Wikipedia:Reference desk/Archives/Computing/2011 January 14

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.

January 14

Google docs arrow keys

Hi. I use google docs and by and large it's great. But sometimes the arrow keys on my keyboard don't work. This happens on linux and mac, always using the latest version of firefox. Does anyone else experience this? It's very annoying because I have to use the mouse to navigate and this takes my fingers away from the keyboard. Does anyone have any advice? Robinh (talk) 10:13, 14 January 2011 (UTC)

Since nobody's answered, a wild guess: Are you using a laptop with a touchpad as well as a mouse, and is it just the arrow keys that cease to work when this happens, or does the keyboard stop working? If so, what's happening might be that you're accidentally touching the touchpad when the mouse pointer is outside your document. That will cause your document to lose keyboard focus, and you'll need to use the mouse to give keyboard focus back to your document. If this is the case, a solution would be to deactivate the touchpad, or to reduce its sensitivity. --NorwegianBlue ^talk 10:05, 15 January 2011 (UTC)

Hi. Nope, the keyboard works fine but the arrow keys do not apparently do anything. It's very annoying (but only a problem about half the time I use googledocs). Cheers, Robinh (talk) 09:10, 16 January 2011 (UTC)

USB media - hard disk or not?

On a recent Linux distribution, which commandline tools can tell me whether a USB device is an external hard disk or another (storage) device like a flash memory stick? I'm writing a script that is supposed to write data to an external hard disk (connected via USB), but I want to avoid accidentally overwriting a flash memory stick that the user might have left plugged in. -- 78.43.71.225 (talk) 11:02, 14 January 2011 (UTC)

You might be able to examine /sys/class/block/<devicename>/device/model to detect a flash disk. (where <devicename> is something like sda, sdb, etc.). -- JSBillings 16:28, 14 January 2011 (UTC)

I found no /sys/class/block/<devicename>/device/model, only a /sys/block/<devicename>/device/model, which contains part of the manufacturer name for each device, it seems. I did find /sys/block/<devicename>/removable, which contains "1" for a flash memory stick, and "0" for both the first hard disk as well as the one connected via USB. I'm not sure that this is a reliable way of detecting a hard disk, though. Any further suggestions? -- 188.105.132.219 (talk) 16:56, 14 January 2011 (UTC)

IIRC, MS-Windows 7 uses the device's sustained read speed to detect if it is flash drive, you might need to do something similar. CS Miller (talk) 17:17, 14 January 2011 (UTC)

Use smartctl? I don't think flash drives and memory sticks have S.M.A.R.T. information, while I think hard drives and SSDs do. 118.96.156.2 (talk) 17:25, 14 January 2011 (UTC)

Hard drives and SSDs do have S.M.A.R.T. information, but it seems not every USB controller is able to forward this info to the host, so this isn't reliable, either. :-( -- 78.43.71.225 (talk) 17:38, 14 January 2011 (UTC)

smartctl, like hdparm -I works on ATA/ATAPI devices (that is, devices on the ATA and workalike busses). In my limited experience (that is, checking a couple of Ubuntu systems today, trying to find a decent answer to this question) suggests that the Linux usb mass storage device driver stack doesn't implement the requisite ioctls (even for an ATA drive in a USB enclosure), so neither seems like a viable option. -- Finlay McWalter ☻ Talk 17:39, 14 January 2011 (UTC)

Do you have "usb-devices" installed? When I run that, I see info under "Product=" that identifies what is plugged in. I can identify USB storage devices because they have "Driver=usb-storage". I tested a thumbdrive and it came up "Product=LEXR PLUG DRIVE". It is possible that "Cls=08" may be a thumbdrive storage class, but I don't have a usb harddrive to compare with. -- kainaw™ 17:39, 14 January 2011 (UTC)

usb-devices is not a supported command on my install of Ubuntu Hardy Heron. :-( According to packages.ubuntu.com, it is also not part of any package belonging to Hardy. :-( -- 78.43.71.225 (talk) 17:53, 14 January 2011 (UTC)

I don't think that's sufficient. Below is a trace of usb-devices on a machine with a SanDisk Cruiser USB flash disk and a LaCie USB disk enclosure:

T:  Bus=01 Lev=01 Prnt=01 Port=04 Cnt=02 Dev#= 40 Spd=480 MxCh= 0
D:  Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs=  1
P:  Vendor=059f ProdID=1019 Rev=00.01
S:  Manufacturer=LaCie
S:  Product=LaCie Desktop Hard Drive
S:  SerialNumber=00E0010080086
C:  #Ifs= 1 Cfg#= 1 Atr=c0 MxPwr=2mA
I:  If#= 0 Alt= 0 #EPs= 2 Cls=08(stor.) Sub=06 Prot=50 Driver=usb-storage

T:  Bus=01 Lev=01 Prnt=01 Port=06 Cnt=03 Dev#= 39 Spd=480 MxCh= 0
D:  Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs=  1
P:  Vendor=0781 ProdID=5151 Rev=02.00
S:  Manufacturer=SanDisk
S:  Product=Cruzer Micro
S:  SerialNumber=2400121D52C0F953
C:  #Ifs= 1 Cfg#= 1 Atr=80 MxPwr=200mA
I:  If#= 0 Alt= 0 #EPs= 2 Cls=08(stor.) Sub=06 Prot=50 Driver=usb-storage

I don't know what the Atr=80 field is, but other than that only the MxPwr=200mA field looks discriminatory. -- Finlay McWalter ☻ Talk 17:52, 14 January 2011 (UTC)

Here's a hack: lsusb -v, and for each device find its MaxPower setting. For me, a genuine flash external disk has a MaxPower of 2mA, whereas a "real" disk in a USB enclosure has a MaxPower of 200mA. Other than that, I don't know of a way to determine the actual media of a linux block device (bar a vendor/deviceID lookup using the data JSBillings' post retrieves, which is a labourious business). -- Finlay McWalter ☻ Talk 17:46, 14 January 2011 (UTC)

Sadly, I'm getting just the opposite result here: the stick draws 200mA, the hard disk only 2mA (it has an external power supply) -- 78.43.71.225 (talk) 17:56, 14 January 2011 (UTC)

Sorry, I mixed the two up - if you look at my trace above, I get the same as you. The SanDisk is the flash stick. -- Finlay McWalter ☻ Talk 17:59, 14 January 2011 (UTC)

Okay, now I will have to look for a USB-powered hard disk to check how that one shows up. -- 78.43.71.225 (talk) 18:08, 14 January 2011 (UTC)

The fly in the ointment will be bus-powered usb disk enclosures (the slim kind that take laptop hard disks). -- Finlay McWalter ☻ Talk 18:13, 14 January 2011 (UTC)

Drat, so we're back...

... to square one, I guess. -- 78.43.71.225 (talk) 18:22, 14 January 2011 (UTC)

On a more fundamental level, what is the difference between a hard disk and a flash stick? Capacity? Moving parts? Nowadays there are solid state hard drives - no moving parts, used like hard disks. mp3 players can be plugged into USB and look like memory devices - some have rotating disks, some don't. A device might accidentally give a hint about its internal mechanics, or it might not. Should you really care?

If you really need to make a guess, how about looking at the device's maximum storage capacity, or available free storage, and deciding that anything over X GB is a "disk".

Or how about this: on disks that are allowed backup(whatever) devices, manually create a file called "this-is-a-backup-disk-please-write-backups-here". Your script will check for that file and if it is there, the script will do its thing. Random memory sticks won't have the file and your script will abort. 88.112.59.31 (talk) 20:34, 14 January 2011 (UTC)

Sadly, these approaches don't work, or at least not for long. I'm trying to build a box where the user can pull copies of a hard disk image onto "the real thing". Some of these disks are SSDs for embedded systems and have capacities of 60 or 64 Gigabytes - and with USB flash memory becoming larger and larger, using the maximum capacity of the medium as an indicator won't work (at least not for long).

The reasons why I need to be able to tell a flash memory stick apart from a SSD or moving-parts hard drive are:

• Sometimes, data (new images) is transferred to this box via flash memory stick. Overwriting this stick, just because a user forgot to remove it in time, would be anything but user-friendly.
• The SSDs or hard disks will be plugged in via SATA interface in their destination systems (in other words, the USB enclosures the users bring along are only temporary casings) - a flash memory stick doesn't have a SATA interface, plus the machines that still use moving-parts hard disks write to the disk pretty often, so flash memory will wear out. (I've seen two new high-quality USB sticks die in a similar application in the last 6 months. I know that these days, this shouldn't happen any more, but I guess these claims only refer to home use - heavy-duty writes like logging or database access, even today, still do kill flash memory sticks, I have seen it happen.)

Since this box will be available to a larger group of users, I have no control over the USB sticks or hard drive enclosures they will bring along, so writing a "it is (not) safe to overwrite this disk" file on the media won't work.

-- 188.105.132.219 (talk) 21:52, 14 January 2011 (UTC)

Why not present a user interface from your script to let the user pick the correct mode of operation? 118.96.154.64 (talk) 23:23, 14 January 2011 (UTC)

(EC with above) An obvious question that comes to mind that I think others may have been thinking of: Are you sure no one is ever going to connect what you consider a HD, that they don't want to be overwritten, to the computer? If you aren't even if you can differentiate the 2, it's not going to ensure you don't overwrite stuff people don't want you overwriting.

Is it possible you're approaching this from the wrong way? I kind of skimmed thru the discussion so forgive me if I missed it but is there some reason you can't ensure the HDs you want to use are identified as such. Whether a file, volume name or simply requiring the disks are cleared before they are brought to be written i.e. just don't overwrite anything.

If this doesn't work, is it possible to just ask the user? Use double prompts perhaps, with clear phrasing, perhaps even requiring they type YES or something (whatever you think is necessary).

The other alternative is probably physical. Is it possible to designate one set of USB ports or one port for bringing data? And one set or one port for connecting the backup devices. Make sure these are properly seperated and perhaps put the ports for backup devices behind a hinged door or something with a clear label indicating anything connected to them may be overwritten. Then design your script so it only writes to disks connected to these ports.

In both cases some people may still be clueless or careless so wind up overwriting what they don't want overwritten but ultimately there's only so much you can do to stop people doing stupid things.

Nil Einne (talk) 23:31, 14 January 2011 (UTC)

The "Store image" function should probably be restricted or removed anyway (either password-protected or disabled – images should be stored using special root access). The case is, what if someone were to present an incorrect, corrupt, malware-infected, etc. image to the system? Will that image be restored into subsequent disks presented to the system for restoration? 118.96.154.64 (talk) 23:50, 14 January 2011 (UTC)

All the users of this system belong to Field CircusService, and the machines they're imaging a) don't run Windows and b) are not accessible to end-users, plus the image they're storing is an individual image for each machine, so even if something does go belly-up, it cannot spread to other machines. This makes me believe that the risk in my particular case is rather low. They know that if they mess with the system, they end up creating more work for themselves (reimaging is faster than reinstalling and reloading the config, but they do have access to the full installation procedure as well, so we're not creating a deadlock situation here). I just want to keep them from accidentally shooting themselves in the foot by overwriting a forgotten USB stick. -- 78.43.71.225 (talk) 13:55, 15 January 2011 (UTC)

Thank you for all of your input. Since it seems nigh impossible to detect the media type via USB, I will probably be using a SATA drive bay for the disks. Of course, this means the users will have to power-cycle the computer before swapping hard drives, as the standard SATA controller doesn't support hot-swapping, but since the only media connected via USB will be a memory stick, all I have to do is check for connected USB devices and complain/abort if there is one. Since all USB devices automatically get mounted under /media/usbn, something like

umount /media/usb*
lsmod | grep usb_storage && rmmod usb-storage

and checking its return code (if the module could not be unloaded, it means that a usb storage device is still in use) should do.

After that,

fdisk -l|grep '^Disk /dev'|tail -1|awk '{ print $2 }'|tr -d ':'

should tell me the name of the last disk available to the system, and if it doesn't equal a disk listed in

df|grep '^/dev/'|awk '{ print $1 }'|tr -d '[0-9]'|sort -u

I'm good to go.

Of course, if anyone knows a simpler way of solving the issue, feel free to comment. :-) -- 78.43.71.225 (talk) 13:55, 15 January 2011 (UTC)

Full Screen Games Minimizing On Start

Hey all,

I've got an intermittent problem that's been doing my head in for quite a while. I've looked around on the internet for people with the same issue, and it seems there are many of us, but surprisingly few people (i.e. zero) with the answer. My problem is that when I launch a full-screen game, it minimizes almost immediately, and no matter how many times I click the tab in the task bar, it won't stop minimizing. A restart of Windows generally sorts the problem out. Now, I will anticipate a few of the possible answers I may get:

• Do you have an IM running, as it sounds like something is demanding attention? Yes and no. I have tried playing games with and without an IM running, and I still get the same problem.

• Do you have the latest version of your video card's driver? My driver is updated automatically whenever an update is available, so, yes.

• Have you tried running games in windowed mode? Yes, I have, and this solves the problem. Unfortunately, not all games have a windowed mode (and many of the ones which have won't stay on screen long enough for me to change the settings).

My computer is a G60 Netbook from HP. This problem has only been occurring since October of last year. I recently reinstalled Windows (due to an unrelated issue), and the problem is still happening. Like I say, after restarting Windows, the problem is usually fixed, at least for a few games, before it starts happening again.

Any help would be appreciated. --KägeTorä - (影虎) (TALK) 19:33, 14 January 2011 (UTC)

I have noticed a similar annoying symptom for the last few weeks when playing Civilization IV recently (which I bought on Steam) on my Windows Vista machine. I do task-switch often via alt-tab, and I do play Civ IV as a full-screen game. When I alt-tab back to Civ IV, it will sometimes switch right back to the Windows desktop. I have found that if I alt-tab to Civ IV and quickly click the left mouse button, it won't task switch back to the desktop, and I have assumed there's some errant bit of code somewhere which believes that I've clicked the Windows desktop and it switches me over there; and the Civ IV click stops this behavior. I'd be interested to know if this helps your symptom, too - Comet Tuttle (talk) 23:06, 14 January 2011 (UTC)

It doesn't help - I often click or press escape after re-maximising the game, and I am not sure if it is helping because the game will sometimes minimise itself again and sometimes not. It is pretty frustrating, and seems to happen more with older games (pre-2008), although it sometimes happens with the newer ones too, especially if I've tried to play an older game first and this has happened. I forgot to say I am on Vista. --KägeTorä - (影虎) (TALK) 02:38, 15 January 2011 (UTC)

I have no references for this, but I just wanted to throw out an idea: Could it be that the games are trying to use resolutions not supported by your monitor? Google suggests that the G60 has a resolution of 1366x768 and a lot of older games especially wouldn't support widescreen. Normally I'd expect the monitor to resize accordingly, but maybe that's the problem and it doesn't/can't. Try going into your display settings and changing the resolution to 1024x768 (assuming you can - things will either look stretched or will have black edges) and then try your game and see if it's made any difference?  ZX81  ^talk 15:22, 15 January 2011 (UTC)

Well, the thing is, once one game starts this pallavah, they all start doing it, no matter what the resolution is. A restart will generally sort the problem out. Defragging sometimes seems to help, too. It's only been happening since October or so of last year, and before that I'd never experienced anything like it. Not only that, I can play the same games on my Ubuntu partition (using Wine) and I don't get the problem even once. --KägeTorä - (影虎) (TALK) 20:09, 15 January 2011 (UTC)

Ubuntu

Between Ubuntu variants, is there really any significant difference? And are there any major flaws or advantages of any specific variants (especially version 10.10 "Maverick Meerkat")? Thanks, --T H F S W (T · C · E) 21:31, 14 January 2011 (UTC)

Regular Ubuntu uses GNOME as its desktop environment (with a bunch of associated helper apps); Kubuntu uses KDE instead. Some folks think GNOME is easier to use, or perhaps a bit more like Microsoft Windows, and find KDE a bit techie; other folks think GNOME is inefficient and unconfigurable. In practice almost all applications will run in either environment (as Kubuntu still installs GNOME's underlying toolkit, GTK, and Ubuntu still installs KDE's underlying stuff, including QT). For a full-featured desktop, it really doesn't matter which you use, and if you don't know (or don't care) then just use regular Ubuntu. Xbuntu eschews both for a lighter-weight environment; it's more suitable for older, cheaper, or otherwise more resource constrained environments. Canonical produces a server variant of Ubuntu too; that's really mostly just the desktop one with the graphical stuff removed (and with some changes to configuration to best suit a server environment). But the desktop variant can install all the server stuff, and the server variant can install the desktop stuff, so they're not that far apart. Only use the server variant if you're setting up a genuine dedicated server machine. Beyond that there's stuff for specific purposes (produced by other development teams) like Edubuntu, which are really just regular Ubuntu with a tailored set of applications that come by default. As to versions, there's almost no reason to ever install anything but the latest one. Only people with very specific requirements to deliberately run older software (mostly to keep uniformity and predictability in an enterprise environment) should ever install the older versions. -- Finlay McWalter ☻ Talk 22:28, 14 January 2011 (UTC)

I'd stick with LTS (long term support) releases, if I used Ubuntu. Latest is 10.04, and it will be supported longer than 10.10. If you know what you're doing it won't matter, but it's slightly more insulation. I'm with Finlay mostly, but think Windows users will tend to prefer KDE. Some of the derivative distros are officially supported by Canonical, some are not, FYI. ¦ Reisio (talk) 22:34, 14 January 2011 (UTC)

It'd recommend LTS for an enterprise with hundreds of machines doing mission-critical work, where they didn't want to take the risk (and cost) of upgrading machines to a newer version during their design lifespan. LTS just means they'll continue to get patches on that install for longer. For everyone else, and I'd say for any new Linux user, the newest release is best; Canonical (and the Debian teams that do much of the work) have done a great job in making the last few Ubuntu versions (for at least two or three years) very stable. -- Finlay McWalter ☻ Talk 22:48, 14 January 2011 (UTC)

The idea of using Ubuntu for something mission critical is frankly laughable. I agree, however, that it will only make a difference to the truly uninitiated. ¦ Reisio (talk) 03:26, 15 January 2011 (UTC)

Marnanel (talk) 03:37, 15 January 2011 (UTC)

Ubuntu is a derivative of Debian. Debian is for home or personal use, not mission critical use. The main issue here is that package maintainers are allowed to alter the code for the package as they see fit without sending it back upstream to a main oversight group. For example, the openssl package maintainer removed a random number seed in the openssl package because he didn't think it was necessary. The change was not part of oversight. For over two years, nobody noticed (except hackers). Then, it was discovered that all of the ssl keys in Debian were weak and guessable. All Debian users had to regenerate them. An alternative is something like RedHat. All changes require extensive oversight because it is designed for a mission critical environment. For home/personal use, it isn't very nice because that is not the target market. It all boils down to two markets: Get updates slow but highly secure vs. Get updates right away and find the bugs later. -- kainaw™ 03:58, 15 January 2011 (UTC)

iPod touch Backup

I am running Mac OS X 10.5.8 and I have a 4th generation iPod touch running iOS 4.2.1. I have app data on it that I would like to save. I'm confused about whether or not backing up my iPod through iTunes saves app data because I've read that only certain apps allow for app data to backed up, or it doesn't backup app data, etc. I have James Cameron's Avatar: The Game on my iPod touch and I've gotten pretty far in it and I want sync a couple of recent previous firmwares and iOS 4.3 beta on my iPod touch so I can get their SHSH blobs with Tiny Umbrella. I've looked at the size of the backup created by iTunes and it is only 44 MB in size which seems way too small to have backed up all app data, contacts, etc.

So, I want to know if there is any program out there that I can use (compatible with Mac OS X) that can backup everything on my iPod touch, in a manner similar or the same as flashing the memory (when I say flash I do not mean Adobe Flash or wiping it clean) without a jailbreak? I think it might be theoretically possible by just ignoring file system privileges in the same manner some builds of xpdf can view a PDF document without respecting the DRM. --Melab±1 ☎ 23:22, 14 January 2011 (UTC)

Help would be appreciated. This is time-sensitive. --Melab±1 ☎ 00:37, 15 January 2011 (UTC)

Keyboard shortcut for the back button

In Internet Explorer, pressing backspace does the same thing as clicking the back button, and over many years of using windows I have got very used to this shortcut. At work I use Ubuntu and I find with Firefox the keyboard shortcut is Alt+Left-arrow which requires two hands and is pretty inconvenient (only the left-Alt works and the keys are too far apart to do this with one hand). Several times a day I automatically press the backspace and ponder for a few seconds how much longer it is going to take to go back to the previous page, before I realise I have done the wrong shortcut yet again. Until recently, I though Firefox didn't support changing the shortcut until I came across one that did. Unfortunately, I was in the middle of a training course I didn't have time to check out the settings. So, how do I change the keyboard shortcut for the back button, so it works like Internet Explorer? Astronaut (talk) 22:02, 14 January 2011 (UTC)

Open about:config and change browser.backspace_action to 0. Algebraist 22:10, 14 January 2011 (UTC)

Thanks. I'm not at work now, but I'll check it out. However, the about:config sounds interesting... where can I find out what else I can do with it? Astronaut (talk) 22:25, 14 January 2011 (UTC)

ALT+lArrow and backspace are actually redundant shortcuts, FYI, you don't have to pick just one. IME Linux builds tend to have the backspace shortcut enabled by default, which actually always bothered me personally :P while the Windows builds do not, IIRC. More shortcut configuration: http://forums.mozillazine.org/viewtopic.php?t=72994 ¦ Reisio (talk) 22:30, 14 January 2011 (UTC)

http://kb.mozillazine.org/About:config_entries for starters. ¦ Reisio (talk) 22:27, 14 January 2011 (UTC)

Resolved

Works like a dream. Thank you. Astronaut (talk) 22:09, 17 January 2011 (UTC)

Hotmail has started delivering spam not addressed to me

In the past few days I've been getting a few spam emails in my hotmail inbox which are not addressed to me, but to possibly similar addresses or sometimes addresses which are not at all similar.

Does anyone know why this is happening please? I installed Thunderbird recently, cannot remember if I did this before or after getting the spam, and it could be just coincidence. Thanks 92.15.25.173 (talk) 22:10, 14 January 2011 (UTC)

Just sounds like regular spam to me. Your address is probably on a spam list somewhere, either from a site you gave your address to, or from one of your friend's compromised computers. Astronaut (talk) 22:23, 14 January 2011 (UTC)

It is just regular spam. I get messages (to my hotmail) addressed to someone else quite often, and usually from contacts with hotmail addresses. --KägeTorä - (影虎) (TALK) 22:58, 14 January 2011 (UTC)

I do not understand why Hotmail is sending me things that are not addressed to me. 92.24.180.142 (talk) 23:30, 14 January 2011 (UTC)

Email contains two different parts: the human-readable part, and the machine-readable part. Most email systems do not require that the two actually match. So, it's entirely possible to send mail to user@server.com and place a human-readable "To: Mr. User (joe@otherserver.com)" as metadata. You can read more information in our article, Email#Message header. Specifically read: "Note that the To: field is not necessarily related to the addresses to which the message is delivered. The actual delivery list is supplied separately to the transport protocol, SMTP, which may or may not originally have been extracted from the header content." If this explanation is too technical, you can sort of think of it with this analogy: suppose a spammer sent you a paper letter via regular postal mail. On the outside envelope, he has to put your actual address (house and street number) so the postal service delivers it. But once you open the envelope, the paper letter on inside can be addressed to anybody. It can say, "Dear Aunt Sue," or "Dear Bob"... it doesn't matter. The mail has already been delivered based on the outer envelope address. Nimur (talk) 00:25, 15 January 2011 (UTC)

Many internet security experts consider the original EMAIL implementation to be fatally flawed for this reason (among numerous other problems). There is essentially no enforcement that senders authenticate themselves (they don't have to prove that they are who they claim they are); and there is essentially no requirement that metadata matches reality. See E-mail spoofing for more information. In a sense, "social network" websites sidestep this issue by consolidating every authentication in the entire world into a single login-system, which is operated by one single social network corporation's servers. Email, on the other hand, is distributed, so there is no central authority to verify credentials. Nimur (talk) 00:34, 15 January 2011 (UTC)

Note of course even if you are note in the To:, you may be in the CC: (which is visible) or BCC: (which should normally be removed) Nil Einne (talk) 10:51, 15 January 2011 (UTC)

I'm shocked that as a very very very very....very rudimentary anti-spam device that the Hotmail computers, and the well paid execs there, do not make sure that the emails are delivered only to the human readable address, or at least that emails where the human readable address contradicts the computer readable part are rejected. At the very least there ought to be an option to refuse emails with these deceptive addresses. Is there one?

How can I complain to Hotmail/Microsoft about their lazy sleeping execs who should have already fixed this long ago? You cannot say that the computers cannot read the human-readable email address - that is no excuse. Even the domain name part of the email address has been different. Its really really dumb to deliver something addressed to Pamela Anderson, Hollywood, USA, in the letter box for The Pope, Vatican, Italy. Duh! 92.15.27.244 (talk) 11:10, 15 January 2011 (UTC)

Uh who said computers can't read it? It's just that they don't because they're not supposed to and in fact doing do would likely break some things like Blind carbon copy and most mailing lists. As Nimur has said, you're letter analogy doesn't really work. If I write a letter to Pamela but address it to The Pope on the envelope, it's not the postal service's fault that The Pope receives my letter. Note that even if server did require To:, this would do little to discourage spam since the spammers can easily change their practices here, it's not likely to be much of a hurdle at all. (Things like requiring authentication would likely help more.) P.S. You can of course easily set up a filter to delete, copy to a spam folder or whatever all messages which don't have you in the To: field. You might want to at least consider looking in both the To: and Cc: but it's up to you. Nil Einne (talk) 12:16, 15 January 2011 (UTC)

Who says "the're not supposed to" look at people's emails? Googlemail got bad press about doing this routinely, but just having a computer check that the mail is being delivered to its ostensible address is nothing that anyone, apart from the spammongers, could object to. And its astounding that it hasnt been done from day one. 92.15.24.16 (talk) 14:23, 15 January 2011 (UTC)

Thanks, will look into the filter you described, but it should be the default setting.

Blind carbon copies ought at least to be signalled as such in the human readable part of the email, with an option to reject them. I suppose BCCs must be how spammongers send a lot of their spam out, otherwise it would have the correct address shown.

I'm shocked Hotmail seem to shrug their shoulders at spam and say "Stuff happens" and go back to sleep again. Emails with the wrong address are really shouting out loud with a bullhorn and waving a large flag saying "Hey, I'm spam!" yet nothing is done about it. 92.15.24.16 (talk) 12:29, 15 January 2011 (UTC)

I don't think you can do much, even with the filters. The filters only let you ignore mail from a specified address in the from field, which can also be easily faked. So, you might appear to recieve mail from panderson@hollywood.com and pope@vatican.org, but when you examine the headers and trace back the IP addresses, they actually come from the same place in China (for example). And when it comes to complaining to Hotmail, I wouldn't hold your breath. In my experience, they seem curiously reluctant to get a real person to answer users' complaints and instead rely on a computer sticking different names on form letters. If the spam really bothers you, I suggest you ditch that account and set up new accounts for different purposes (one for family & friends, one for shopping, one for work/school, one for riskier sites such as forums, etc.) Astronaut (talk) 14:35, 15 January 2011 (UTC)

Surely it must be elementary programming to check that ostensible address matches the delivery address. Why don't they pull their finger out and do it? 92.15.24.16 (talk) 14:44, 15 January 2011 (UTC)

Because as pointed out already, that's how Blind carbon copy works and to "fix" this, would break that (and people do actually use BCC for valid reasons). So I can't see them wanting to make a change that would break a feature that a lot of their members probably use/rely on.  ZX81  ^talk 15:17, 15 January 2011 (UTC)

I really cannot see why people cannot understand that I want to reject all emails that are not ostensibly addressed to me. BCC is irrelevant to this: all the software has to do is see if the delivery address is within the ostensible address string, and if not bounce it back - so so simple. I am on a few mailing lists etc and the emails are always properly addressed to me, so even if the use BCC they still have no problems ostensibly addressing it to me. Its a very simple and straightforward request. Its easy to understand - if the ostensible address is not mine, then I don't want it. Whats so difficult to understand about that? I've never had any valid email that's not been ostensibly addressed to my address. I cannot imaginine even the most dim-witted real postman having difficulty with the concept that you only put mail into the letterbox it is ostensibly addressed to. 92.15.24.16 (talk) 18:37, 15 January 2011 (UTC)

Well BCCs don't necessarily have your email address in the header (that's why they're called BCCs...) and as such that's why Hotmail can't just block them without blocking all BCCs. Some emails do contain it, it all depends on the sending server, but as far as I know there's no requirement for the sending server to include it. Since Hotmail is used by millions of people and they have to support BCC I'm afraid your only solution would be to set up your own mailserver and tell it not to accept BCC's, but then if someone tries to send you a BCC you wouldn't get it, nor would you know they even sent it...  ZX81  ^talk 18:54, 15 January 2011 (UTC)

Our article mentions some of the problems with the way what's supposed to be done with Bcc is defined. Nil Einne (talk) 20:18, 15 January 2011 (UTC)

Also, if Hotmail were to make a change like that, it would stop spam for a couple days, and then the spammers would just change what address they send things to. The reason why Bayesian spam filtering works well is that it can react quickly to new spam techniques, and that it also identifies characteristics that make a message likely to be non-spam for a particular user, which spammers can't fake without knowing the recipient's specific interests. Paul (Stansifer) 15:46, 15 January 2011 (UTC)

The point is not sending it to my address. 92.15.24.16 (talk) 18:37, 15 January 2011 (UTC)

They are sending it to your address or else it wouldn't get to you. This should be obvious.

What you're complaining about is that the "To" field does not match your email address. As has been explained above, if an email is sent to you with your address in the "Blind Carbon Copy" field then your email address will not appear on the email. (That's what the "blind" means. A regular carbon copy will show the address of the CC recipient.) There are entirely legitimate uses of this. (Bob want's Joe to have proof that Bob sent an email to Sue, but he doesn't want Sue to know that Joe is involved. So he sends the email to Sue with Joe's address in the BCC. Joe's email address will not appear on the email, but he will receive it.) APL (talk) 08:43, 16 January 2011 (UTC)

Um there are standards (RFC) that email servers are generally supposed to follow and while Microsoft and other companies do sometimes ignore them, in this cases there's no point since as me and others have pointed out, this will cause far more problems for the many people (including mailing lists) who do use BCC correctly, but will cause very few problems for spammers since they can easily change their software and don't give a crap about revealing other people's email addresses or the potential confusion about whether something was sent to a discussion list or sent to a person privately. There have been plenty of proposals which will semi-break the way e-mail works (e.g. hashcash) in an attempt to thwart spam but to be blunt, your one seems one of the more pointless ones that isn't going to achieve anything other then cause problems for those who aren't spamming (probably why you don't tend to here it from those who actually manage these sort of things). Nil Einne (talk) 20:00, 15 January 2011 (UTC)

As for the filters, I don't use hotmail (or webmail products in general) so was just speaking in general terms and apologise for not making that clear. Definitely few decent email programs would have problems setting up filters which would basically discard any emails which don't have your address in the to and cc lines (although with gmail for example where your address can actually take numerous forms it may not be so easy if possible to support all those forms depending on the boolean operators available). I'm quite confident the old Eudora has been able to do this for a long time (probably very 5 or earlier) and from a quick look at Thunderbird it should be no problem there either. If you filters lack a 'not' or 'else' function and also don't follow a defined order (with the ability to end filtering in a given filter) then it probably won't be possible. Nil Einne (talk) 20:09, 15 January 2011 (UTC)

Now they've found your email unfortunately you will get more spam. One bit of free advice - never click on a link in such spam, in particular never click on a link in spam saying you want to be removed from their mailing list. That simply tells them you read the spam and are a more worthwhile target. Dmcq (talk) 21:20, 15 January 2011 (UTC)

If I don't want garbage addressed to other people stuffed into my inbox, then I don't see why I should get it. Why bother whining? Why not simply fix the problem? If that means I never get something related to "BCC" again, then so be it. No more excuses. If a real postman did this, they'd rightly get sacked. 92.29.122.203 (talk) 11:10, 16 January 2011 (UTC)

There's no problem to fix. If you want to violate established internet standards for something that will achieve little, that's your choice and you've already been told of ways you can do it, just don't expect companies who actually think about this sort of thing and so know why it's a bad idea and why it's pointless to do it. And yes why you're whining at us none of which I suspect work for Microsoft because you don't like standards and think they should be ignored for no reason is beyond me. If a real postman decided to open your mail because they didn't trust the address on the outer envelope and didn't deliver the mail because the address inside was different they would indeed be sacked and not just for the privacy violation. (This isn't the best example but the postman analogy is starting to fail and this is the best example I can think of which fits your situation.) Nil Einne (talk) 15:09, 16 January 2011 (UTC)

Of course, spam filtering can be done pretty well. Spam Assassin is a good program that assigns a "score" to an email based on a large number of rules. (If the recipient is only a BCC it penalizes the email a few points.) Then the user can set up their email client to filter out any email below a certain score.

So far as I know, none of this could be done by an end user to a Hotmail account, sadly. But if you ran your own email server you could easily install this software or others like it and control it anyway you like. (And when it threw out an important email that you needed to read, well you could complain to yourself for setting it up, I guess.) APL (talk) 19:11, 16 January 2011 (UTC)

As others have said, the Simple Mail Transfer Protocol, which is how emails are routed to the recipients, ignores the To: and Cc: fields. Please see the example in the SMTP article for how it is actually routed. CS Miller (talk) 15:58, 16 January 2011 (UTC)

Real postmen deliver junkmail that's not specifically addressed to me all the time. Sometimes it's to "Joe Smith or Current Resident", sometimes they give up pretending they know who lives here and write "Our Valued Customer" even though I'm not, and sometimes they don't even put the address. I'm not sure how this works, but you can get the post office to deliver a piece of mail to every resident in a certain area. APL (talk) 19:11, 16 January 2011 (UTC)

Of course, at least where I come from, you can reserve yourself against that. Taemyr (talk) 22:51, 16 January 2011 (UTC)

At least I get to recycle all the spam the mailman dumps though the letterbox. Now if only there was a way of doing that with email spam... ;-) Dmcq (talk) 00:24, 19 January 2011 (UTC)