OrbWeaver

That wiki entry reads more like a technical white paper than a feature documentation; I certainly wouldn't expect ordinary users to plough through all of that. It's probably about time DarkRadiant's own online documentation contained some instructions for using the game connection feature. I hadn't added it up until now because the feature was somewhat experimental and in flux, but now that it's been in the software for a few months and has visible UI elements associated, it should be documented. Of course we all know that users hardly ever read documentation, so improving discoverability of the feature is still important even if it is also covered by the user guide.

If people aren't using the feature, it is most likely due to a lack of discoverability — users don't know that the feature exists, or what benefit it can provide for them. This is probably not going to be fixed by tweaking the GUI layout or having more/fewer buttons, but there are some things which could be done to improve discoverability: Have the connection always active when the game is running, rather than having to go into the console and enable com_automation. Currently there is no way for users to even know that they have to do this without visiting the forum and reading the appropriate thread. If a user clicks on one of the camera-toolbar buttons, they will just get an error dialog and probably conclude that the feature is broken or useless. DR could dynamically detect when the game is running and automation is enabled, and automatically enable the respective buttons, rather than having them active (but not useful) at all times. This would be a clear change of state when automation was possible, which would encourage users to click on the now-active buttons to see what they did, thereby discovering and exploring the feature. There could be some kind of indicator widget in the status bar, maybe with a pop-up bubble the first time (and only the first time) automation is detected, to make users aware that there is something new they can do.

Well it seems there was something problematic with my installation, because I updated to the latest SVN, exported the normal map from GIMP in BC5 format, and it works fine in game. I'm not sure what the problem was before — perhaps I was just testing with a game DLL which didn't have the latest changes. BC4 definitely doesn't work but according to the description here, this is a 1-channel format so it wouldn't be expected to. I'm pretty sure I tested all of the GIMP formats the first time, but maybe I was just being dumb and tested BC4 multiple times but never tried BC5 @duzenko I've uploaded the images here if you want to confirm that the formats are correct. The diffuse is DXT1 and the normal is BC5, although in this particular case I would probably keep the normal map as PNG since this is less than 10% of the size of the DDS (because it's a smooth gradient normal map which compresses well with PNG). @MirceaKitsune Feel free to download the linked archive and see if you can open the _local map. I'm 99% sure you will be able to because I can read and write it fine on Ubuntu using GIMP.

Sure, I'll see if I can dig out the normal map. I'm sure it's something I'm doing wrong, like not using the right compression format or using an out-of-date or broken DLL version.

I think you're confusing the Game Connection feature which @stgatilov introduced recently, with the "Connect entities" command which has existed in DR from the beginning. Game Connection is a tool which allows you to synchronise the game camera position with the camera in DarkRadiant. It is not related to connecting entities.

I took a normal map which was working, then saved it using the GIMP plugin in 3DC+ format (I also tried using the Compressonator with various formats which looked like RGTC, e.g. "ATI2N"). After this, the normal map would either show only a single channel (e.g. horizontal bumps are visible but vertical bumps are missing), or no bumps at all; I don't recall exactly which.

Supposedly yes (I believe it's equivalent to "3DC+" in the GIMP plugin), however as I mentioned I have not been able to get this to work at all with the latest build. Only plain DXT-compressed normal maps render correctly.

Note that I have made a couple of minor improvements to the GUI. There is no full-fledged management widget or anything, but you can quickly enable/disable the camera sync or move the DR camera to the current game position using buttons on the camera view toolbar. I also rationalised some of the menu options, so enable/disable is a single toggle rather than two separate menu entries.

We're already using DDS format, and have been doing so since the initial release of the mod. The recent comments in this thread are about using RGTC compression within DDS for normal maps, which should give better quality than regular DXT1 but is not necessarily as widely supported by editing tools.

As long as the aspect ratio is the same, changing texture resolution alone should not break existing FMs. The compiled map (along with all models) will be storing texture coordinates as 0 - 1 normalised UV coordinates, which don't care about texture resolution. I believe the problem @AluminumHaste is referring to is that the actual visual alignment of certain textures has been changed, which means that they would need to be reapplied manually in DR by the mapper. This would make a direct image replacement unsafe.

"You may not include any Thief2X resources in any fan mission package" seems pretty conclusive. They are not granting permission to use their assets even within fan missions in T2, so there is obviously no general permission to include their assets in a completely different game.

Right now most GPUs are practically impossible to get hold of. There seems to be some global supply shortage. Amazon are currently listing an RX 550 (which I wouldn't use for anything more than web browsing and video playback) for over £220, which is more than I paid for my nearly-top-of-the-range R9 290 several years ago, and almost as much as 5600 XTs were selling for a few years ago. I haven't seen anything available in the mid-range affordable price bracket since the beginning of Covid. It's either ridiculously inflated bottom-dollar GPUs or the totally unaffordable super high-end.

Indeed, I'll give them credit for making it somewhat believable. They might want to improve their targeting though. If the objective is to sell women's cosmetics, a forum full of male gaming nerds probably isn't the most effective audience.

I wonder if this is going to turn out to be one of those tag-team spam threads, where one person asks for a recommendation and then their mate shows up 24 hours later to post a link to the desired site. Do girlfriends really ask their boyfriend to buy "some make-up" without specifying what they actually want (surely make-up is a very personal decision, not something you want an uneducated buyer to surprise you with), and since when do gamers use £4000 professional-grade workstation graphics cards? Apologies to the OP if this is for real, but these are some very unusual requirements.

Indeed, that would be the scientific way to do it, and very interesting if it ever happened, but I wasn't going to suggest it because there's no way it could ever be done given the ethical concerns you mention. Which does unfortunately lead to one of the fundamental paradoxes with social science research — "We can't test whether X is dangerous because we'd have to expose people to X which might be dangerous, therefore the experiment would be unethical. As a result, we just have to assume X is dangerous without ever knowing the truth!". The problem with this is that there is no way to separate cause and consequence. It might be that speech campaigns by far right organisations cause violence, but it might also be having a large number of far right individuals in a town causes both "hate speech" and violent attacks in parallel without an actual causal link between the speech and the violence (they are both consequences of a common cause). So unlike the interventional study, this "observational study" would not yield valid results. You are right, although this wouldn't be definitive proof (it is technically a post hoc fallacy), it would provide evidence if there was a consistent and widely reproduced temporal link in a large number of settings. If event B happens almost every time A happens, and there is no other obvious reason for the temporal correlation, it does point to a possible causal link between A and B which is worthy of investigation. But you'd need to be careful to distinguish between a causal relationship and just an "informational" one — did the speech cause the violence or was the speech just an indicator of some upcoming violence (perhaps by the same people who published the speech)? Are we looking at cumulonimbus clouds which cause thunderstorms or altocumulus castellanus which merely indicate they are likely? That would be interesting background information, but it's important not to extrapolate from "agreement with statements" to actual violence. There is a huge difference between somebody reporting that they have a different view of someone's rights, and actually committing violence against those rights. I wouldn't consider that a flaw at all. I would much rather see actual evidence-based policy making which changes in response to new evidence, rather than dogmatic faith healing which is assumed to be true no matter what. We should of course point out that even if we did all these experiments and gathered proof that speech could cause violence, this would still only be half of the puzzle — it would also be necessary to prove that laws against speech are effective at reducing violence. Even if you do, after all, prove that X is dangerous, you also need to prove that your laws will actually reduce the incidence of X, rather than increase it (by psychological backlash), drive it underground where it is even more difficult to keep track of, or just be widely ignored. Good catch. I never really thought of it like that, but I guess it is inconsistent not to demand evidence for restrictions on direct calls for violence. I suppose I'm not so bothered about this because specifically calling for violence is easy to identify, not likely to be confused with anything else, does not have any fundamental value in terms of discussion[1], and laws against it cannot so easily be abused to forbid dissent from an ideology in the same way that generic "hate speech" so frequently are. But there are forms of expression which could literally be interpreted as calling for violence (e.g. "#KillAllWhiteMen") which I don't think should be criminal acts, so perhaps it is better to keep an open mind even in this category. [1] Although this is a dangerous argument to make — some people would argue that dirty jokes and porn don't have any value in terms of discussion, but that isn't a good reason to outlaw them.

I think we're basically in agreement then. Speech that directly and explicitly advocates violence should not be, and generally isn't, protected as free speech. Linking other forms of speech to violence is too unreliable to form the basis of legal action, for the reasons you mentioned. Removing speech after the violence has happened is too late to protect the victim(s), and prosecuting the person who made the speech would be unjust because you would be punishing them for the behaviour of somebody else. Trying to censor speech in advance, on the other hand, involves guesswork and speculation about what might cause violence, which is unprovable and subject to the particular biases and prejudices of the censors. Indeed. This is why correlation fallacies are so dangerous. It's often the case that A happens before B but B could still have happened without A, which is why a mere sequence of events can never prove a causal relationship. Even using the claims of the violent criminals themselves are not reliable, because criminals very often try to blame other people for their actions. I've seen this before with anti-porn campaigners, who sometimes talk to convicted sex offenders who say "the pictures made me do it", then the campaigners accept this as fact and use it as justification for more censorship, without ever considering whether people in jail for sex crimes might have a strong motivation to find somebody else to blame.

There seems to be some contradiction here (or maybe I just don't understand what you propose). On the one hand, you don't want to censor actual source material or media, such as religious books, films or video games. But on the other hand you want to censor "speeches that trigger violence". By "speeches" are you exclusively referring to actual public speeches made by politicians? What if the speech is recorded in a book, or posted online? Is Mein Kampf a speech or source material? What about Protocols of the Elders of Zion? At what point does a speech (worthy of censorship) become media or source material (which should be given a pass)? And what would be the test for whether a speech actually triggers violence? Hopefully not yet another correlation fallacy, like "a violent attack happened after a speech therefore it was caused by the speech". Would you restrict the scope of the law to speeches which directly advocate violence ("Go out and beat up these people, they are scum!") or is it sufficient to argue that criticising some group of people might encourage violence against those people (as modern hate-speech campaigners often do)?

And that's an evasive non-answer. Do you or do you not agree that "The Koran causes terrorism" involves the same logic as "Hate speech causes violence against minorities"? If you agree that they are the same, why do you support censoring "hate speech" but oppose censoring the Koran? If you don't agree that they are the same, what is the important difference between the two arguments which makes the first one invalid but the second one valid?

Looking to Wikipedia for a balanced discussion of "hate speech" is like using a Fox News talk show to decide whether Trump was a good president. Although they do at least mention that the definition of the term varies between countries, and does not exist at all in certain legal systems. Which basically brings us back to square one. There is no precise, universal definition.

Why single out working-class people? Everybody is affected by psychological biases, especially confirmation bias (seeking out information that confirms your views, ignoring everything else). This applies to educated, intelligent, supposedly rational people just as much as it applies to less-educated people. One of the main purposes of free speech is to cut through the cognitive biases and air a range of opinions, allowing the best arguments to win the debate. Obviously this can't happen if powerful actors legislate to protect their views by redefining any disagreement as "hate speech", which is why more aggressive censorship laws are associated with governments who promote ridiculous views (such as the North Korean leader being some kind of nature-defying god). Sure, but the bourgeois rich are the ones who make the laws, which is why the hate speech laws we see in the West are heavily skewed towards issues which the privileged elites care about (such as using the correct transgender pronouns), whereas there is very little desire to outlaw insults against the more socially conservative working classes ("chavs", "white van man", "thick Brexit-voting racists" etc).

I think the idea of such mods is they are a 1-for-1 replacement of stock textures, chosen by players to enhance graphic quality (if they have appropriate GPU, disk space etc), rather than extra texture packs for mappers to use. Of course if the improved textures are widely considered beneficial and don't significantly increase download size or hardware requirements, including them into the core mod might be a good choice.

That's about the most honest thing that's been posted in this thread so far. "Hate speech" is an invention of the political and cultural elites to promote their particular ideology, while ensuring that those disgusting "working class people", who are too stupid to properly understand controversial issues, don't have the chance to answer back. Of course this view of certain social classes as essentially brain-dead sub-humans is a form of hate in itself, but rampant hypocrisy and double standards are standard operating procedure for most authoritarians.

We should probably avoid showing particles at all in the non-lighting preview render mode. I can't see any use for them here (other than perhaps sizing large particles relative to geometry, e.g. for fog).

Again, that's a "vaccines cause autism" correlation fallacy. There is no specific evidence that the First Amendment is the cause of homicide in America, rather than other factors such as the widespread availability of firearms, lack of access to healthcare or mental health support, extreme inequality or wider cultural issues. It is never valid to look at a particular country's crime rate, pick a random aspect of that country's laws, and then argue to those laws are the direct cause of the crime statistics. That is just political rhetoric, not science. In my experience, hyper-moderated gaming forums are some of the most toxic I've ever seen. This doesn't prove that moderation makes them toxic of course, but it does indicate that moderation isn't a total solution. Even if you ban the worst types of language, racial slurs and the like, people still find ways to indirectly insult each other with endless comments like "Wrong!", "Learn to read", "Some really stupid comments in this thread", "So much suck and fail" and so on. Perhaps such comments drive away fewer people than directly calling them a "cunt" or a "n*gger", but it's certainly not the case that aggressive moderation automatically leads to a pleasant, welcoming community. At worst, it might even lead to a community of idiots like RPGNet, where special snowflake social-justice moderators give you a long list of political opinions you must hold (even off-site) before you're allowed to use the forum.

I'm not in a position to check right now, but is there a filter in the Filters menu which might help? E.g. "Particle emitters"? If not, it probably isn't a lot of work to add one.