is an old problem with NI, which is really annoying. I've discussed it before several times, and I don't know know if you fixed it or aware of it. NI damages TIS files structure related to doors somehow when you extract them. There is my post on this issue with screenshots somewhere where I compare TIS file extracted with NI and with IEDSP. If I remember correctly, it has something to do with the messed file header. Anyway, it would be nice if you could take a look into this issue.

[Sam.]

Sam, there is an old problem with NI, which is really annoying. I've discussed it before several times, and I don't know know if you fixed it or aware of it. NI damages TIS files structure related to doors somehow when you extract them. There is my post on this issue with screenshots somewhere where I compare TIS file extracted with NI and with IEDSP. If I remember correctly, it has something to do with the messed file header. Anyway, it would be nice if you could take a look into this issue.

I'll add it to my growing list of things to do . I've looked all over SHS, G3, and PPG forums, and I haven't been able to find the topic/post you described (unless you mean this). If it was here at SHS a long time ago, the screenshots will be gone... Any chance you can dig it up for me?

[Sam.]

Very interesting... NearInfinity does remove the full 24 byte header from TIS files.

Tried to upload a screenshot but the damn uploader won't let me upload anything bigger than a small BG1 portrait... What the hell good is that?!?

Do you think you can take a look at this, Wisp?