Wikipedia:Reference desk/Archives/Computing/2010 August 4

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.

August 4

Shady Facebook Group

I've been invited by my Uncle Don to join the quite obviously spam group "Make Money Online" ([1], if that works for you guys). I'm quite confident that Don didn't really create this page himself, not least of all because all of his aunts are dead, while the creator of the group claims that "his aunt has been making tons of money...". However, from my end of things, it does look like Don's account created it. How did this likely happen? Did Don give his password to someone he shouldn't have? Or did I do something wrong on my end to invite this sort of thing? Buddy431 (talk) 01:39, 4 August 2010 (UTC)

I can't view the page due to not having a FB account... but have you considered directly asking your uncle if he created it? 24.189.87.160 (talk) 06:56, 4 August 2010 (UTC)

I've had probably 3 friends send me such spam messages over the past year. They attributed the problem to a computer virus, which may mean the virus snooped packets, or installed a keylogger, to log their usernames and passwords when they logged onto Facebook, and either (a) sent all the info to the bad guys, who have a script to make all the victims join this group and spam their Facebook friends about it; or (b) the virus takes care of these tasks itself without needing to send any info out to the bad guys. Comet Tuttle (talk) 16:57, 4 August 2010 (UTC)

Sometimes multi-level scams like these come with canned backstory's that you're just expected to copy/paste. APL (talk) 19:31, 5 August 2010 (UTC)

What is this image process called, and how can I do it?

Hello! Given this:
+ I want this -->
That is, is it possible for image software to take two single-layer near-identical images and extract the differences to a transparent background? Is there a name for this process? Is it something I could do with GIMP or ImageMagick? Any pointers how? Thank you!--el Aprel (^facta-_facienda) 03:27, 4 August 2010 (UTC)

It is called difference masking. I don't know about those programs, but in Photoshop, you open both files, then drag one image onto the other and change the blend mode of the top layer to "Difference." Then, you go to Select --> Color range and select the background. Hit DEL, and remove the bottom layer. You now have the objects on their own layer with a transparent background.--Best Dog Ever (talk) 04:43, 4 August 2010 (UTC)

Some versions of photoshop leave some weird dithering issues around the edges. This is generally the right approach, but computationally the issue's pretty straight forward. Shadowjams (talk) 06:48, 4 August 2010 (UTC)

You can write a simple script in MATLAB or GNU Octave (or any language that is convenient to load the images as arrays of pixel-values). Pixel-by-pixel subtraction of Image 2 from Image 1 will come close to your desired output. But what it really looks like you want is a more complicated algorithm (pseudocode below):

if ((img1[x][y] - img2[x][y]) <= threshold)
    output[x][y] = white
else
    output[x][y] = img2[x][y]

You can write this as a very simple for-loop in Octave. You can set the threshold to account for various noise, or set to 0 for exact-match detection. You might need to perform image registration to align the images perfectly. Nimur (talk) 15:52, 4 August 2010 (UTC)

I rarely use MATLAB, but wouldn't you want: if(abs(img1[x][y]-img2[x][y])<=threshold) -- kainaw™ 16:37, 4 August 2010 (UTC)

(Yes, probably would want an absolute-value in there for most cases). Nimur (talk) 22:36, 4 August 2010 (UTC)

This is surprisingly simple in GIMP. Under File, select Open as layers. Select both images and they will open in one image as layers. In the layers dialog, select the top layer. Then, in Mode, select Difference. Done. -- kainaw™ 16:45, 4 August 2010 (UTC)

Great answers so far, thank you. I'd also like to add the following question: What if these photos didn't have the exact same background? What if I stood still taking pictures of a person walking somewhere and wanted to create an animation of just the person? The background would be the same, but it would be slightly shifted and have small differences at the edges from the imprecision of my hands trying to hold the camera in the same place. Can these programs show any tolerance to a situation like this, and only extract the main differences?--el Aprel (^facta-_facienda) 19:08, 4 August 2010 (UTC)

Perhaps image registration that Nimur mentioned might help accomplish this? Is this technique available in free and open software?--el Aprel (^facta-_facienda) 19:13, 4 August 2010 (UTC)

I can't recommend a technique because the ideal technique depends on the image. There are probably hundreds of different techniques for extracting an image from a background. Photoshop, for example, has the magic wand, the quick-selection tool, the extract filter, color-range command, the lasso, the magnetic lasso, channels, the marquis, and so on. I watched two DVDs of video tutorials on masking (essentially the same thing as selecting) in Photoshop -- [2], [3]. So, if you want some hints, I can give them to you, but I'd need to know more about the image you want to mask. You can combine selection techniques (in fact, you should) to achieve better results. Masking an image can take hours. The ease with which you mask depends on how similar the subject is to the background. An ideal situation would be a subject against either a green screen or blue screen. The more contrast, the easier the program will be able to distinguish the subject.--Best Dog Ever (talk) 20:19, 4 August 2010 (UTC)

That's a difficult challenge - not totally impossible, but difficult. Image registration will be a precursor step - aligning the backgrounds. This is accomplished by numerical optimization to find the maximal alignment (cross-correlation) between the images. Then you could either subtract or mask. There are other, much more complicated ways to do this: motion vector estimation, for example, is a critical piece of modern lossy video encoders, like H.264 - and allows an estimation of what pixels have "moved" between two images. Very quickly, these types of image processing tasks cross a line from "basic addition, multiplication, and subtraction of images" to "advanced nonlinear optimization problems in high-dimensional vector spaces", and are open research topics. Some advanced image tools, like PhotoShop or Gimp, may support some best-effort implementations of these kinds of operations. Nimur (talk) 22:36, 4 August 2010 (UTC)

I found a plug-in for GIMP that performs image registration and works quite well (check it out here), but I'm still having some problems with the basic step of getting the differences of two images. When I followed Kainaw's GIMP instructions above for my sample images, I got this image:

It's not exactly what I was looking for, but it's close, and it looks like I might just be missing an extra step or two to get what I need. It seems like the brighter leaves bleed through the image of the molecule, and the bottom-left symbol is a bit transparent. Is there a way to rectify this?--el Aprel (^facta-_facienda) 04:49, 5 August 2010 (UTC)

To remove the bleed (assuming that the bleeded image has constant, not variable, transparency), I'd try to cut out the molecule image onto its own layer, overlay the a copy of the background and set its mode to "difference", and then adjust the transparency until (hopefully) the molecule looks approximately like normal again. If different parts of it have different amounts of transparency, I think the only solution is to do the opacity reduction with the airbrush tool, a steady hand, and some good luck. Adding and removing transparency in parts of an image can be achieved by editing the alpha mask. Paul (Stansifer) 14:04, 5 August 2010 (UTC)

Wait a second; the problem isn't bleed at all. That's what the difference image is supposed to look like. What you want is to (a) turn all non-black areas to white in the difference image, and (b) use the result as a transparency mask in the original image to turn the background transparent. Paul (Stansifer) 14:10, 5 August 2010 (UTC)

Exactly. The process Paul Stansifer just described is equivalent to the pseudo-code I posted earlier: masking, and then selecting image2 data where necessary. This is not something Gimp can do in one step; and it is why I suggested the algorithm in Octave's image-processing kit. You could perform that process in Gimp in several steps: use your result with the black background as an intermediate step; then, chroma-select black areas; then create a mask from that selection; then use that mask to select pixels from image 2). Nimur (talk) 20:09, 5 August 2010 (UTC)

Wow, excellent work, everyone. While I'm not familiar with Octave yet, I will keep Nimur's advice in mind if I need this trick for a large quantity of images. With a small number of images, though, GIMP's being working great! Thank you.--el Aprel (^facta-_facienda) 02:27, 6 August 2010 (UTC)

Whatever happened to pump and dump spam?

Whatever happened to pump and dump spam? A few years ago, my inbox was full of emails trying to get me to buy worthless stocks that the spammer had bought in advance, trying to drive up the price so they could sell at a high price. But I've not seen any of this type of spam for ages. Why is that - did enough people catch on to the scam to make it worthless (I doubt this, seeing that I still get plenty of messages from various Nigerian princes), did the people doing it get arrested, or are spam filters just much better at blocking this sort of thing than they are at catching prescription medicines and fake designer goods? 59.108.42.46 (talk) 11:18, 4 August 2010 (UTC)

I've only had a chance for a brief look, but a couple of anti-virus/anti-malware companies speculated that it was caused by improved spam filters and a regulatory crackdown at the SEC ([4]),([5]). This website, which analyses spam, suggests that the spammers used deliberately malformed image files to bypass optical character recognition in spam filters; it's possible that OCR has improved enough that this technique is no longer effective (personally, I no longer see image spam of any kind - a few years ago, it constituted about 30% of the spam I saw). Apparently, there is also now software which detects trading patterns which indicate pump-and-dump scams ([6]). The drop seems to have begun in late 2007/early 2008, but there was also a hosting provider McColo which was notorious for having spammers as customers, which was closed down in late 2008. This might account for some of the drop; although many of the spammers later set up shop elsewhere, it's possible that they switched to a new business model. I haven't been able to find a definitive answer so far, though --Kateshortforbob _talk 16:03, 4 August 2010 (UTC)

I still get image-spam for generic drugs, so I'm not sure if it is better filtering. However, I'd agree with Gabrielle Glaister about better regulation. The 'stocks' were generally on the non-SEC regulated pink sheets. According to the Caveat Emptor - Buyer_Beware section, any 'stocks' that are spammed are automatically suspended. It would be easy to set up a few hundred email accounts on the major webmail accounts and have them auto-forwarded to central account ran by the market; if a company is listed on several of these accounts then it would be ipso facto evidence that it is being spammed. The same could be done with the major ISPs with their co-operation. Of course, it could be possible that a company P'n'Ds its rivals to give them a bad name, but I've not seen or heard of it. CS Miller (talk) 19:26, 4 August 2010 (UTC)

I think you mean Prima facie evidence. AndrewWTaylor (talk) 09:18, 5 August 2010 (UTC)

Those sorts of scams relied on the patsy's willingness to believe it would be possible to make money buying the stocks. Several years ago, the economy was only going up, and it was really easy to believe that a random stranger could provide you with hot stock tips. The recent financial crash means that hardly anyone views the stock market as a means to a quick buck. It's much harder to believe that investing in stocks will make money, so a pump-and-dump will have a much lower success rate than previously. The pump-and-dump scammers probably are now falling back on "traditional" 419/v1agr4 scams. -- 140.142.20.229 (talk) 21:50, 8 August 2010 (UTC)

raw to png

I need a program that can convert .raw to .png, and back again. Any suggestions? Thanks 82.43.88.151 (talk) 12:54, 4 August 2010 (UTC)

I don't think ".raw" is one format. However, ImageMagick can handle the more common raw formats. See [7]. --Stephan Schulz (talk) 13:08, 4 August 2010 (UTC)

IrfanView and Raw Therapee can also convert many RAW formats easily to a png, though going back to a specific RAW format isn't possible. The obvious question is, "why are you trying to do this"; why just not keep the photo is a RAW format until the very end, when you then convert to the PNG? Buddy431 (talk) 13:45, 4 August 2010 (UTC)

RAW formats are, by their very nature, platform specific, camera-specific, model-specific, sensor-specific, etc. There has been a small effort by some of the major camera vendors to provide consistent output raw files - truthfully, these are not raw data any more (they have been reordered into standard output formats). There seems to be a common misconception: RAW does not merely mean "uncompressed" or "losslessly compressed" - it also means "in the specific pixel, bit, byte, and Bayer pattern order of the particular sensor that generated it". It theoretically also means "with absolutely zero digital processing applied" - but the definition of "absolutely zero" varies from implementation to implementation. This said, once you have images out of the camera, it's better to switch to a standard but losslessly-compressed (like PNG), or uncompressed (like BMP or TIFF) image format if you need to preserve image fidelity. "RAW" files are really a programmer's debugging tool. They have caught on in digital photography because of the side-effect that they are also uncompressed. If you want to convert an image to RAW, you can easily write a program to do so. Nimur (talk) 16:04, 4 August 2010 (UTC)

I can't write a program to do so, I wouldn't know the first thing about writing programs. 82.43.88.151 (talk) 16:08, 4 August 2010 (UTC)

You should probably use a format like TIFF, which will not modify the raw data in any way and will be uncompressed; but has a standard file format, and can be used with a huge variety of existing digital imaging software. Nimur (talk) 22:28, 4 August 2010 (UTC)

Nimur, I don't understand what you're saying here. Of course you could define a TIFF variant that stored the same data as your camera's raw format, but no off-the-shelf software would support those images. Or you could use a commonly supported TIFF variant, but that would not preserve the original data, which I think was the whole purpose of this exercise.

There are basically three or four parts of converting the digital sensor output to a JPEG image: demosaicing, optional denoising, clamping and quantizing to an RGB color cube, and the actual JPEG compression. Of those, the JPEG compression is easily the least lossy step, if you use a high quality setting. The only reason to keep camera-raw images around is to preserve the image's original color gamut and/or to benefit from future improvements in demosaicing and denoising algorithms. You lose all of those benefits (except denoising, I suppose) if you convert to PNG or TIFF, unless you define a specialized version of PNG or TIFF that no standard software will understand.

82.*, what is your ultimate goal? Do you want to losslessly compress your camera-raw images to take up less space? There is a file format called Rawzor designed to do exactly that, but the format seems to be undocumented, without even an open-source decompressor, so I can't in good conscience recommend using it if you care about long-term availability of your images. You could also use Zip or 7-Zip. -- BenRG (talk) 23:16, 4 August 2010 (UTC)

A brief test of compressing a Canon RAW image saved less than 2% compressing with .Zip, so I'd conclude that they're fairly well compressed already. --Phil Holmes (talk) 08:21, 5 August 2010 (UTC)

Not only compressed, but some data is encrypted as well. Look at Digital Photography Review's interview on the subject; and Decoding raw digital photos in Linux, including source-code for reverse-engineered versions of the proprietary lossless compression used by major vendors: dcraw.c, and additional tools for Canon: clean_crw.c. Nimur (talk) 20:16, 5 August 2010 (UTC)

I've been reading a little about Adobe's DNG format. It's a TIFF variant (I should have known), it can probably accommodate all of the metadata from other common camera-raw formats, it supports lossless JPEG (which compresses pretty well), and it's a widely supported open format. However, I don't know what tools can be used to losslessly convert from other camera-raw formats to DNG, and I suspect that many DNG-writing tools will write uncompressed output. Dcraw supports raw/linear TIFF output, but I don't know if it's DNG and it definitely isn't compressed. -- BenRG (talk) 06:41, 6 August 2010 (UTC)

Yes, DNG is actually just a TIFF that stores the camera's capture, exposure, and other meta-data in a "free-form data" region of the TIFF spec. It's commonly referred to as a "RAW", but it is not actually a raw format. A quick web-search will obviate that this format is not well-liked by the camera manufacturers; those that offer it on their camera do so begrudgingly. (This says nothing about the utility or technical merits of DNG - it only illuminates a conflict of business-interest between camera-vendors and software-vendors). Adobe's own DNG Profile FAQ helps explain some technical details of DNG; at first glance, these "portable RAW" files make perfect sense - but only to consumers who plan to use PhotoShop or similar products. As I mentioned above, actual raw image formats from the manufacturer are not very useful to the average photographer. Most photographers who want a "raw" really just mean that they want an uncompressed and unprocessed version of the image. What these photographers do not understand is that an "unprocessed version of the image" is literally nothing more than a bit-stream out of the CMOS or CCD sensor - and unless you like reading ADC control bits, silicon-specific process controllers, and receiving the image in whatever pixel-order was convenient for the manufacturer of the silicon to route into his output pins, actual RAW format is totally useless. And, camera vendors like Canon and Nikon, take pride in their "flawless" final image output that highlights the hardware perfection of their products. They do not want to expose any "intermediate" steps. They see no benefit in encouraging end-users to use additional post-processing; the more image-processing that gets offloaded to 3rd-party PC software, the more commoditized the camera hardware becomes.

Anyway, why is it so hard to get the "un-processed" sensor data in the first place? Shouldn't this just be the "direct" output from the sensor? Well, in a perfect world, yes. A CMOS or CCD sensor is just an array of voltage-inputs to an analog-to-digital converter, and by reading its outputs, you should ideally get a "raster scan" of the pixel values at each point in the image; and you would call that the "RAW" output from the sensor. But this is not at all how an actual consumer-grade CMOS of CCD sensor actually works. This is because of engineering realities related to the manufacture of cost-effective CMOS and CCD sensors. The data has to be compressed, or re-ordered, or patterned, or read out in strips, or every 6th line is read, while a capacitor charges up to re-read the next set of every 6th-line, and so forth - in fact, the sensor might not even be in RGB format, it's much more likely that it uses a YUV colorspace that preserves its Bayer pattern - and then the data is spit out over a proprietary protocol onto the sensor's IO output. These bits/bytes/data are not an "image" yet; a variety of on-camera electronics and software must first process the sensor RAW output and generate (1) an uncompressed image. Then, if desired, software and hardware noise-reduction and image processing may take place, producing (2) a "clean" version of the uncompressed image. Finally, the device may compress this image to (3) a JPEG, and save it in this file format. Most photographers really want something "somewhere between stage (2) and (3)" when they say they want "raw" data from the sensor, without properly defining exactly where between stage (2) and (3) they actually mean. (For example, are noise and dead-pixels desired or undesired in the RAW?)

As you go farther up the "expensive"-ladder, you pass the consumer-grade DSLR, into the professional full-frame sensors, and ultimately into the specialized imager technology for scientific and industrial applications; and you can start expecting reasonable, meaningful data to actually come out of the sensor (i.e., if you just read the bits, you could reconstruct an image from it). But down at the reasonable price-points where mere mortals and consumers operate, there is so much ugly "detail" in terms of getting the optical image out the wire of the sensor, that the result is a very complicated raw data stream. Nimur (talk) 15:35, 6 August 2010 (UTC)

You're making this more complicated than it really is. Camera raw formats store digitized sensor outputs with as little digital postprocessing as possible. That's what DNG stores, so DNG is a camera raw format. There would be no point storing the data in some weird interleaved straight-off-the-hardware bit order, but that's not what camera raw is about and never has been. It's about keeping your postprocessing options open. The color channels from a Bayer sensor are not sRGB or YUV, but something more like LMS or XYZ, but in any case, they are stored uninterpreted and untransformed in a camera-raw file. -- BenRG (talk) 10:16, 7 August 2010 (UTC)

DVD Ripping

My DVD collection is taking up too much space so I am trying to rip my collection to my hard drive. I used Corel DVD Copy 6 to rip my Star Trek The Next Generation DVD's but I found it the be extremely unreliable (crashing frequently etc.). What DVD ripping software would people recommend? I want something reliable, and I want to be able to play the movies on my PC, PS3 and XBox. Thanks 86.45.185.245 (talk) 15:03, 4 August 2010 (UTC)

DVDShrink, DVD Decrypter or AnyDVD for ripping the DVD to disk, then you should probably encode the ripped files into a better format like XviD .avi or x264 .mkv, using HandBrake, MEncoder or AutoGK. HandBrake can directly copy from DVDs, but it only reads from non-copyrighted protected dvds (your star trek ones are most likely protected) 1230049-0012394-C (talk) 15:27, 4 August 2010 (UTC)

I've heard many good things about DVDFab. It lets you back up to your hard drive, another DVD, or can output to specific mobile formats. --Zerozal (talk) 18:35, 4 August 2010 (UTC)

A minor note is that to use DVDFab without paying money is a trial for 30 days.Sir Stupidity (talk) 08:52, 5 August 2010 (UTC)

Apple Travel Adapters

Does anyone know of a maker of travel power adapters (http://store.apple.com/uk/product/MB974ZM/B?mco=MTY3ODQ5OTY) for the Magsafe power supplies for power books and other macs that isn't made by Apple and is therefore not £30 for a few pieces of plastic? I'm sure someone must make the same thing, unbranded with Apple, for like a third of the price?

86.153.255.200 (talk) 16:22, 4 August 2010 (UTC)

The basic set of travel adapters comes with new computers, and if you buy used you should always ask if they are included. The kit is mostly a replacement for those, with (I think) a few funky extras. best places to look for off-brand units are eBay and Amazon (if anyone sells off-brand units, they'll have stores up there). --Ludwigs2 16:41, 4 August 2010 (UTC)

No adapters but the local one came with my recent MacBook Pro (bought in Switzerland) and none came with either of my previous two Apple Laptops bought in the US. You can, however, use ordinary adapters. The bricks themselves are multi-voltage multi-frequency, the adapters just provide the electrical connection to the mains. You can also pull off your local end piece, and use a normal local HiFi power cord. Most people should have some of those lying around. --Stephan Schulz (talk) 19:34, 4 August 2010 (UTC)

E-mail postage system

I have heard and read about a proposal (I'm sure there have been several) about a new e-mail system in which users have to pay a small amount of "postage" (say, 1 cent US) per individual e-mail sent. The objective is to greatly reduce spam, as it now would cost a spammer US$10,000 to send out his 1 million e-mails; and the cost to an individual is low. Editorializing and implementation details aside, does Wikipedia have an article on any such proposal? Has any consortium seriously advocated such a proposal yet? Comet Tuttle (talk) 18:02, 4 August 2010 (UTC)

See Internet taxes. It has been proposed to thwart spam. IN the article, you will also see that it has been suggested as a way to move money from the (white) United States to the (black) Africa in an attempt to make up for the racial divide of the global economy. -- kainaw™ 18:13, 4 August 2010 (UTC)

That part of the article is, incidentally, unsourced. (The apparent source says nothing about Internet taxes whatsoever.) --Mr.98 (talk) 22:53, 4 August 2010 (UTC)

The UN referred to it as a "bit tax", not "email tax" or "internet tax". They defined it as "a very small tax on the amount of data sent through the Internet." It appears that that the reference to the UNDP report was replaced with just "Id". -- kainaw™ 00:24, 5 August 2010 (UTC)

See also Hashcash for a non-payment scheme. All these schemes have drawbacks, which is why they're not in use. --Sean 19:07, 4 August 2010 (UTC)

Attention economy#E-mail spam covers something similar. However, the extraordinary success and utility of email basically preclude any non-backward compatible changes. Also, for many people, Bayesian spam filtering has so effective that spam is basically a non-problem. Paul (Stansifer) 19:13, 4 August 2010 (UTC)

Thanks for the links. Hashcash is awesome. Comet Tuttle (talk) 20:55, 4 August 2010 (UTC)

Button on taskbar disappeared

I was at a library with Firefox, but this happens with Explorer too. If you go here then you see a list of open programs, each one a "task in progress". I did this in order to beable to describe what I was doing: that has nothing to do with my specific situation. Yesterday each one of the buttons represented some web site. On one of them, a popup ad showed up despite the popup blocker, and the button for the site I was on disppeared, replaced by the one for the popup ad. I never saw the site again (except by going to it all over again) until I clicked on the red X for all the rest when I was finished.

I've had the buttons rearrange their order, too, when one of the sites was acting up. What's going on?Vchimpanzee · talk · contributions · 18:38, 4 August 2010 (UTC)

The taskbar is a list of open windows, not open programs. A popup window can close the window that it popped up from. As for order, you can customize the order to be alphabetical, it order opened, related to screen position, etc... -- kainaw™ 18:40, 4 August 2010 (UTC)

I'm just telling you what the other site said. But the window wasn't closed. It was there, but I couldn't get to it without closing all the others. And the buttons got rearranged without my consent.Vchimpanzee · talk · contributions · 18:50, 4 August 2010 (UTC)

Are you positive that it didn't go into "grouping" mode where multiple windows of the same program are grouped into a single button? -- kainaw™ 18:53, 4 August 2010 (UTC)

They did not.Vchimpanzee · talk · contributions · 19:12, 4 August 2010 (UTC)

This does not answer the "why", but are you aware of Alt-Tab? If you ever "lose" a window, hold down the Alt key and hit Tab a few times. This will let you get to any existing window, whether hidden or not. Comet Tuttle (talk) 20:37, 4 August 2010 (UTC)

A hint about the "how": this sometimes happens on IE8 if the circle keeps rotating and rotating. The sort of thing that could shut the Internet down, for example, though that rarely happens on my computer.Vchimpanzee · talk · contributions · 21:58, 5 August 2010 (UTC)

It happened to me again today. Alt-Tab worked.Vchimpanzee · talk · contributions · 17:23, 9 August 2010 (UTC)

Google Wave killed, any Wave like replacements?

Google today announced they are killing Google Wave [8]. It's an amazing tool I use to collaborate with several remote freelancers all in different locations. We can't live without it (once we finally wrapped our minds around it). Are there any commercial products or services that offer anything similar? The main feature is persistent, time re-windable shared group chat/email/discussion/task tracking/meeting scheduler/to do list/brainstorming with live typing? --69.148.250.94 (talk) 23:44, 4 August 2010 (UTC)

You may have noticed articles noting that Google has already open sourced "much of the code". I haven't seen news of any replacement site yet. Comet Tuttle (talk) 00:19, 5 August 2010 (UTC)

Yeah, I noticed that. But I have intention of compiling code myself. I just want a sevice to log into! :) --69.148.250.94 (talk) 04:58, 6 August 2010 (UTC)

You may not currently be able to find all of Google Wave's functionality in a single tool, but you might like to check out this recently reposted Lifehacker top ten of web based collaboration tools [9]. Equisetum (talk | email | contributions) 11:56, 5 August 2010 (UTC)

You can host your own servers for MediaWiki, Jabber, and all of the other constituent parts of a collaborative workflow (including Google Wave, when the source becomes available). This cancellation of Google Wave is exactly the reason why it's so frightening how much of our world's e-Commerce has been implicitly subject to vendor lock-in. When Google decides that GMail is not profitable, what will all the GMail users use? They will have to re-learn how to actually use email. This means: running an email server, and running an email client. If you want a "convenient web-based interface," that is fine; you can set that up on your own server as well. But the status-quo right now seems to be "assume somebody else, like Google, will do all the set-up for me, for zero charge." It's 2010. Every single part of a modern IT infrastructure, from the router to the enterprise-scale server to the Web 2.0 Application layer, can be run on cheap commodity hardware using free and open-source software. Do not allow Google, or IBM, or Apple, or any other company to hold your business hostage to their whims, just because access to their server-hosted tools is "convenient." Nimur (talk) 15:51, 6 August 2010 (UTC)