S3PE Error Message

[Simmerqueen]

When I attempt to use s3pe I get an error message at the point where I am going to save the file it has created. This is the message:

[Inge Jones]

Can I have a bit of context please?

1)What sort of content was in that file? Was it one of the game files, or some custom content you made?
2)What version of s3pe (check Help|About)
3)What types of task had you been carrying out on the file?

[Simmerqueen]

The content was custom content that I had downloaded - all .package files and mostly clothing/makeup. There were 147 files in the folder and I was merging them into one package. The merge completed successfully and the files were displayed, and then when I went to save the file so I could continue with my next folder, I got this message.

The version is 0912-13-1729 (just downloaded this morning).

[Inge Jones]

Then I suppose it's possible to consider whether that message was actually correct and you had indeed run out of memory. CAS parts are quite huge. What are your specs?

[Simmerqueen]

No, I don't think that's very likely to be the problem. My computer has 4 gigs of RAM and a 3.33 Core Duo processor. My husband built it and it's a dedicated gaming machine with nothing else but sims on it.

[Inge Jones]

Ok can you try doing this same process to just combine 10 downloads into a file and see if it lets you save ok. Let me know please. If it manages 10, try 50 and so on.

[Simmerqueen]

Okay, this is strange. I started with 10 files and worked my way up - it did fine until I got to 125 files, and at that point I got the error message again but - and this is where it got strange - it still created the compressed package file even when it threw the error. I'm assuming it might have done before as well, but since it errored out I didn't think to look for the completed package and probably failed to notice it.

So I guess the question is, in your opinion is it likely that the compressed package contains all the proper info even after an error like that? And I guess the other question still in my mind is why are others that I know of able to compress 400-500 files at a time and definitely don't have any more computing power than I have, yet mine freaks out at 125. Seems odd.

[Inge Jones]

I was just trying to make sure it wasn't a spurious error message caused by a fundamental bug in multiplexing packages, which I am relieved you have shown is not.

Now it seems to be a case that the error message is at least relevant - but why? You could completely disprove running out of physical memory by turning on your task manager just before the action that brings up the error (ie just before trying to save the larger file) and watch the resource usage as you are saving, until the error comes up. If that shows the graph never reaching anywhere like capacity, then we have a situation where for some reason the tool is not using all the memory available to it. That could be anything from having a limit to its own data structure size, to just needing to run in some sort of compatibility mode - maybe your OS is not allowing it as much memory as it needs.

[Simmerqueen]

I just now completely ruled out the memory issue (at least for the computer itself) by bringing up the task manager while it was saving, as you suggested. In fact, it never did reach anywhere near capacity. But again - it gets strange. This time, the save of the compressed file with 125 .package files in it didn't error out, which is good but why? Can't imagine what is going on.

ETA: I tried compressing a folder with 387 CAS .package files in it, just to see what would happen. It errored out while it was compressing it first, and I clicked continue. Then it errored out while it was saving it, and I clicked continue again and it did create the merged package file. Again, I confirmed no excessive memory usage.

I guess the concern for me is that I have no way of knowing whether some info was left out of the package due to the errors, and therefore no way of knowing how it would affect the game. Does this reasoning make sense?

[Inge Jones]

Yes I understand why you're bothered and it bothers us too. It will have to be investigated. Next time it happens can you drop down the details box and we can get a better idea of what it thinks it was doing. Is your system natively 64bit?

[pljones]

In the error box, there's a "Details" button. If you click it, it will display details of where the error occurred. You can copy these details to the clipboard. If you could post the details, it could help pin the problem down.

[Simmerqueen]

I just re-ran the 387 file folder and this is the detail from the error that it generated.

[pljones]

Thanks. That's definitely internal to .Net rather than s3pe.

In Windows Explorer, please go into the folder with all 387 packages, select them all and let me know the total size. (It should say in the status bar -- if not, right-click->Properties...) I doubt if the absolute size is the problem, though.

What may be happening is "memory fragmentation": just like on a hard disk, free memory gets allocated by the OS on a "first come, first served" basis. As programs request and free memory, free space gets fragmented. If s3pe suddenly says "Give me 1.2GB of RAM" and the largest contiguous space is 0.9GB (even if the total is over 1.2GB), Windows may be deciding not to provide the memory. I don't actually know if that's the problem.

You could try splitting the job up. Create 6 folders with 50 (or so) packages in. Then create a package from the content of each folder. Then merge those six packages, one by one maybe.

[Inge Jones]

what a palaver

[Simmerqueen]

The size of the 387 file folder is 1.66 GB. The 125 file folder that errored out yesterday (although not today) is only 454 mb however.

I'll definitely try splitting the job up if that seems to be the best solution, and then merging the results. I just was not wanting to end up still with a sizable number of folders (if I could only do less than 125 packages in each one) but the method you've suggested seems to be a possible work-around to that scenario.

ETA: I tried re-merging the merged files and I got the error, even when just combining them one at a time, so unfortunately that didn't work. Also, Inge I forgot to reply to one of your questions - my system is 32 bit.
ETA: A different error message, regarding an arithmetic operation:

[pljones]

If you're running a 32bit PC, then the maximum addressable RAM by a program under Windows is, by default, 2GB. Trying to create a single file of 1.66GB compressed resources will take a lot more RAM than 1.66GB.

I think you'll find you're running out of memory. (The process is active, so can't use your swap file.)

Some programs can use up to 3GB but I think you need to edit the .Net binaries to enable this. (There was a similar problem with SimPe.)

[Simmerqueen]

Well that would explain the 1.66 gb folder, but what about the 454 mb one? Also, why did the 454 mb one error out on some attempts and not on others? Due to those inconsistencies, it still doesn't really make sense to me. In addition, it appears that this program really only works properly for people with 64 bit systems, which is something that perhaps you should emphasize in the documentation.

[Inge Jones]

Simmerqueen you're being annoying and consumer-rights-ist now. I am not sure what guarantees you thought the tool was offering you?

We have both spent quite a bit of time trying to help you with your problem now, and I think you should thank us instead of complaining. NB I am using this tool perfectly well on a 7 year old 32bit so your observation is frankly inaccurate.

[pljones]

And it's being developed under WinXPSP3 32bit, too, on a machine that reports "992MB of RAM". In fact, I've no idea where "it appears that this program really only works properly for people with 64 bit systems" comes from, so I'm locking the thread as I can only assume it's an attempt to troll.

[Inge Jones]

Just noticed Simmerqueen had out of memory problems with SimPE too, and received advice there on how to configure her computer to use her extra memory. Maybe she needs to reapply this configuration - perhaps it got unset.