Why must RCE's be fatal?

Q · #1 July 29th, 2002, 06:20 PM

In the Last days I had some trouble with range check errors that always resulted in a game crash. And if I search this forum it seems that I am not the only one who had this problem.
I understand that in a rather complex game like SE IV such errors can happen. But my question is if there is no way MM could prevent the game from crashing when a RCE happens: if the game would give a detailed error message (where happened the error?) and then it would be possible to continue the turn with the lowest or largest value that is allowed.
As it is now you are just left with guessing what caused the error and you have to send the entire game to MM in the hope that he finds the problem. But frequently your game is lost.
Has this been discussed or is it just impossible? I am not a computer programmer but I am struck by the detailed error Messages you get from the game when something is wrong with the AI files compared to the total lack of information when a RCE occurs.
May be it's all the blame for Windows

?

[ July 29, 2002, 17:22: Message edited by: Q ]

Singollo · #2 July 29th, 2002, 07:00 PM

My first post on this board, ever. What an event.

Q, I think the issue revolves around the nature of the error message. The Range Check error is delivered by Windows--at least the ones that I've encountered are--and would probably not be under the influence of MM. Something like the AI errors or mod incompatibilities is SE-generated, so MM can be more precise with the error description.

I know this doesn't help much, but I thought I'd toss my two cents in. It is my experience, however, that Range Check errors often just...go away. I encounter them every now and then, and I just boot up a saved game from the turn before, and I'm usually fine.

Q · #3 July 29th, 2002, 09:27 PM

Welcome to this forum Singollo and thank you for your answer.
Yes I realized that the error message is from Windows (therefore my comment on Windows in the original post) and might be out of MM's control. But I don't know that and therefore this topic with a question mark.
Concerning the severity of these errors: in my experience, especially in simultaneous games, RCE's can ruin your day (and certainly your game)!

Atrocities · #4 July 29th, 2002, 09:34 PM

Welcome Singollo

I love your signature. Very funny.

Q When are you getting the errors? While in playing your turn, or processing the AI's turn?

I used to keep track of the RCE's I used to get.

http://www.angelfire.com/tv2/NewAgeS...rors/home.html

here is a link to an old discussion on RCE's

http://www.shrapnelgames.com/cgi-bin...=005317#000006

[ July 29, 2002, 20:47: Message edited by: Atrocities ]

Singollo · #5 July 29th, 2002, 10:34 PM

Have you encountered an instance in which the Range Check error was unavoidable...i.e., when you reloaded it automatically happened again, each and every time? I've never seen that, which is why it hasn't exactly been catastrophic for my games. I've seen some of your other Posts about the error, though (concerning vessels with large super-large tonnage, etc.), and it does seem as though it can potentially ruin a game.

Thanks for the welcome all. I've been lurking for a while now, and although I don't wish to usurp the tone of this thread, I would like to say that the quality and helpfulness of the advice dispensed on this forum is top-notch. I've gained and learned a lot just by logging in and looking around a little, and as far as I've seen, the flaming here is kept to a minimum (which is rare, in my experience). Kudos to everyone! When I work up the nerve, I'll post a question thread of my own...

Atrocities · #6 July 29th, 2002, 10:37 PM

Yes. The Fighter bug was a major source for this RCE re-occurring. The only way to stop it was to play the AI turn that was causing it and break up the AI fleet so that it would not launch all of its fighters at once. Time consuming, but it helped to circumvent the RCE.

Q · #7 July 30th, 2002, 07:24 AM

This recent RCE occurs in the combat simulation and in the turn processing of a simultaneous game. First I had it in the simultaneous game and it took quite a time till I found out that it was connected to the death star ship.
The error has now been reported to MM and I will see what he finds out.
But I believe this is certainly not the Last RCE and therefore some more information when it occurs and a fail-save mechanism that would allow to continue the game would be great if possible.

Mephisto · #8 July 30th, 2002, 11:23 AM

IIRC RCEs will pop up when for some reason the program tries to allocate to large a value to a variable (like 257 to a byte variable). There is no way make this no-lethal because if you could you already would have found the source of the problem and would have fixed it in the first place.

minipol · #9 July 30th, 2002, 12:40 PM

i'm a programmer and such errors can be a real pain. anyway, with the proper coding, they can be catched and dealt with. if you don't catch it, the errors is passed on to windows that then deals with the error for you since you didn't catch it. that's how the mechanism works. so if they would do checks whenever they worked with arrays, they wouldn't be getting this. but it's a pain to program it especially if you didn't do this from the beginning and it slows down the code a bit but makes it more robust

Jmenschenfresser · #10 August 1st, 2002, 06:09 PM

Speaking of which, it's obvious why over 255 as the number of systems produces an RCE, but has anyone ever tried to enter a really small number like 5 or less.

I tried that for testing mods...so that the empires would start next to each other. Sometimes I got an RCE and sometimes not.