Niemand

Unfortunately, I have to respond by asking some questions: What else does the error message say? (We need to eliminate or rework those dumb error numbers; they look like useful information but mostly aren't.) Are you on Windows or Mac OS? (I can't find anything in the Mac code that might generate an error number 11, but I may not have looked closely enough.) What do you mean by: Quote: Whenever I make a scenario file it is visible on the open-menu and can apparently be copied to the scenario section while still in open-menu but it is invisible/absent if you try and look for the actual folders. It may help to note (if you haven't already) that when the editor generates a scenario, it places the whole scenario directory inside the Blades of Avernum Files directory, rather than the Blades of Avernum Scenarios directory. You can use the editor to edit the .bas file while it's in either location, but the game can only find the scenario if it's moved into the latter location.

I just finished The Gathering Storm as well, and quite enjoyed it. I've also been working on reading The Fountainhead, which is a stunningly pointless book.

It certainly does not suck; I enjoyed it.

Are you sure that you have set each character's memory cell 3 correctly to correspond to the node at which it should being talking? You may also want to assign a unique personality number to each character and to the corresponding nodes, although I don't think that leaving that out would produce problems like what you describe. (I regret that it's been rather a long time since I've actually written any dialogue code.) To your second question: That depends, possibly, on whether you want to literally open a door, or unlock a locked door. For either one, you want to use the code section of the relevant dialogue node. To open a closed door, all you have to do is place a call to set_terrain() in the code section, which sets the terrain on the space in question to be the appropriate open door. If you want to unlock a locked door, the easiest way would be to set the memory cells of the door so that it checks a Stuff Done Flag to see whether it is in fact locked. Then all you have to do is use the set_flag() call from your dialogue node's code section to set the flag you chose to be used by that door.

Simplistically it would seem that Newton's law of cooling would imply that the warmer house would lose heat faster, and so would require more fuel to maintain its temperature. Off of the top of my head I can't think of any reason that this wouldn't describe your example.

That fix doesn't solve everything, but it does get to the point of displaying a properly sized window with the controls laid out correctly. That window still appears in a bad place, jammed up under the menu bar, and the controls themselves don't draw properly, but it's progress.

Quote: As I recall, the size is not consistent between times; somehow either the variables to hold the size are not being initialized, or the size is being grabbed from out-of-bounds memory (or just the wrong place in general). I think it's the former; I just can't understand how. I'm trying to understand where a window size is supposed to come form in the first place: there appears to be none specified in edit-terrain.xml at all, for example. Is it supposed to be dynamically sized to fit by recalcRect()? EDIT: Yeah, I think that at least one problem is manifesting in recalcRect(). I got sick of stepping through the loop over all of the controls so I just added a print statement for the new bounds values after each iteration: Code: 8cTextMsg "$}.0~.Wtw{hOO$KF!5wz~Wx%MU;/pK @/eQvg+l"r:588 b:3998cTextMsg "$~z0^Sph9Oa[3\U,_7Z+),LZ<UrAO]A"r:588 b:39910cTextField "arena"r:588 b:3999cLedGroup "blockage"r:10047 b:12933 It looks like the size of an LED group is being drastically miscalculated. Aside: What on earth is the reason for generateRandomString()? If you need unique identifiers, why not use sequential numbers? EDIT 2: Code: --- button.cpp (revision 142)+++ button.cpp (working copy)@@ -265,6 +265,7 @@ } void cLedGroup::recalcRect(){+ SetRect(&frame,0,0,0,0); ledIter iter = choices.begin(); while(iter != choices.end()){ if(iter->second->frame.right > frame.right)

O_0 That was. . . impressive. It's been over ten minutes since I ran that thing, and my system still hasn't fully recovered. As best I can tell, what it did when I tried to edit a terrain was that it created a gigantic window, far larger than my screen, and tried to tile it with one of the background textures. It was so large, though, that the space needed to buffer it greatly exceeded the capacity of my video RAM. By the time I was able to get past it scribbling all over my screen and force quit it, it had forced everything else but the OS out of RAM and into swap space. During all of this, a lot (thousands) of log messages identical to this were printed: Code: time hostname BoE Scenario Editor[11402] <Error>: kCGErrorFailure: CGSLockWindowRectBitstime hostname BoE Scenario Editor[11402] <Error>: kCGErrorFailure: _CGSBindWindowBacking: cannot map backing data shmemtime hostname BoE Scenario Editor[11402] <Error>: kCGErrorFailure: _CGSLockWindow: Unable to lock window Anyway, now that I can very much see it malfunctioning, I'll try to get some idea what's going wrong. I suggest that we take this to a thread of its own, as it may require more discussion and it would be nice to cut down on the thread hijacking/rambling in this forum.

WebKit would be massive overkill for just parsing some XML. I think CM is using TinyXML/TinyXML++, which is simple, portable, and gets the job done pretty well. (And, as a result of this, I would assume that the OS update would make no difference whatsoever for a parsing problem, although I doubt he was seriously suggesting that it would have.) I'm also rather baffled as to how reading the dialog dimensions could fail. I wouldn't mind taking a poke at it to see what's going on; how would I run whatever test you've been using to debug it, Celtic Minstrel?

While it's far from the only influence I can think of, one of the more unusual ones that springs to mind is a book that my father read to me sometime around when I was ten to twelve. Since I had been quite small my father had traditionally read aloud to me for a maybe twenty minutes to half an hour before I went to bed every night. As I got older and could do a lot of reading for myself, he transitioned to heftier and more serious books that I wouldn't have been likely to read or get through on my own, and one of the last of these was Mila 18. For those not familiar with it, the book is a dramatization of the events leading up to and during the Warsaw Ghetto uprising of 1943. The story was incredibly grim, and somewhat frightening to me at the time; I still recall vividly how I imagined the scenes that were portrayed. It wasn't until some years later that I learned how closely the book was based on actual history.

While I can't speak for why it doesn't yet work, as I understand it the idea is that this new dialog format can then be interpreted and displayed by code using Carbon, Win32 APIs, wxWidgets, or whatever, unlike the old classic Mac OS resource formats, which could only be read and displayed by functions specific to Carbon.

Correct math, as well as correct definitions of the Celsius and Fahrenheit scales, one would conclude.

After glancing at the town's script, it looks like there may be a bug such that you are now stuck, no allowed action in the game will get you back in the boat. A tentative solution: Use the Character Editor on your save file. (You should probably make a backup copy of it first.) Click the 'Advanced Features' button and choose 'Set Stuff Done Flag'. If you want to proceed, click yes, and then in the following three boxes enter the numbers 6, 15, and 0, in that order, one per box. Save your changes, and load the modified save file. You should now have the option to get in the boat again. From my quick examination it looks like the above should work, but I have not tested it, and so you may want to wait for word from Ephesos, the maker of the scenario in question.

For Mac OS there exists a GUI version which allows the user to select a file and then writes the report to a window. It does nothing that the command-line version didn't do, but it's a bit more convenient. I have a somewhat patched version of Alint which I use in my experimental Ascript editor; the idea was that I would get Ascript in good working order and then release it and the fixed-up Alint source. I do genuinely intended to finish Ascript, I just sort of. . . haven't gotten around to it. And so the Alint fixes languish on my hard-drive. I don't recall any of them being sensational, but if you would like I could make a list of them, and likewise if you have pinpointed particular bugs that ned fixing, I could see about doing that and then make the source available ahead of Ascript. Finally, I do have to say that I agree with Nikki in that Alint is already 99% correct, so our thirst for new scenarios vexes us more acutely. A scenario may be more work, but think of the glory, the everlasting fame, the riches of recognition which will surely be yours! P.S. I have just realized that I hate both Alint's reporting of incorrect argument types ("want 'ii'"? How is a designer supposed to know what that means?) and Ascipt's reporting of Alint-detected errors (In one place it marks that an error has be found, but won't say what it is, and in another it tells what the errors were, but doesn't give itself enough space to show it all without the user resizing things.) Ugh. I resolve to address both of these unsightly blunders. Possibly beginning right now, as it would be more interesting than my E&M exam. EDIT: CM, I could be mistaken, but I was under the impression that all of us who have weighed in on this discussion are male, and I was specifically convinced of this in the case of Nikki, to whom I think you were referring.

There's a snow storm going on, and I couldn't sleep, so the obvious thing to do was to wander around out in it for an hour with a camera. Resulting Pictures I was eventually forced to turn back due to poor visibility; not in general, but because of the accumulation of snow on my glasses. In fact, for a while I pretty much couldn't see what I was taking pictures of. I also ended up having to discard a number of pictures as later examination showed that they had been taken with the camera lens entirely covered with water. Luckily, the camera dried off in my pocket when I took a break for most of the way back, and so I was able to get some good pictures in the forest. I think that pictures 16 and 21 are my favorites from this set, though. The difference depending on where I was was drastic; when I was out by the lake I couldn't see because of stuff blowing around, but I could hear the lake water crashing around most impressively. In the woods it was incredibly quiet, and the trees filtered out much of the blowing snow, so I could see well enough to appreciate the odd twilight look; it was quite bright given the time was somewhat after midnight.

I can't necessarily answer for ES's preferences, but a PM or an email is the typical way to submit beta feedback.

Ouch. I hope that heals soon, as I can only imagine it is most unpleasant.

I was going to point out that you have to confirm that you have read the rules to register an account, but checking the registration page reveals that in fact you only have to agree to be bound by them. This interacts curiously with the fact that rules do, actually, explicitly require all members to read them.

Definitely. Many of us hadn't any when we arrived, and standard procedure for the rest is to check it at the door. Where it tends to get consumed by the fluffy turtles.

I would be happy to convert graphics files for you, however, it might be best if I don't promise to test, since I'm still only part way through doing some testing for Nikki.

Yes, Avernumscript has a while loop construct. (foreach makes no sense in a language with no collections or range objects ) Be aware, though, that iterating over most or all of a zone's spaces can put you dangerously close to the node execution limit. Code: //assuming variables i, j, and height exist//height is set to the height to which to floodi = 0;while(i<64){ j = 0; while(j<64){ if(get_height(i,j) < height){ set_terrain(i,j,[number of water cube terrain]); set_terrain(i,j,[number of water floor]); set_height(i,j, height); } j = j + 1; } i = i + 1;}

Or in any other font which includes the character lowercase i, assumedly. Particularly since, as best I can tell, both i characters in your post are identical, being the code point U+0069, and thus no matter what font you use (excepting fancy behavior like the automatic application of ligatures) they will appear identical. Also, I'm not aware of that character having a particularly special appearance in the Arial typeface or the Microsoft Unicode Arial font specifically. In fact, I just tested, and the only difference from my system's Helvetica glyph is a slightly different height and baseline.

Originally Posted By: Nalyd We have avatars, Alo. He knows this; I'm pretty sure he was speaking about historical reasons for frequent PDN changes.

A few people here switch displayed names regularly, but have sort of canonical names by which they the are known. Primary among these are Alorael and Slartucker. Note that the names we typically use to refer to these persons do not necessarily have anything to do with their given names, but are just additional names they assigned to themselves on these and related boards. Only a few people here talk in the third person, and it is generally supposed to have nothing to do with displayed name switching, although now that you mention it, it does seem like there might be a rather marked correlation. Clearly we need a statistical study. Finally, since you have mastered differentiation between 'there' and its homonyms, llloyd, I suggest you tackle punctuation next.

A lot of your problem sounds like it may be the fragmentation of the data into many small files. This often drastically slows down processes like transfer or deletion. If it took 3 hours just to mark the files as deleted, I would expect it to take at least that long to pour them into the network connection for uploading.