Feedback talk:Joe Kimmes/Archive Oct-Dec 2010

Costume Brawl 2010

Hi Joe. Is it a bug that this year's costume brawl builds are exactly the same as last year's? --Musha 19:48, 21 October 2010 (UTC)

Gw.exe Excessive Disk Writes

I have a fully up to date version of Guild Wars (including -image), yet it seems to do a lot of disk writes. Whenever I start Gw.exe it will immediately write 15,596,728 bytes to the hard drive. As soon as I load into the game this number goes up just over 100 MiB. If I logout and load into the same town again, it stays the same. Every time I load into a different map, it increases by about 5 MiB.

To compare, WoW writes less than 100 KiB if I launch the game, enter the world, and exit again.

I would appreciate if you can look into this, as I am moving Gw off of my SSD for now.--129.13.72.197 03:56, 4 November 2010 (UTC)

I tried hooking the WriteFile function and simply returning TRUE right away, and the writes disappeared with no apparent problems in Guild Wars. Of course I did not try saving templates or local settings but at least it indicates that these writes really are pointless/unneeded.--129.13.72.197 18:06, 4 November 2010 (UTC)

Are you sure this wasn't due to game updates? Is this happening everytime?--Lania 18:43, 04 November 2010 (UTC)

As I said my game is fully updated (-image'd) and this is happening every time.--129.13.72.197 19:43, 4 November 2010 (UTC)

Oh and I can reproduce this on my laptop as well.--129.13.72.197 19:44, 4 November 2010 (UTC)

Can you tell which files the program is trying to write to? I can't test it myself since I'm at work, but I'll look into it too when I get home. I can only speculate what's going on at the moment. --Lania 20:16, 04 November 2010 (UTC)

The whole thing is pretty much being written to the Gw.dat file.--129.13.72.197 21:16, 4 November 2010 (UTC)

And is it removed when the program is closed? — Raine Valen 21:19, 4 Nov 2010 (UTC)

It's not so much about the Gw.dat file growing - in fact it isn't, it merely seems to read some files into memory and dump them right back onto the disk, likely overwriting it with the same data.

Try showing the I/O bytes read and write columns in Task Manager and look at the values changing while you load into the game.--129.13.72.197 21:42, 4 November 2010 (UTC)

Here is what I think might be happening... GW uncompresses data from compressed portions of GW.dat as the game loads, and dumps them to the physical memory. As the size of the GW physical memory usage grows or the ratio is hitting it's limit (as in computers with low amounts of RAM), GW dumps the data back into GW.dat into uncompressed portions on GW.dat. (that's only if GW.dat increases in size when the program is still open) If that's not the case then, yeah GW is overwriting parts of GW.dat fairly often then. *still not home so still speculating* --Lania 23:24, 04 November 2010 (UTC)

Nope, that cannot be it. I have 12GB of RAM on this PC. But it's true that it seems to be extracting the file from the Gw.dat and then for some unknown reason writing the same file back.--129.13.72.197 23:53, 4 November 2010 (UTC)

Yeah you're right, it's writing to GW.dat fairly often in 4,2,1.8,1.2MB, and 3,4,8kb, segments all at the end of the file. There are some writing activity at exactly 4096 byte chunks at the beginning too. Seems related to some kind of paging operation since GW self limits it's own physical memory usage near 500-600MB. --Lania 01:47, 05 November 2010 (UTC)

Oh also, the GW.dat file increases in size about 4MB when the game is running, and shrinks by the same amount when the game is closed. So it seems like a section of the file that is getting updated frequently. --Lania 01:50, 05 November 2010 (UTC)

I'm inclined to believe that it is other characters' armor sets. — Raine Valen 1:53, 4 Dec 2010 (UTC)

That seems unlikely, if the low-resolution armor textures have to be recreated each time they're loaded, then the program would need to load all the high-resolution sets to remake them, which ruins the entire point of not needing to load them. On top of that, disk i/o is probably the biggest bottleneck in desktop computing aside from that whole von Neumann thing. There's no purpose in writing to the DAT file unless something is being saved, and if you're used the -image flag, everything should be saved. –Jette 02:22, 4 December 2010 (UTC)

It actually makes sense to me. Considering that the armor texture is different for every possible combination of armors, you're looking and thousands of different combinations, multiplied by the number of hairstyles (and their appropriate alpha channels), hair colors and dye combos. While it makes sense to keep a raw copy of the different skintones and body/face combos in the file, keeping track of the number of armor textures would severely bloat the dat file.-- Pyron Sy 03:14, 4 December 2010 (UTC)

Texmod says that textures are generated on every load. — Raine Valen 3:42, 4 Dec 2010 (UTC)

Which is half-correct. Most armor files are kept split into five parts (I'm guessing here, it's late and I can't be bothered digging my old files; the numbers are close enough to accurate): three are 256*256 squares, the other two are 256*128 rectangles that fill the last square to make a single 512*512 image, which is "stretched" across the 3D wireframe. But all that's being done is movement; the textures aren't being changed, they're only being assembled. Moreover, those images aren't saved afterwards and must remain in high-speed memory (i.e. RAM) until they're no longer needed, when they should be discarded. Writing to disc should take no part in this. Joe would know better since he has access to the source, but as far as I can tell, unless A) a crash dump or other log is being written, B) new data is encountered and is being added to the archive, or C) the user requests a data write such as a screenshot, there's no need for GW to write any information to the disk during an ordinary execution. I don't design games or anything here, but I can't see any reason for what's happening. –Jette 04:53, 4 December 2010 (UTC)

Got a couple of suggestions i`d like some Dev feedback on if possible Joe, http://wiki.guildwars.com/wiki/Feedback:User/Balky/How_to_remove_Bots_Gamewide http://wiki.guildwars.com/wiki/Feedback:User/Balky/Add_PvP_Arena_Moderators

Many thanks

Balky 13:20, 7 December 2010 (UTC)

the first day of spring

Hey Joe, let's play a game. Tomorrow Mike walks into your office and says he's letting players run their own scrawny little GW servers from home. He says the system will work as follows: first a disgruntled user will run the GW client and pick an alternate server from a drop-down list or enter an IP or server name or whatever. Then they enter their credentials for their account (the same one used on the official servers), which gets zipped over to Bellevue or wherever it is you guys keep your database. First, the credentials are checked. If that works, the account's status is checked: is it banned, which campaigns and stupid microtransaction add-ons does it have, etc. Then that information (but not the e-mail address, password or other personally identifiable information) is transmitted over to the unofficial server, which is hosted by some fat nerd with a creaky old Dell from the 90s and a copy of FreeBSD out of his mother's basement. After that, the player is allowed to do anything he/she could do on a normal GW server and no more: locked professions and campaigns remained locked, skills remain locked, etc. Some stuff could be changed by the people running the unofficial server, like skills, but an ordinary player could never use monster skills, skills from campaigns they don't own, etc. I'd assume local administrators could do whatever they want like on the test servers (which would be weird for admins who don't own all GW campaigns, since you'd have a bar with Spectral Agony, Banish Enchantments and Glyph of Doom on it but RC would be locked).

In this hypothetical scenario, would you be able to come up with all this junk by yourself in a reasonable period of time or would they have to yank programmers from GW2? –Jette 00:57, 29 November 2010 (UTC)

And Joe, when you've picked yourself back up off the floor, can you arm yourself with a digital camera and show the above post to all programmers who have contributed to the engine, net code, world map interface, account permissions, and any other relevant part of the game. The photo essay on human expression will be truly awesome! Thanks. :) -- WarBlade 05:11, 29 November 2010 (UTC)

Yeah, that would take GW2 programmers for sure; in fact, there's very little of what you describe that would be doable by me at all.

Ignoring whether it's possible or not, the whole validated-login-auth thing is meaningless once we put the server code in the player's hands. There's nothing we could do to prevent a server admin from changing the whole game, so it would be pointless to connect it to the 'real' Guild Wars. - Joe Kimmes 19:31, 14 December 2010 (UTC)

Aw. :U Well, I guess I can keep sending my weekly "pls gief sauce when GW2 comes out" email to Mike O'Brien... thanks, though. Someone on John Stumme's page wanted to know something code-y, if you want to take a look, although I'm still not 100% what it is they're asking or why. –Jette 22:51, 14 December 2010 (UTC)

Buff retracing

→ moved from Feedback talk:John Stumme#Script_question:_.27Bonded.27_skills_and_maintained_enchantments

While we are at it, Joe, why do Arcane Conundrum and Renewing Surge don't end prematurely on their casters' deaths? 88.153.105.75 00:21, 15 December 2010 (UTC)

Why not put it here? It may come as a surprise, but he'll probably be more likely to answer questions addressed to him and on his own page. --ஸ Kyoshi 05:27, 15 December 2010 (UTC)

Those skills have a negative effect on the target apart from any effect on the caster, why would they expire with the caster? - Joe Kimmes 18:03, 15 December 2010 (UTC)

I believe the implication is that since Parasitic Bond expires if the caster dies even though it has a -1 degeneration effect, Arcane Conundrum and Renewing surge should too. I think the answer is more of a design one than a mechanical one; with PB the intention is to use it as a cover hex and/or some sort of awkward heal, the 1 pip is just a side-effect. Arcane Conundrum and Renewing Surge are more offensive, with the energy gain being a bonus rather than the primary aim of the skill. –Jette 18:18, 15 December 2010 (UTC)

Shame, Guilt. Hmm, why do these end then? 88.153.105.75 19:02, 15 December 2010 (UTC)

They do? I'd never noticed. –Jette 20:21, 15 December 2010 (UTC)

These two are different from PB/AC/RS hexes above, because they don't do anything at all until they end, and their end effect does something to the target, rather than just the caster, thus they can't be grouped together. Rose Of Kali 22:45, 15 December 2010 (UTC)

Oh, and the reason PB expires with the caster's death is in the name itself - it implies a living parasitic link between the caster and target, as if the caster is leeching that pip of health off the target like a parasite, so it would make no sense if it didn't end with caster's death. At least I think that's it. :P Rose Of Kali 22:56, 15 December 2010 (UTC)

At the risk of pointing out the obvious, it could be because both Arcane and Renewing were updated to add the on-end effect and previously did not have it. --129.161.205.103 23:07, 15 December 2010 (UTC)

There are (figuratively) countless examples of skills that affect the caster, even barely, that end on caster death or over reach (becoming outside of party range from the target). There's no other way to categorize them, you're not going to be able to soundly justify every single one. The reason that Renewing Surge and Arcane Conundrum do not end under the normal rule is the same reason that Ray of Judgment did not scatter NPCs. –~=Ϛρѧякγ (τѧιк) ←♥– 23:39, 15 December 2010 (UTC)

Looking at the skills mentioned, yes, this is basically a Ray of Judgment scenario (skill's functionality changed and a piece was missed). There's no intentional design/lore reason behind it, as far as I know. - Joe Kimmes 06:31, 16 December 2010 (UTC)

enjoy coca-cola

Is the Miniature Polar Bear supposed to have a preposterously low drop rate? We've gathered a sample of over 3,000 opens of the Wintersday Chest and not once has a Polar Bear appeared. I understand if the designers don't want you letting players in on secrets about how drops work, but this is a bit absurd. I've never opened the chest myself, and certainly have no desire for the mini (I'd consider it if carried a coke bottle), but the infrequency of the drop rate — even compared to other "so rare it's stupid" items like Ghostly Heroes and those creepy beating-heart staves — seems to imply that there could be something wrong. Did the staff come in and tell you "hey, make this item virtually unattainable" or is there another factor at work? A misplaced decimal, a weird requirement such as quest, anything. I don't want to report it as a bug or something until I know whether or not someone intended it to be nightmarishly uncommon, and you're the only one I know who fully understands the drop tables, so I thought I'd ask. –Jette 01:41, 19 December 2010 (UTC)

I am curious about the answer, too (although I am fine if there is a ridiculously low drop rate for something so "nightmarishly" easy to farm ;-)  — Tennessee Ernie Ford (TEF) 01:55, 19 December 2010 (UTC)

If its so easy to farm then how come YOU don't have one? :P Or do you..? Anyway, I, too, would like very much a confirmation of this mini's droprate. --Musha 06:33, 19 December 2010 (UTC)

Some interesting statistics have been reported here. Could you please take a look Joe and verify if this is accurate and if it was intended? --Musha 06:56, 19 December 2010 (UTC)

I think TEF means the quest is nightmarishly easy to farm, so the best way to make the rewards rare is to make them drop incredibly little. --ஸ Kyoshi 00:29, 20 December 2010 (UTC)

I'm pretty sure he's talking about this. But I'm also quite curious about the answer to this question.--BriarThe Spider 10:46, 20 December 2010 (UTC)

Oh, I loved that movie! But no, I wasn't. It was just an adverb, one that should have been pruned like almost every other adverb, if I ever bothered to edit my posts before posting them. –Jette 10:51, 20 December 2010 (UTC)

This entire section makes me think "WTB Mini Kanaxai!" --Neil • 11:53, 20 December 2010 (UTC)

The difference being that polar bears come from a chest but Kanaxai comes from some sort of random prize thing. If you had fifty bots farming the wintersday quest you would inevitably get a polar bear, but no amount of work gets you a Kanaxai (which is mostly fine; he's kind of ugly. Mallyx is much cuddlier). Sometimes I say "WTS mini polar bear, PM offers" in Kamadan while I'm on DND just to see what happens, though. Usually it's funny. –Jette 12:15, 20 December 2010 (UTC)

Stumbled across this...I'm the same guy who posted those two 1000+ data sets in the chart and was really hoping someone on staff would answer this. This year should put my lifetime runs on that quest at 4000 without a polar which is a bit absurd even among the rarer items in GW. 76.17.97.158 16:04, 20 December 2010 (UTC)

Heh, I've wasted two entire Christmas events on farming for this bear, and came out with nothing. Probably somewhere on par with IP's numbers above. Remember, that's weeks of doing nothing but those quest runs! I've actually lost sleep towards the end just to get more chances before it was all over. Not doing that again. >_< And then I heard about people getting as many as 3 of them in one event. F*** that bear. Hard. And then kill it and dump it in a ditch after taking your money back. (Too bad I can't make those prostitutes look like a polar bear, I think that would make me feel a little better.)

Anyway, these minis/items that have a stupendously low drop rate and sell for hundreds and thousands of ectos are one of the things I hate about this game. I much rather prefer something like the Heavy Equipment pack that costs lots of zcoins and takes many many hours, but at least you know for sure that after all that work you're gonna get what you were working for. Even if it cost 500 gold zcoins, that's still better than the polar bear drop rate. <_< /rant

^^THAT, by the way, is the reason I was ecstatic to hear about the way dungeon rewards will work in GW2. THANK YOU! Rose Of Kali 16:24, 20 December 2010 (UTC)

Cool. My reward system actually got considered for... something. Just sad they used the itemized one.--BriarThe Spider 16:38, 20 December 2010 (UTC)

It was the only reasonable way to go (imo), I don't think they needed an external suggestion for that one. I'm just happy for all the things they are doing right in GW2, and there are many. But anyway, this isn't the place for that discussion. :P Rose Of Kali 16:59, 20 December 2010 (UTC)

"you're the only one I know who fully understands the drop tables" - Unfortunately untrue - even I can't claim to understand the full workings of the loot system. I do know how the dungeon treasure chests work though, and can confirm that the MPB drop rates are working as designed, by which I mean absurdly low. - Joe Kimmes 20:48, 20 December 2010 (UTC)

Thanks for the quick answer.  — Tennessee Ernie Ford (TEF) 21:02, 20 December 2010 (UTC)

Would it be possible to heighten them a bit? --BriarThe Spider 00:18, 21 December 2010 (UTC)

absurdly low' that's exactly my opinion on that for the past 3 years... and then there are some people who say we do not farm to get nice things in this game. :P Jaxom 00:57, 21 December 2010 (UTC)

Okay, thanks, I appreciate it. Seems odd to me, but at least we know it's intentional. –Jette 00:58, 21 December 2010 (UTC)

Meh, I wouldn't increase the drop rate. I'd just add it to the Lair of the Snowmen chest with drop later even lower than absurd... yeah, ludicrous drop rate! MithTalk 16:03, 21 December 2010 (UTC)

Persistence pays off, I finally got one. Can proof the table results on the chest page with a screen (though I shouldn't have to...) 76.17.97.158 19:09, 21 December 2010 (UTC)

So, its intentional, eh? Is there any way you could tell us what the actual numbers are for the drop rate? It is currently "reported" at .02%. But, honestly, I have no idea how that number has been calculated since all the reports on the droprate research have been '0.' Obviously, its not 0% since the darn thing exists. --Musha 03:42, 22 December 2010 (UTC)

Musha: no drops reported tells us that the drop rate cannot be higher than a certain number — no one is reporting that it is 0.02%, they are saying that's the highest it is likely to be based on 3000-6000 chests open without any reported MPBs. The actual number might be lower...or the reporters might all be really unlucky, in which case it could be higher. (Again, ignore my comments about the specific rates; GW-Susan did the calculations and you should go by her numbers, not mine.)  — Tennessee Ernie Ford (TEF) 06:28, 22 December 2010 (UTC)

I won't say the exact number, but the estimates players have come up with are fairly close. - Joe Kimmes 20:16, 22 December 2010 (UTC)

Ok, cool. Thanks for verifying that much at least. --Musha 21:36, 22 December 2010 (UTC)

Well, perhaps you might confirm whether or not the actual droprate is higher or lower than the player-estimated 0.02%. ...Please, for the love of Balthazar, let it be higher... ~Mervil 17:51, 23 December 2010 (UTC)

Scripts

Hello, and merry xmas.

Would it be possible for you to reveal if, when it comes to scripts, the engine produce opcodes directly executable, pseudo code (semi interpreted mode), or if they only always run in interpreted mode ? What have impacted the final decision (beside the need to keep scripts consistent with the different updates of the exe)? Yseron - 90.14.228.35 02:14, 27 December 2010 (UTC)

Charrslayer shield

I know this is probably the wrong place and I'll probably get yelled at, but so far none of the other reports getting repeatedly archived since the release of EotN have had any results, so I figured I'll give it a wild shot right here. Maybe you'll see the horribly disfigured "thing" here and poke someone who can fix it.

The Charrslayer Shield is grossly under-textured and is missing most of the design, which can be clearly seen in the comparison between the shield itself and its icon. It's just a step above a basic wire frame. This is pretty much unheard of in the entire world of Tyria. Please, can someone fix this? :( Rose Of Kali 02:46, 28 December 2010 (UTC)

Joe isn't an artist, and I do think this is a job for an artist. (though I 100% completely agree that it needs to be given that seemingly missing bit). -- Konig/talk 03:24, 28 December 2010 (UTC)

people waste money on these items LOL wut? the +vs charr mod = not worth it for me... would get some of them (they have a lot of cool skins) but that one unremoveable mod is ftl.- Zesbeer 20:02, 28 December 2010 (UTC)

I'm yet to meet a single player that made one of those 'charrslaying' weapons. MithTalk 01:19, 29 December 2010 (UTC)

Hi Mith. im Neil.--Neil • 01:28, 29 December 2010 (UTC)

How dare you break my impeccable record! Anyways... Why do you make a weapon that is only effective against a kind of creature that is less than 1% of the enemies in the game and can't be modded? (And that is also rather ugly)MithTalk 01:46, 29 December 2010 (UTC)

Same reason people get "of Demonslaying" or "of xslaying" weapons. To be used better than the rest against those enemies! They're rather common in farming builds and in elite areas (demonslaying in DoA, for example). -- Konig/talk 03:22, 29 December 2010 (UTC)