Wikipedia:Reference desk/Archives/Computing/2010 May 27
Computing desk | ||
---|---|---|
< May 26 | << Apr | May | Jun >> | May 28 > |
Welcome to the Wikipedia Computing Reference Desk Archives |
---|
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages. |
May 27
editwindows mobile os backwards compatible with ce.net?
editCan I run an application written in CE.net on a device running Windows Mobile? Is WM backward compatible with CE.net? Thanks —Preceding unsigned comment added by 68.28.42.225 (talk) 00:07, 27 May 2010 (UTC)
Port 22
editIs there any legitimate or accidental reason why an IP address belonging to either a DSL modem or a county government for a county in another state (not sure which it is, WHOIS says "Org-Name: Butler County") to try to access port 22 on my network, or is this a sure sign of malicious activity ir a virus infection? The IP is 66.117.197.99 PCHS-NJROTC (Messages) 02:23, 27 May 2010 (UTC)
- Port 22 is assigned to SSH so it's most likely what you describe (something looking for people running unsecured or poorly secured SSH server), but it's obviously easily possible it could be accidential. E.g. if someone typed in an IP address wrong, if someone has the correct IP address but it's been re-assigned... In terms of legitimate, well there are some possibilities like your ISP is testing you from some remote location perhaps to see if you are violating their TOS and they're using something you won't recognise to avoid tipping you off. Perhaps you're on some CIA watchlist, they're obviously not going to use a CIA identified IP if they're testing your defenses (I believe you live in the US, so it's possible this could be considered legitimate if authorised by law). In fact, if the question is 'Is there any legitimate or accidental reason why' the answer would usually be yes since it's unlikely you can't compare up with some possibility even if it's ridicilously unlikely Nil Einne (talk) 04:44, 27 May 2010 (UTC)
- According to my
/var/log/auth.log
file, I've gotten over 500 failed attempts to log in via SSH in the past few days. They're almost all malicious, and I seem to recall that this is about normal. As long as your passwords are good, you should be safe. (And it doesn't hurt to make sure that root is an invalid user to SSH into.) Paul (Stansifer) 12:37, 27 May 2010 (UTC)- There's no need for conspiracies by the CIA. This is very common. If it's more than just a few incidental logins, it's almost certainly a brute force attempt against SSH servers. The IP its coming from may be an infected machine too. If you aren't running an SSH server, or have it firewalled off, you shouldn't need to worry about those attempts. Even if you are running it, the key is to make sure you have complex long passwords. There are other security options available as well. This is very common and anyone running an SSH server will frequently see these failed login attempts in /var/log/secure. Shadowjams (talk) 20:29, 27 May 2010 (UTC)
About PC RAM
editI m having 4 gb ram, but in system properties it shows only 2.75 gb. the RAM is 2-2 gb two pieces. And my motherboard is AMD-GA-MA785GM-US2H,It has 512 mb graphic card built in. The shopkeeper said that RAM is using by Graphic Card but i m not sured. So resolve my doubt n tell me whats the problem with RAM. Reply me at <e-mail removed> —Preceding unsigned comment added by Abhi29jain (talk • contribs) 02:23, 27 May 2010 (UTC)
- I removed your e-mail address since we answer questions on the RD, and you're risking spam by leaving it in a public place; as mentioned in the header. Anyway as for your problem, at a random guess, you're using a 32 bit operating system, probably Windows which limits the amount of RAM that is addressable. If you upgrade to a 64 bit operating system like Windows 7 x64 or Windows Vista x64 or Windows XP x64 you should see all your RAM. I'm lazy to explain more particularly since I have no idea if you're going to read this since you want me to e-mail you, perhaps someone else will Nil Einne (talk) 04:38, 27 May 2010 (UTC)
- The maximum memory available in 32-bit Windows is 3.25GB. Minus 512MB used by the integrated graphics and you are left with 2.75GB. F (talk) 09:09, 27 May 2010 (UTC)
- Reference: KB929605. 124.214.131.55 (talk) 11:25, 27 May 2010 (UTC)
Problems
editI have a hp pavilion dv5-1170ei and it had problems with the audio and the quick launch buttons after upgrading to Win 7. I decided to update the bios because that was the only thing that i had not updated. Now the computer is frozen with the BIOS update still running. I once had the same problem with a dv5-1199ei and it worked fine after upgrading the BIOS. By the way i upgraded from Vista Home premium to Win 7 Ultimate.
I don't know what to do.-129.78.64.100 (talk) 03:16, 27 May 2010 (UTC)
- Restart your computer first to verify BIOS is not loading and the update operation is hung. If your BIOS fails to load, you may need to reset your BIOS to default and try to flash it again. I'd re-download the BIOS flashing program to be sure it isn't a corrupt download as well and be doubly sure you got the right BIOS update.
- If you do not know how to reset your BIOS, here is a simple article detailing a couple of methods. It does not mention, however, that you must disconnect ALL power sources from the laptop (including batteries), so please do this. Before touching any of the internal electronics, be sure as well to discharge any static electricity from yourself by touching a metal surface.
- Resetting the BIOS will also reset your time/date settings and any other changes you have made to BIOS, and so you will need to change those back.
- Given that this procedure involves working inside the laptop (and while not an especially challenging operation) it is a great way to ruin your several-hundred dollar investment if you are not careful. If you don't feel confident doing this, please take this to your local PC repair shop. -Amordea (talk) 05:35, 27 May 2010 (UTC)
keyboard commands history
editDid the assignment of Z/X/C/V to Undo/Cut/Copy/Paste originate at PARC or what? —Tamfang (talk) 06:40, 27 May 2010 (UTC)
- Yes. It is commonly assumed that they chose those because they are in a row in the lower left of a standard qwerty keyboard. However, like much of Xerox PARC's innovations, they didn't do much to popularize it. Microsoft adopted the same control keys in Windows and from there they became standard. -- kainaw™ 18:57, 27 May 2010 (UTC)
- That's not how I remember it. Command-Z,X,C,V has been the standard on Macs since 1984, while Windows originally used Alt+Backspace, Shift+Delete, Ctrl+Insert and Shift+Insert respectively. I think Windows 95 was the first version to officially switch to Ctrl+Z,X,C,V, though many programs supported them before that. -- BenRG (talk) 01:26, 28 May 2010 (UTC)
What DSL cable do I have?
editMy broadband router is connected to the DSL jack on my phone line with a thin cable that has small 6-pin connectors, the same at each end. I'm not sure if all the pins are used, can only see two wires attached. I'd like to get a longer cable, but can anyone tell me what jack it is? Seems too small for an RJ45, and I don't want to just guess.... —Preceding unsigned comment added by 86.170.214.177 (talk) 11:52, 27 May 2010 (UTC)
- Ok, I've searched some more and think it's an RJ-11. Does that sound right? Another question - in the US, is this used as a phone jack? I'm in the UK and our phone jacks are different, flatter and wider with a clip at one end. 86.170.214.177 (talk) 12:06, 27 May 2010 (UTC)
- Yes, it's (essentially) an RJ11; that's typically what's on both ends of the cable runs from the DSL splitter to your broadband router. Registered jack goes into horrible detail (strictly what you call RJ45 isn't, and what we're calling RJ11 here is really probably RJ14, but RJ11 will do you fine). FYI, the BT plug (properly called a BS6312 plug) is described at British telephone sockets. -- Finlay McWalter • Talk 15:30, 27 May 2010 (UTC)
- Being a US native, I couldn't help but giggle at this question. RJ-11's are so ubiquitous here that it's even hard for me to imagine someone not being familiar with one (which I don't mean as a slight, it just amuses me). But yes, the RJ-11 (and its cousins, but they tend to get lumped in with the others and labeled as RJ-11 as well since they all use the same size/shape plug but have differing numbers of connectors, with the RJ-11 being the most common type) is our landline standard here in the US. But hey, we're also still using imperial units, so we might just be backwards in this department as well. -Amordea (talk) 21:21, 27 May 2010 (UTC)
- They're actually very common in the UK too. As with the US (and I think just about everywhere) RJ11 is the equipment-end of cables for phones, fax machines, modems and the like. BS6312 is only for the utility end. -- Finlay McWalter • Talk 21:27, 27 May 2010 (UTC)
- What you want to look for (online or in a shop) is a 'modem cable'. They are available in various lengths from many retailers, even supermarkets. If you are unsure, take the old cable to a shop to be sure you get the right one. The cables have a RJ-11 plug on the end which goes into your broadband modem, and the other end has a standard BT plug on it. Alternatively, if you have a broadband filter at the wall socket, you will probably need a cable with a RJ-11 plug on both ends. Astronaut (talk) 01:27, 29 May 2010 (UTC)
Customizing Win7 taskbar/tray
editI could swear that I read a tip the other day about how to force an app to display in your system tray (AKA notification area) instead of the taskbar, but I don't remember where. I should have bookmarked or wrote it down. Anyway, does anyone know how to do this off the top of their head? I just want an app to display in the tray to conserve desktop space. I'm using Windows 7. A Quest For Knowledge (talk) 12:26, 27 May 2010 (UTC)
- I believe what you are referring to requires individual program support. Some programs are built to run in the notification area, some are not. What you read could have been referring to the ability to show/hide notifications. But in Win7, you can easily pin programs to the taskbar anyway, so I am not sure why you would want to put them in the notification area (which is not what it was designed for anyway). To pin an item to the taskbar, you can simply click-and-drag the executable link to the taskbar and drop it onto it. If taskbar real estate is an issue, consider using small icons (right-click the taskbar, click properties, click the "use small icons" checkbox). -Amordea (talk) 19:17, 27 May 2010 (UTC)
- Er, did you try a web search? RBTray, Trayconizer, PowerMenu, lots of others. -- BenRG (talk) 20:16, 27 May 2010 (UTC)
- Huh, I had no idea there was so much support for this sort of thing, even if third-party. My (admittedly very brief) web search did not turn any of these programs up. Well, learn something new every day. -Amordea (talk) 20:50, 27 May 2010 (UTC)
- Well, the issue for me was Windows Live Messenger. On Windows 7, it has two entries in the taskbar. Anyway, the tip I was thinking of was in the June issue of Maximum PC magazine. Basically, you go into the app's shortcut, change to the Compatibility tab and select Windows Vista. A Quest For Knowledge (talk) 19:26, 28 May 2010 (UTC)
Outlook mail accounts
editIs it possible to use yahoo or gmail accounts with Microsoft Outlook? —Preceding unsigned comment added by 122.255.2.187 (talk) 14:39, 27 May 2010 (UTC)
- Yes, at least for Gmail. You need to enable IMAP in Gmail's settings and then configure it according to these instructions. It works perfectly well. -- Finlay McWalter • Talk 15:04, 27 May 2010 (UTC)
- I've never done it with Yahoo! Mail; it seems the free version now supports POP3 (a simpler, less powerful counterpart of IMAP) - some instructions are here. -- Finlay McWalter • Talk 15:14, 27 May 2010 (UTC)
Megaupload
editIf I upload a common file to megaupload, say for example the vlc-media player installer file, the upload progress bar goes straight to 100% without any uploaded data coming from my connection (I monitored it upload and download speeds and amounts). When I download the file, it is in perfect working order. Obvious I didn't upload this file, so does megaupload detected if the someone else has already uploaded the same file, and simply link my upload to it? 82.44.55.254 (talk) 15:22, 27 May 2010 (UTC)
- Probably. It would be a pretty easy thing to do, by just exchanging md5 hashes of file contents, and it would save time for you and space for them. It's how I would do it, anyway, if I were running a site like that! --Mr.98 (talk) 15:58, 27 May 2010 (UTC)
Windows and Linux virus
editCan a virus pass from one OS to the other? It could be a kind of "write once, run everywhere" virus. I often use the same USB in a Windows and Linux machine, could that be a source of infection?--Mr.K. (talk) 17:30, 27 May 2010 (UTC)
- There are, in essence, no Linux virusses. While some proofs of concept have been made, and one can surely speculate about how one might exist, there is no evidence at all that any Linux viruses exist "in the wild". In theory Wine runs many Windows programs, so a Windows virus might (again, in theory) run in Wine, but in practice Wine doesn't do many of the lower level things that virusses reply on, so again there's no evidence that a Wine-compatible virus really exists in the wild. -- Finlay McWalter • Talk 17:36, 27 May 2010 (UTC)
- The Linux malware article talks about viruses and trojans written for linux, and how they're "on the increase in recent years and more than doubled during 2005 from 422 to 863". Also, to address the OPs question, could there possibly be viruses or malicious programs written in crossplatform languages like Java which could affect different operating systems? That seems like a possible threat, although I personally don't know if such things exist. 82.44.55.254 (talk) 18:11, 27 May 2010 (UTC)
- Unlike programs written in machine code, or in a language such as C that compiles into machine code, programs written in Java compile into Java byte code, which needs the Java Runtime Environment to run. At least in my experience, this prevents them from (1) starting on their own without the user's intervention, and (2) interfacing with the system directly. Of course, there are exceptions to both, but I hardly think they matter. A malicious native program could very well start a Java program on its own, but for that, you would need a native program in the first place, and it could very well do the malicious stuff itself in that case. Java programs can do damage to the system in a system-legal manner, for example by deleting files, but they cannot actually call low-level system calls. JIP | Talk 19:03, 27 May 2010 (UTC)
- I see no reason why there couldn't be a Java-hosted virus. Microsoft Word macro viruses were all the rage once. -- BenRG (talk) 01:17, 28 May 2010 (UTC)
- I understand that. It seems the issue here is to make people execute malicious code in the environment it runs in. For machine code, this isn't an issue, because such code can run directly under the OS without human intervention. But for things like Java byte code or Microsoft Word macros, there is human intervention required to start up its host environment. A "social attack", like providing links to a malicious Word document and writing "Look at this document!!!! It's kewl stuff!!!111!!" would be one way. Once the user opens the document in Word, the viruses go about their business. But if the user doesn't do that, the viruses can't do squat. It would be the same thing with Java viruses. JIP | Talk 19:31, 28 May 2010 (UTC)
- Strictly speaking, nothing is going to start up on its own. Every attack is either a social attack or an attack on a vulnerable service that does things for the user (sometimes a combination of both). You could even imagine an attack that requires Java because the vulnerable service allows scripting in Java (ewww) and has a hole in its sandbox. So I can imagine a totally Java-based virus (maybe with a bit of system-specific code for installing services, etc.), but there are a lot of reasons why such a creature isn't likely, not least of which is that Java isn't going to be too familiar a language to people who are used to thinking about buffer overruns, etc. Paul (Stansifer) 14:19, 29 May 2010 (UTC)
- I understand that. It seems the issue here is to make people execute malicious code in the environment it runs in. For machine code, this isn't an issue, because such code can run directly under the OS without human intervention. But for things like Java byte code or Microsoft Word macros, there is human intervention required to start up its host environment. A "social attack", like providing links to a malicious Word document and writing "Look at this document!!!! It's kewl stuff!!!111!!" would be one way. Once the user opens the document in Word, the viruses go about their business. But if the user doesn't do that, the viruses can't do squat. It would be the same thing with Java viruses. JIP | Talk 19:31, 28 May 2010 (UTC)
- I see no reason why there couldn't be a Java-hosted virus. Microsoft Word macro viruses were all the rage once. -- BenRG (talk) 01:17, 28 May 2010 (UTC)
- Unlike programs written in machine code, or in a language such as C that compiles into machine code, programs written in Java compile into Java byte code, which needs the Java Runtime Environment to run. At least in my experience, this prevents them from (1) starting on their own without the user's intervention, and (2) interfacing with the system directly. Of course, there are exceptions to both, but I hardly think they matter. A malicious native program could very well start a Java program on its own, but for that, you would need a native program in the first place, and it could very well do the malicious stuff itself in that case. Java programs can do damage to the system in a system-legal manner, for example by deleting files, but they cannot actually call low-level system calls. JIP | Talk 19:03, 27 May 2010 (UTC)
- The Linux malware article talks about viruses and trojans written for linux, and how they're "on the increase in recent years and more than doubled during 2005 from 422 to 863". Also, to address the OPs question, could there possibly be viruses or malicious programs written in crossplatform languages like Java which could affect different operating systems? That seems like a possible threat, although I personally don't know if such things exist. 82.44.55.254 (talk) 18:11, 27 May 2010 (UTC)
Making sense of syntax
editI am having problem understanding the last line of the Python code below.
If it were object.method(), it would make sense for me, you would be calling the method of an object. But, this way, it seems like two methods concatenated... :(
>>> from urllib import urlopen >>> url = "http://www.gutenberg.org/files/2554/2554.txt" >>> raw = urlopen(url).read()
--Quest09 (talk) 17:50, 27 May 2010 (UTC)
- In Python, pretty much any value is an object.
urlopen(url)
returns an object which has aread
method. It's as if you did
temp = urlopen(url) raw = temp.read()
- but in one line. --Spoon! (talk) 18:13, 27 May 2010 (UTC)
- I don't understand Python, but from a quick look at the above, it looks pretty similar to the situation with languages like Java or C#. If a method returns a reference to an object, you can call other methods on that object simply by concatenating the method calls. The only difference is that return values from methods aren't lvalues, so you can't do things like:
PlaceToStoreObject(location) = new Object();
(not legal)
- but you can "cheat" with special array syntax, like so:
GetObjectArray(location)[0] = new Object();
(legal, I think)
- C and C++ have the
*
operator, which transforms rvalues into lvalues by dereferencing them, but Java and C# do not. JIP | Talk 18:57, 27 May 2010 (UTC)- From a language perspective, it's perfectly fine in C++ to chain function calls together using the dot operator (a reference-returning function call is an lvalue), although this does go against the Law of Demeter design principle, which basically says: "Use only one dot!"
decltype
(talk) 10:05, 28 May 2010 (UTC)- Drat, I always forget about the C++ reference type. I suppose that's because such a thing doesn't exist in either C, C# or Java. I figure the language still has sane enough type checking so that returning a reference to a local variable, like
int a
, is legal, but returning a reference to something likea+1
isn't. JIP | Talk 19:34, 28 May 2010 (UTC)- Kind of. Something like
int& f() { int i = 0; return i; }
is allowed, and is a common source of undefined behavior, although most compilers do warn about obvious cases like that. However, you cannot bind a non-const reference to an rvalue, so something likeint& f() { int i = 0; return i + 1; }
should not compile.decltype
(talk) 20:16, 28 May 2010 (UTC)
- Kind of. Something like
- References and lvalues are irrelevant here, surely? C++ allows you to call methods on temporary objects. Even non-const methods. Even the
operator=
method. You can assign to a function return value (of class type) in C++ simply because assignment is a method call in C++. In other popular languages assignment is magical, and the things you can assign to (like variables and slots) are not objects and hence can't be returned from functions. -- BenRG (talk) 21:06, 29 May 2010 (UTC)
- Drat, I always forget about the C++ reference type. I suppose that's because such a thing doesn't exist in either C, C# or Java. I figure the language still has sane enough type checking so that returning a reference to a local variable, like
- From a language perspective, it's perfectly fine in C++ to chain function calls together using the dot operator (a reference-returning function call is an lvalue), although this does go against the Law of Demeter design principle, which basically says: "Use only one dot!"
- I don't understand Python, but from a quick look at the above, it looks pretty similar to the situation with languages like Java or C#. If a method returns a reference to an object, you can call other methods on that object simply by concatenating the method calls. The only difference is that return values from methods aren't lvalues, so you can't do things like:
- Python syntax is mostly about expressions and statements. Expressions have values, and statements do things. (Confusing the issue somewhat is that expressions can also do things when they are encountered.) So
1
,1+1
,"google.com"
,urlopen(url)
urlopen("google.com").read()
are all expressions. On the other hand,x=4
,import foo
, andreturn 1+1
are all statements (but they contain expressions, which I've underlined).
- Every value (and thus, the result of any expression) in Python is an object, and all objects have methods (to see what methods an object
o
has, look at the value ofdir(o)
). Try the following expressions out (interactively, Python treats an expressione
as the statementprint e
):
dir(1234.5678) (1234.5678).as_integer_ratio() #the parentheses are only needed because "." has two different meanings. (2).__add__(2).__mul__(3) # "2+2*3" is actually short for this "hello".upper()
- Expressions can contain other expressions, arbitrarily deep. Statements can contain expressions, but expressions can't contain statements. See also Python syntax and semantics. Paul (Stansifer) 19:08, 27 May 2010 (UTC)
- It's really not true in Python that "expressions have values and statements do things". The expression-statement distinction in Python is rather arbitrary.
dict['key'] = value
is a statement butdict.update(key = value)
is an expression,exec
is a statement butexecfile
is an expression,import
is a statement but__import__
is an expression, etc. Most method calls are of the formidentifier.method(args)
, and it's not necessarily obvious at first glance that the thing on the left of the.
is allowed to be an arbitrary expression (after all, the thing on the right isn't). One could even argue that restricting the thing on the left would be consistent with Python's avoidance of compound operations likea[i++]
andreturn x[y] = z
. So I don't think that this behavior follows from some grand principle of Python design. The only principle of Python design seems to be "whatever GvR likes". (Though I do think he has good taste, overall.) -- BenRG (talk) 01:10, 28 May 2010 (UTC)- Ah, you caught me in my idealism there. The only real distinction is that expressions have values, and statements don't. Statements exist solely for effect. But expressions can have effects, too, ruining the beautiful symmetry ) : — I feel like it's clearest to avoid effectful expressions, but this is sadly impractical in Python a lot of the time. Not to mention pretty much all other languages. Paul (Stansifer) 01:42, 28 May 2010 (UTC)
- It's really not true in Python that "expressions have values and statements do things". The expression-statement distinction in Python is rather arbitrary.
- I don't think the confusion is with statements/expressions, but confusion about objects and operators. Within an expression, the general scheme is "get object(s), apply operator to object, repeat". Let's take "
raw = urlopen(url).read()
" as an example. What happens is that the object named as "urlopen" is obtained, and the "function call" operator is applied, with the object named as "url" as a parameter. This operation returns an object. Then the "obtain member" operator is applied to that object (the one that was returned), with the identifier (string) "read" as a parameter. This returns an object (a member function object), on which the "call function" operator is applied and an object is returned. Assignment is a statement, rather than an operator, but for the last step, a label of "raw" is applied to the object which is the final result of the right hand side. In general, you can have arbitrarily complex statements to the left of the "obtain member" dot e.g. "(str(45/5) + " MI$$I$$IPPI ".strip().replace('$','S')).lower()
" -- 174.24.200.38 (talk) 03:12, 28 May 2010 (UTC)
Blender question
editI am a Blender noob. I am playing around with the interface and learning how to make some simple objects.
One thing I notice is that all of my renders, no matter how much I play with their materials, look very shiny and flat and very 1980s computer graphics (kind of like this. What I'd really like to know how to do is to make them look more "organic" in the sense that even though they are of, say, metal things, they have a nice texture to them. A good example of the effect I'd like to achieve is in this image here: File:Engine_movingparts.jpg.
What exactly gives it that organic feel, like it was made out of a rough plastic? I'd be eager for any explanations even if they are not Blender specific. I assume it is some kind of settings relating to how the objects reflect light, and maybe some kind of very subtle texture bitmap applied to it, but I'm kind of clueless in my essential noobiness. --Mr.98 (talk) 20:16, 27 May 2010 (UTC)
- This is of interest to me too. I've played with Blender but never gotten very far. One thought is that the more organic pictures might have a subtle depth of field. The other issue is that the raytracing / reflections are much more subtle on the better images. Shadowjams (talk) 20:24, 27 May 2010 (UTC)
- That image looks like it benefits from radiosity, which simulates light diffusely bouncing between parts of the object. Also, turning down specular highlights tends to make things look more plasticy and less metal. Paul (Stansifer) 20:30, 27 May 2010 (UTC)
- This Blender page discusses the various parameters - "gloss" seems to best describe what you're talking about, in the Blender sense. Diffuse reflection covers the physics, and points us to Radiosity (3D computer graphics) and Photon mapping. Both of these are rather expensive, so it'd be tempting for some applications to apply a
globallocal illumination like Phong shading with some anisotropic noise, and that'll look reasonable. If you really needed a brushed metal appearance, you'd probably look to implement one or several shaders that did a noise-guided texture and possibly bump map, but that's probably overkill for most applications. -- Finlay McWalter • Talk 20:34, 27 May 2010 (UTC)
- Thanks much for the comments. I've been popping around these links and others and I think what I need to do is really grok how ambient occlusion works, because I think that's the effect that I'm really going for, in the end. --Mr.98 (talk) 20:38, 27 May 2010 (UTC)
- Did you realise that the image you linked appears in Blender_(software)#Features where it is described as " with ray tracing and ambient occlusion using Blender and YafaRay" - looking at the blown up image I notice a lot of noise - indeed YaFaRay uses montecarlo based approximations for illumination - in fact the image seems to gain a pleasurable graininess from the approximate nature of the sampling - more accuracy might result in a 'flatter' looking surface - and in fact look less good (subjectively).77.86.125.207 (talk) 20:51, 27 May 2010 (UTC)
- Thanks much for the comments. I've been popping around these links and others and I think what I need to do is really grok how ambient occlusion works, because I think that's the effect that I'm really going for, in the end. --Mr.98 (talk) 20:38, 27 May 2010 (UTC)
- Yes, the graininess comes from having a low number of samples (probably the default 5 samples). I like it that way too, looks like a mezzotint. 81.131.23.148 (talk) 03:53, 29 May 2010 (UTC)
- I think you want to turn down the shinyness of the surface - ie reduce specular reflections - they do look '80s'. In phong shading either just it turn off or reduce the value of n in the equation given in Specular_highlight#Phong_distribution (1 or 2 maximum). Possibly this hasn't been done in the example image - it's difficult to tell since the angle of illumination wouldn't show any specular highlights anyway - I think rough montecarlo sampling is the thing that is causing the effect you like in the image.77.86.125.207 (talk) 21:19, 27 May 2010 (UTC)
- It's enormously easy (ambient occlusion) - you just click the button to turn it on. You can then set the energy and decide whether it should add light, shadows, or both, but the default settings are usually fine. 81.131.39.118 (talk) 21:31, 27 May 2010 (UTC)
- If it's a question of understanding the process behind ambient occlusion it's easy to explain - each point visible in the scene has many rays cast from it in all directions outwards - if the ray escapes the scene (ie no other object is in the way) it contributes to the ambient lighting - obviously you get a percentage of rays escaping - and convert that to a brightness figure.77.86.125.207 (talk) 23:05, 27 May 2010 (UTC)
- It's enormously easy (ambient occlusion) - you just click the button to turn it on. You can then set the energy and decide whether it should add light, shadows, or both, but the default settings are usually fine. 81.131.39.118 (talk) 21:31, 27 May 2010 (UTC)
- The User:Wapcaplet created the file and says will send you the scene file if asked - quickest way to learn is to copy modify.77.86.125.207 (talk) 21:23, 27 May 2010 (UTC)
- I have found that ambient occlusion (located among the World buttons, on a tab marked Amb Occ) is a great help in this. It lights the scene as if the light was coming from everywhere, which means that the shadows fall in the cracks and corners. (You can, if you wish, light the scene using only ambient occlusion and no light sources, although I guess you'd lose spectral highlights.) For an additional hit of realism you could deal with the leakage of colours from one object onto nearby objects ... I think this is radiosity, as Finlay mentioned; I've never felt the need to try it, ambient occlusion was the key thing for me. 81.131.39.118 (talk) 21:29, 27 May 2010 (UTC)
- Back in the days for easy lighting I just use Yafray and its Global Illumination, looks great every time (if not a bit slow). It seems they've removed Yafray support in 2.49 though, but the Blender renderer seems a lot better too. --antilivedT | C | G 11:32, 28 May 2010 (UTC)
- The image to the right is great. --Andreas Rejbrand (talk) 23:37, 28 May 2010 (UTC)
Sibelius files
editWhy can't Sibelius files be uploaded into any Wikimedia project? --84.62.209.203 (talk) 20:44, 27 May 2010 (UTC)
- We only allow uploads of files in "free" formats, that is formats that can be read and written by open source and free software, and that aren't encumbered by patents. As far as I know, only Sibelius itself adequately reads and writes its file format. -- Finlay McWalter • Talk 20:48, 27 May 2010 (UTC)
- If you have some, why not convert them to MIDI files so that we can all read them? Dbfirs 15:47, 30 May 2010 (UTC)
Budget program recommendation
editMy husband and I are looking for a computer program to keep track of our budget. We prefer using the envelope method.... For a while we used Quicken and did complicated things to make it act like the envelope method, but we're tired of that now, and not looking to pay for an updated version of Quicken.
I have a Mac, and he has a PC, so either kind of program works.
We have used the trial versions of Budget and Common Cents and ran into major frustration with non-intuitiveness and outright bugs. I liked the trial version of Moneywell but my husband doesn't want to spend quite that much unless we know it will work very well.
Here are the features we want (we don't need any other fancy widgets but can ignore them if necessary):
Ability to keep track of ~30 envelopes/categories, ability to import transactions in QIF, CVS or Quicken format (just one is fine), ability to split transactions between more than one envelope/category, ability to easily transfer funds from one envelope/category to another, ability to easily start envelopes/categories with a certain amount of seed money (it's crazy how hard this is in some of hte programs we've looked at).
extra bonus: ability to access the budget from an iPhone would be nice but certainly not necessary.
Thanks for any help you can give! —Preceding unsigned comment added by 98.240.235.130 (talk) 22:00, 27 May 2010 (UTC)
- I had never heard of "the envelope method", but our article Envelope System lists some software that apparently uses it. Comet Tuttle (talk) 23:03, 27 May 2010 (UTC)
- Shouldn't that article be at Envelope system, to comply with our capitalization standards? Anyway, I created the redirect Envelope method. Buddy431 (talk) 01:48, 28 May 2010 (UTC)
- Take a look at GNUCash. It is double entry accounting (not as complicated as it sounds) and also has a budget function. Shadowjams (talk) 06:20, 29 May 2010 (UTC)