just a few noob questions i need answered.

i know how to get my weapon out but i'm clueless on how to hit anything. when i left click on an enemy it says "you see a rat/goblin" and right clicking seemingly does nothing.

another thing i have no idea about is how many cycles should dosbos be set to for arena?

please help guys i'm just an elderscrolls fan trying to play the basics.

At the top of this forum there's a pinned thread that says "READ FIRST...". That looks like a good place to start: it has the instructions for attacking in Arena, among other useful information.

Read that thread and the one on using DOSbox.

DOSBox is flaky in terms of speed. For example, 30000 cycles is fine on this machine, but I have to set the cycles much lower on my laptop to get the same performance. It seems as though DOSBox has timing issues across processors and the best way to get your game running properly is to adjust it while playing using the hotkeys until it runs properly. It's funny seeing Railroad Tycoon take up to 30k cycles on some systems and only 1k here, and that's an ancient game, so expect to have to fiddle with Arena/Daggerfall a little.

DOSBox is setup with a autocycle feature, this is the best if you don't want to use it. Also, the issue isn't with DOSBox, its with your processor speed. Depending on the how many resources the game needs and the general power of the computers processor you'll need different cycle settings if you do it manually.

For example; Daggerfall on a typical 3.0GHZ processor takes about 15-20,000 cycles, while Battlespire takes quite a bit more, in the high 30,000's if you wish to run it on the 'high' settings, and even then you'll have issues with performance.

The issue IS with DOSBox, not a processor. Think about it. SNES emulators, NES emulators, even Atari emulators run properly on all processors, both AMD and Intel. However in DOSBox 30k cycles may equal 200MHz on one system and only 20MHz on another. In other words, running SNES9X on my P2/233, P3/800, and this P4/3.20GHz systems all run at the proper speed, no tweaking required. DOSBox on the other hand, runs entirely different on different systems. 30k cycles on a P2/233 is not 30k on my P4. DOSBox does not properly emulate cycles across processors. Saying that the piece of hardware that runs the OS and every other game around flawlessly is the problem and not DOSBox is highly incorrect. I have worked on emulators in the past and would be happy to break it down in detail and possibly provide source-code in PM if you want more information.

Also, the "dynamic" operating mode of DOSBox does NOT mean that you don't have to adjust it. It means that if you have it set to emulate 30k cycles and a game like the original Doom only uses 15k, it will only emulate the 15k cycles needed. It does this by reassembling code on the fly which can lead to performance boosts, allowing you to emulate many more cycles than with the "static" mode of operation. Read th DOSBox documentation if you need more info on that subject.

Clydeshademore, if you post your system specs (RAM, CPU, soundcard, and videocard) I can probably recommend a good setup to begin playing with. It may not be perfect, but it'll be within 10k cycles of being correct. I've been using DOSBox for years and on systems above the 2GHz barrier it becomes second-nature.

DOSBox and SNES9xw are completely different things. If we're talking about an SNES emulator, well... an SNES is a console, and a huge advantage consoles have, from a performance standpoint, is that every console of a particular make has identical hardware (Well, idealy. I guess we have XBox 360, XBox 360 Lite, and XBox 360 with Almonds now). Any particular SNES game is going to work the same on any given SNES.

DOSBox, on the other hand, is a processor architecture/OS emulator. And it's emulating an OS that was in use for something like 16 years, which ran on a really, really huge variety of hardware. In the case of Arena, Arena wouldn't run the same on two computers running DOS natively if they had vastly different hardware.

In some regards, adjustable cycles is sort of a built-in MoSLO. Older games were not meant to run on fast processors, and if you give them too many cycles, they're too fast. Likewise, newer games work fine on newer processors and need more cycles. It's not like with an SNES emulator, where every game has the same amount of memory and the exact same processor available to it.

If your computer is slowing down from trying to allocate too many cycles, some of that's from the emulator, as is the nature of emulators, but at that point, your computer's processor is being the bottleneck.

Also, how does something reassemble code on the fly? That doesn't even make sense.

This is straight out of the DOSBox readme.

CPU Cores  On x86 architectures you can try to force the usage of a dynamically  recompiling core (set core=dynamic in the DOSBox configuration file).  This usually gives better results if the auto detection (core=auto) fails.  It is best accompanied by cycles=max. Note that there might be games  that work worse with the dynamic core, or do not work at all!

This was added in 0.6x and offers a HUGE performance boost to most games. It works VERY well in 0.72 and it is how I play DF on this XP system if I don't feel like booting my P2/233. Also, the architecture being emulated does NOT matter. To emulate an SNES system, you emulate a specific CPU/speed, specific sound processing unit (SPU), and specific video processing unit (VPU). In DOSBox you're simply emulating an x86 CPU, an SB16 for the most part (some people emulate others), and whatever video chipset DOSBox emulates. Also, my computer has never slowed down, least of all when emulating DOS via DOSBox. If I max out DOSBox the most CPU usage I see is around 70% on one CPU and 20% on the other. If your system physically slows down when using DOSBox then you're either using an AMD which is known for slowing down a bit under a full load, or you have a system problem.

It is very apparent that you are unfamiliar with emulation in general. Most of this information is in the readme or changelog for DOSBox, which indicates that you have yet to read it. If you wish to continue discussing emulation, take it to PM. This is not the place for this.

Go into your "dosbox.conf" file and set "cycles=max"

This will make DOSBox guess the correct amount of CPU cycles and if you have a relatively recent PC it should run better. There's not much point in manually setting CPU cycles since this feature was introduced (and it's a relatively recent addition).



The problem with your argument is that PCs are not identical, and emulating entire generations worth of hardware is a significant task. To pass it off as being the same as emulating a SNES (which is a 3.58 mhz processor) is pretty short-sighted. PCs have varying hardware between machines, and the DOSBox team has to emulate a LOT of redundant hardware that might be required for a specific game.

A Super Nintendo is a Super Nintendo is a Super Nintendo.

So, with your logic, emulating an Xbox should require the same amount of processing power as a SNES, right?

Are you Xenithar, by any chance? I recall him using a similar name.

I wasn't saying that emulating a PC was the same as an SNES. However, the current emulation is flaky in that you're emulating cycles which don't match up to any specific hardware at all. If the program was emulating SX hardware (increments of 25MHz) or DX hardware (increments of 33MHz) then it would be much more reliable. In fact, the program would perform best if it simply emulated the various older chips and allowed the user to choose one. It would be less confusing (P2/233 equates to what in DOSBox on what hardware?) for new users, it would be much faster, and it would be more stable. This however, is not the place to discuss improvements to another program.

As for being short-sighted, you're not reading the post. What I am saying is that Daggerfall (for example) runs on this XP machine at 30k cycles as good as it did in DOS, but 30k cycles on my friend's old XP machine won't run it for crap with the exact same DOSBox configuration. It isn't that his slower processor is chocking, because it was sitting around 60% usage and never hit 100%. DOSBox does not emulate cycles the same across processors, that is the problem. 30k on a 3.20GHz chip is NOT 30k on a 1.8GHz chip. I have literally copied my configuration from one machine to the other and it is not the same, nor does it max the CPU out. That is a programming bug somewhere that somehow varies cycles based on actual processor speed or something. I'd have to browse their code to figure it out, and I am not going to do that. The point is, 1 DOSBox cycle on a 1GHz chip is NOT the same as 1 DOSBox cycle on a 3GHz chip, which is a bug.

Oh and I am not Xenithar. I go by Sephiroth, and I went by Setzer in the middle 90's. Before that and back before the net was around, I went by Spinal on the DWANGO networks. I knew a Zenithar on a MUD a long time ago, but the only Xenithar I know of is DarkXen from Digital Dominance.

Edit: I was going to be gracious and assume that I just misread things, but apparently others agree so this has been removed.

That doesn't make any sense. 1 cycle is 1 cycle no matter what. However, 10k cycles use considerably less CPU power if you use the dynamic core, instead of the more accurate (but slower) normal core.

Also, the dynamic core is automatically used on DOS games that use "protected mode" while it's off on the older titles.


If you want DOSBox to automatically adjust it, use cycles=auto.


Or it's possible that you're just flat out wrong with several of your statements.


I've played over 50 games in DOSBox, and can safely say 10,000 cycles is about a 50mhz 486. It varies slightly from game to game, but in general it holds true.


This is all misinformation. 1 DOSBox cycle is 1 DOSBox cycles no matter the hardware. The cycles are the amount of CPU instructions DOSBox attempts to emulate each millisecond. If it's set to high so the computer can't handle it, it will obviously stutter.

When you tried it on your friends computer you must have done some sort of error. If it never hits 100% I would guess he got a dual-core (or hyperthreading) CPU so DOSBox can only use slightly more than 50%, since it's a heavily single threaded program. If you don't belive me, feel free to post about it on the http://vogons.zetafleet.com/viewforum.php?f=31, and you will have plenty of people saying the same thing.


If it isn't, then it's his identical twin.

If it isn't, then it's his identical twin.

It couldn't be. If it were, I would be white-hot with rage and annoyance after reading his post.

Okay, I think we're all in agreeance, then. MOVING FORWARD.

Since this is your first time playing arena, the wall-destryoing and floor-destroying spells are *great* fun. Also, Arena's even more roguelikeish than Daggerfall in several areas. Also, the limitted classes aren't really a bad thing. Arena's a shorter game with a definite focus on the main quest, and trying different character types is a lot more meaningful when there's a bigger difference between them.

Anyway, I have a few complaints about Arena:

1) Limitted shop variety. This is also something that bugged me in Oblivion. You only have mage guilds, weapon/armor shops, and taverns in Arena, AFAIK.
2) No need to eat/pay taxes/whatever. As I said, it's sort of roguelikish, but there's no persistent "drain" element going on, like need to eat. Given that the TES games are SORT OF life sims, this bugs me about the entire series.
3) Punching crashes the game a lot

Again, your lack of knowledge on the subject astounds me. Settings cycles to auto results in an average of 3000 cycles being set, not the 30k needed to run the game. This is due to the intro video only needing a few cycles. After that I have yet to see DOSBox figure out it needs 30k cycles or so and set it up. This generally results in taking eons just to exit the game. I can physically tape this with my camera if you're going to choose to continue living in la-la land and posting mis-information.

As for cycles, 10k is far from a 50MHz 486. I say this because I own a 486DX (33MHz) system and to make DOSBox run Doom II the same as it does on that box requires 15k cycles. However, on the laptop it only requires about 5k to get that level of performance, and the laptop is a slower system! This is where figuring out configurations for your unique system come into play. I even wrote a front-end for DOSBox which comes stock with settings for games based on this system, but those settings won't run the games on my friend's P4/1.8GHz system. That's the entire problem. If a game runs at 30k here, it may run at 20k or 40k there, depending on the game. His processor isn't maxed out so explain that with your obviously superior wisdom. Same game configuration, same game, same everything, except one processor is 1.8GHz and one is 3.20GHz. If you can explain that in reasonable terms, I'll shutup. The thing is, you can't because the cycles in DOSBox are tied to the processor in your system, and with varying processors you get varying results. Again, I can tape this and post it as well, if you choose to continue posting incorrect information.

I am not saying that your results are not right for your specific system, but they will NOT work on a global scale. If you don't believe me, take your configuration file and plop it onto a system with about a 1GHz slower or faster chip in it and try the game. You will notice a LARGE difference in the performance, and that is the point that I am trying to get across. Assuming you're on a 2GHz system, copy your DOSBox directory and games to a flash-drive or something and plop them onto a 3GHz system and play using that same configuration. You will notice the difference.

I may get the camcorder out and tape Daggerfall on this system, then copy it to the laptop and tape it. I'll tape the configuration file open in Wordpad so you can see that no changes were made, and then I want you to explain how the game runs differently on two systems with the same exact configuration. All you have managed to do is attack me and continually scream "you're wrong I'm right", but you have offered no evidence to back this up. I happen to own about seventeen computers at this point and have tested and verified everything that I have said. Now you need to backup your claims with proof as I am about to do with this video, or you need to admit that you're only going on experience from your machine and that you haven't tested DOSBox over a wide range of computers yet.

I can also care less if you consider me to be somebody else. You're entitled to your opinions.

Continue to flame people and you're banned.

So wait, attacking me is OK, but making a simple statement (that is NOT flaming) is grounds for banning? I do expect a full explanation as to why that is so, because there are statements just like that made against me. Here let me quote a few.



The first and second pretty much call me an idiot and a liar, and the third calls me a liar as well. So what's the difference here? Also, if you read my post, you'll see two or three times where I tried to get this discussion into PM to keep it off the board, and others kept it going. So I ask again, what exactly makes me out to be flaming here that hasn't been done by others as well?

Those quotes just tell you that you're wrong (which you are). Not that you're stupid or have a "lack of knowledge".

They don't attack you as a person, but the information you post. Quite a difference.

Correct, which is why I didn't bother reporting them or anything. Mine states opinions as well, nowhere did I flame anybody, call out names, or do anything against the guidelines. It is your opinion that I am wrong. It is my opinion that the other user is unfamiliar with the aforementioned program. My second quoted statement was a fact based on my opinion. I was astounded by the lack of knowledge I saw. That is hardly flaming. Had I called the person an idiot or something, then things would be different.

As for what we're attacking, I am simply responding to what the other user posted as well. How is that ANY different? I don't know the guy so I can't possibly attack him. I am only going on what that user has posted, the same as he is.

I understand that you agree with the other person, but that doesn't make my opinions flames either. Nor does it change the fact that I have tried multiple times to take this to PM and keep it off the board, so it does offend me when I am warned for posting my own opinions, which others do all the time, after I tried to get this situation off of the public board. And For the third or fourth time I will ask those users to PM me if they want to continue this discussion. I am not going to start cursing or calling names just because it's in PM and not on a forum, but this is not the place for this discussion.

You can try around 5k but 7k runs it a little better.

I wanted to post the videos I've uploaded to Google Video now, so that you may see why I have stated what I did. The first is from this PC and the system specs are listed in the video details. The second video is from the laptop, and I simply copied DOSBox and Daggerfall from this PC directly to the laptop to run it. The laptop has decent specs also, and actually has a newer, 64bit processor with dual-core capability. As you can see, the newer, faster CPU actually runs worse than this older 32bit P4. The funny thing is that it barely runs at all on my friend's P4/1.8GHz system! So by logic, P4/1.8GHz < AMD 64bit dual-core < P4/3.20GHz. It makes no sense, but this was the reasoning behind my statements above. If anybody can figure it out, please PM me with the solution.

http://video.google.com/videoplay?docid=2171213273052929142&hl=en
http://video.google.com/videoplay?docid=-738138188289075286&hl=en

Oh, and the configuration was dynamic core and 30k cycles, in case you're wondering.

"Settings cycles to auto results in an average of 3000 cycles being set, not the 30k needed to run the game. This is due to the intro video only needing a few cycles. After that I have yet to see DOSBox figure out it needs 30k cycles or so and set it up."

Auto is rather broken in DOSBox. This is a known problem.

Also, it's worth mentioning that there are very, very few games that should be run at more than like, 20K in DOSBox. Running Daggerfall at 30K chokes my 3.2GHz P4 just the same as it chokes the... whatever 2.2GHz processor my laptop has. And both run fine at 20K.

"except one processor is 1.8GHz and one is 3.20GHz. If you can explain that in reasonable terms, I'll shutup."

You're still not really describing your situation clearly. Also, the videos aren't present. So, questions.

1) How fast is that 64Bit AMDyou mentioned? You said it was slower than the 3.2GHz one before, didn't you?
2) How well does DOSBox work on 64Bit architecture, anyway?
3) How many cycles were you attempting to run Daggerfall with? If it was 30k, then odds are, you were choking the 1.8GHz processor and not the 3.2GHz processor. As I said, though, 30K is generally too high for any game unless you're using the dynamic core in which case you really still shouldn't go out of the 15k~ range for most higher-end games. It's also worth noting that Daggerfall isn't emulated all that well anyway.

"If you don't believe me, take your configuration file and plop it onto a system with about a 1GHz slower or faster chip in it and try the game."

I actually have done this, and they run the same. I mean, literally, when I got my laptop, I just copied over my entire D-Fend folder to it. Old configs and all. In the case of Daggerfall, the program was running at 20K cycles with the dynamic core option. Desktop CPU is, again, 3.20 GHz, and the laptop is 2.2GHz. Desktup is a P4 and the laptop is some kind of Intel (Centrino?).

http://www.pcstats.com/articleview.cfm?articleid=1665&page=6

64 bit programs are not always faster, and in fact are sometimes slower than 32 bit programs. Also, having a dual-core processor wouldn't make a difference in DOSBox's performance unless the user was doing something like encoding a video at the same time. DOSBox isn't designed to take advantage of multiple cores. Last I checked, when running DOSBox on my AMD X2 I was using 50% of my CPU (i.e. one of the cores).

I posted earlier in the thread to set "cycles" to max. Auto works horribly, but max runs far better for me. Arena was unplayable at default settings, and is now quite playable after changing it (although running it at max detail still leaves much to be desired).

I would've taken this to PM as well, but your information shouldn't be posted as fact and left to sit here without being corrected (or at least debated).

Yeah I know DOSBox won't take advantage of the 64bit architecture or the secondary core, I was simply stating what the systems were for clarity. The videos should be up soon, I think they have to be verified by Google staff, in case I was trying to upload porm or something. It usually takes about 24hrs.

I also find it amazing that you run Daggerfall with 20k cycles. I have to run it at 30k~35k for max performance, and 20k is crap on this system. However, when I lower the cycles to 20k on the AMD laptop, it runs almost as good as it does here at 30k. Hell, I run Hexen at 25k and Heretic at 20k. I run Shadow Warrior and Duke Nukem 3D at 60k and Blood at 70k. The funny thing is that Hexen runs great at 25k on this system and the laptop, but my buddy needs his at 20k to attain the same speed. I'm not going to go around posting videos all day, but it's this kind of stuff that brought me to my original posting here. Different systems seem to emulate cycles differently, and without digging through the sources, I haven't a clue as to why.

As for going out of the 15k range, Duke3D, Shadow Warrior, and Blood are unplayable at 15k, as is Strife. Strife I think runs around 30k also. I have never had problems running those games at up to 70k, and as such 70k uses about 90% of the CPU, so I imagine my max would be around 80~90k cycles if I had to go that high. I also note that the 1.8GHz system can go up to 45k cycles before it maxes out, so Daggerfall shouldn't be an issue. That system maxes at like 51k or 52k, so 30k should be nothing.

As for specs, they're posted on the video, but until then, here you go.

P4/3.20GHz HT 800MHz FSB (32bit, only HT, no dual-core)2GB DDR400 (PC3200) Dual-ChannelnVidia 7800GS 256MB AGP8XCreative Audigy II w/4.1 Speakers120GB Native SATA HDDP4/1.80GHz 533MHz FSB (32bit, only HT, no dual-core)512MB RDRAM (I forget the speed)nVidia 5700LE 256MB AGP8XCreative SBLive! w/2.1 Speakers40GB ATA133 HDDAMD Turion64x2 1.60GHz (64bit, dual-core)1GB DDR333 (PC2700) Single-ChannelATI Mobility Radeon 9800 256MBBuilt-in Soundcard120GB ATA133 HDD

I am running three OSes on the laptop right now, Debian/GNU Linux, XP Pro x64 Edition, and XP Pro 32bit. Plain old XP Pro will soon be removed in favor of just having x64, since my 32bit games perform the same under both OSes, and I only see gains in 64bit games. The 64bit OS however, does perform better on that CPU than the 32bit one. Debian is also 64bit, but I have not run DOSBox under Linux yet, as that OS is designated for work.

*EDIT*

I just ran Daggerfall with 20k cycles and while it still ran, it would have given me a headache in a short period of time due to screen-tearing and such. It appeared to be running around 25fps, whereas at 30k it hits 60fps, like it does in real DOS on my old P2. I didn't even try 15k. I have noticed that going above 35k doesn't hurt game performance at all, and it runs as smooth as ever, but the music slows WAY down above 35k while in menus and such. Apparently DOSBox only applies cycles to the emulated CPU, and not the SPU (SB16 or whatever is being emulated).

Mine states opinions as well, nowhere did I flame anybody, call out names, or do anything against the guidelines.

The difference really is that telling someone they're wrong, while it can cause hurt feelings, can on the other hand help advance the discussion by setting the stage for where you're going with the rest of your post. But saying that you're astounded by someone's ignorance doesn't advance the discussion and can only hurt feelings.

It's okay if you state what the systems are for clarity, but saying that the Pentium 4 runs better than the Turion when you're using a 64-bit OS that could have a performance hit isn't right. Windows XP probably does run better as 64-bit, but that doesn't mean DOSBox will.

I'm bowing out of this discussion, though. I wasn't planning on posting here again for awhile and I felt like I needed to chime in.