Computing desk
< September 15	<< Aug \| September \| Oct >>	September 17 >

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.

September 16[edit]

Remoting into Ubuntu Desktop VM from Ubuntu Desktop Computer[edit]

Hi.

I have the following Linux machines.

Computer 1: one physical computer with one physical six-core processor and 24GB of RAM, running Ubuntu 10.04 Desktop Edition.
Computer 2: one virtual machine running Ubuntu 10.04 Desktop Edition, with two logical processors and 16GB of RAM assigned to it.

I want to use Computer 1 to graphically remote into and control Computer 2. There appear to be two programs for doing this; Remote Desktop Viewer and Terminal Server Client. Which of these two programs should I use, and what are the differences between the two?

Terminal Server Client asks for the following information under the General tab. (I understand the fields in the other tabs well enough.) What should be entered in each field?

Computer
Protocol (drop-down menu - options are RDP, RDPv5, VNC, XDMCP and ICA)
User Name
Password
Domain
Client Hostname
Protocol File

Thanks to all. Rocketshiporion♫ 03:15, 16 September 2010 (UTC)[reply]

Enable sshd on the target computer. Enable X-tunneling on the target computer. SSH from the remote computer to the target computer. Execute any graphical program you like from the command line and it will be displayed on your local computer. If you understand how to do this, then you are set. If you need help enabling sshd (or even understanding what SSH is), please ask. -- kainaw ™ 03:01, 16 September 2010 (UTC)[reply]

(Expanding answer after question was expanded) Remote Desktop Viewer is used to remote into a Windows machine, not Ubuntu. Terminal Server Client is similar. I've never used it for anything except remoting into Windows. You appear to want to remote from Ubuntu to Ubuntu, not Ubuntu to Windows. So, you need to first understand that Ubuntu is Linux. Linux is not a graphical desktop like Windows. It is a kernel. The graphical desktop you are using is most likely Gnome. That is a program running on Linux. It isn't a self-contained desktop. It is an X-windows manager. Instead of remoting into the desktop, you connect to the machine and tell it to display the X-windows elsewhere (on the machine you are physically using). Tunnelling X (as I mentioned before) is a popular and easy method for doing that. -- kainaw ™ 03:21, 16 September 2010 (UTC)[reply]

I don't exactly know what is SSH, and I have no idea how to do X-tunneling... —Preceding unsigned comment added by Rocketshiporion (talk • contribs) 03:46, 16 September 2010 (UTC)[reply]

SSH is a secure remote shell protocol; "X forwarding" or "X tunneling" refers to the use of SSH to manage a connection to a remote X Window System (which is the basic system upon which almost all Linux graphical programs are built). Nimur (talk) 08:15, 16 September 2010 (UTC)[reply]

First, I must note that I use RedHat, not Ubuntu. However, this is how it SHOULD work in Ubuntu (unless they screwed it up to make it user friendly).

I will call the one you want to connect to the server and the one you want to connect from the client. On the server, ensure sshd is running. As root, run /etc/init.d/sshd status. If you don't have sshd, you need to apt get install sshd. If sshd is stopped, run /etc/init.d/sshd start to start it. Because you will be forwarding X, you need to ensure the server allows it. In /etc/ssh/sshd_config, look for an entry for X11Forwarding. Make sure the option is set to yes and there is no # in front of it. The # means it is commented out (ignored). If you changed it, restart ssh with /etc/init.d/sshd restart. Now, the server is ready - assuming no firewall issues.

From the client, you should have ssh installed - it is standard. Enter ssh -X -C server-username@server-ip. You must put in your username and ip for the server. You will log in and be at the shell on the server. The -X turned on X tunnelling. Test it by entering something like xclock. The xclock program will be running on the server, but displaying on the client because X is remotely displaying everything.

When I show Windows users how to do this, they tend to feel very uncomfortable with it. So, they SSH in and run gnome. This launches an entire gnome session with remote display. So, you see the desktop of the remote machine in a big window. It isn't really the desktop of the remote machine. It is a second desktop being displayed remotely. So, you are running gnome twice (once local, once remote) on the server. Further, you get no benefit from this. It just makes everything much slower. After a while (and a lot of complaining), they only run the programs they want to use and the windows open on their local machine instead of the server. -- kainaw ™ 14:13, 16 September 2010 (UTC)[reply]

I have known about X forwarding over SSH for over a decade and didn't even know it was possible to run the entire Gnome desktop over X forwarding. I guess you learn something new every day, or something. JIP | Talk 14:54, 16 September 2010 (UTC)[reply]

NX_Client may be of interest to you. I used to use it to access one debian computer from another 129.67.37.143 (talk) 11:45, 16 September 2010 (UTC)[reply]

Seconded. Works great (even over low-bandwidth connections) and is able to selectively display a single application from the target computer or an entire desktop. Though if you simply want to connect to your virtual machine running on the same hardware (no physical LAN/WAN involved), VNC should do just as well. IIRC, both Gnome and KDE have built-in support for this; you could also simply start xvncserver on the virtual machine, which will create a virtual desktop that you can connect to with xvncviewer. -- 78.43.71.155 (talk) 18:55, 16 September 2010 (UTC)[reply]

Free Complete Multi-Root Operating-System[edit]

Hi.

I am seeking a complete free multi-root operating-system, other than ReactOS (as ReactOS is an an uncompleted operating-system). Is there currently any such operating-system in existence?

Thanks in advance to all respondents.

—Preceding unsigned comment added by Rocketshiporion (talk • contribs)

What do you mean by "multi-root"? This terminology is not commonplace to describe operating systems. In the mean-time, see comparison of operating systems. Nimur (talk) 08:10, 16 September 2010 (UTC)[reply]

By multi-root, I mean an operating-system (similar to Windows), which accomodates multiple root drives (A:,B:,C:,etc.); as opposed to Unix/Linux, where everything (even physically seperate drives!) has to be mounted as a subdirectory of the single root /. Rocketshiporion♫ 13:44, 16 September 2010 (UTC)[reply]

I actually probably could have guessed this earlier from Root directory#Multiple root directories but didn't look at the results properly. FreeDOS? Nil Einne (talk) 14:26, 16 September 2010 (UTC)[reply]

From Path (computing), perhaps some of the other clones/copies of other operating systems with similar behaviour? I was looking in to this myself but didn't find anything definitive. But perhaps some of the Amiga OS clones? Nil Einne (talk) 15:48, 16 September 2010 (UTC)[reply]

If you follow the loosely related development path from Amiga OS to BeOS to Haiku, Haiku appears to be using a single-root filesystem that looks functionally like *nix, but uses less cryptic directory names. Strangely, I can't remember what the filesystem was like on the Amiga even through I developed Amiga software for about 10 years. -- kainaw ™ 16:53, 16 September 2010 (UTC)[reply]

ReactOS is an NT clone, and NT has a single-rooted object hierarchy like Unix. The drive letters are implemented in the user-mode Win32 library that most (not all) Windows applications use. A path like C:\dir\file is translated to \??\C:\dir\file before being sent to the kernel. ?? is a subdirectory of the root directory, despite its strange name (prior to WinXP it was called DosDevices). \??\C: is a symlink to the actual filesystem mount point, which is typically something like \Device\HarddiskVolume1. The current directories for non-active drives are stored in environment variables named =C:, =D:, etc. Most libraries hide these from you, but you can see them if you parse the environment block yourself. If you want a similar hack on top of some other OS, I suppose there's Wine. -- BenRG (talk) 19:00, 16 September 2010 (UTC)[reply]

Purely out of curiosity, why do you care about having multiple root directories? If it's just for looks, you could always mount things as /a:, /b:, etc. (and make /c: a symlink to / for the full DOS experience) if you wanted to. (It would even be the same length if you dropped the colon, at which point it looks almost like just a different syntax for the same thing.) --Tardis (talk) 15:06, 16 September 2010 (UTC)[reply]

In response to Kainaw, the Amiga's file system hierarchy went DFn: for floppy disks, DHn: for hard disks and CDn: for optical drives. Here n starts at 0 for the first mounted disk and increments for each subsequent disk of the same type. Of course there was RAM: for the RAM disk and then assigned shortcuts such as SYS:, TEMP: and (I vaguely think) UTILS: Zunaid 21:24, 16 September 2010 (UTC)[reply]

In response to Tardis, I'm quite used to storing files/folders directly on the root of a volume (i.e. e.g. F:\), and I personally don't like the idea of storing my files/folders in a non-root directory such as /home/rocketshiporion/. As for mounting a volume as /a:, /b:, etc.; AFAIK non-root names cannot contain the : character.Rocketshiporion♫ 22:25, 16 September 2010 (UTC)[reply]

It would have taken you less time to verify that it is possible than to falsely assert that it is not. -- Finlay McWalter ☻ Talk 22:32, 16 September 2010 (UTC)[reply]

The file name rules on Unix are incredibly simple: everything is valid. That is to say, every C string (thus: no NUL characters) whose length is neither 0 nor "too long" (at least 256 or 1024 characters) can name a file. Slashes of course separate it into components, all but the last of which must (after symbolic link resolution) refer to directories, but since multiple slashes are (except perhaps at the beginning of the name) equivalent to a single slash, everything is valid. --Tardis (talk) 22:53, 16 September 2010 (UTC)[reply]

I have checked it now, and found that while it is possible to create the directory /home/K:/ , it is nigh impossible to create the directory /K:/ (i.e. directly at the system root) as there is no option at all in the right-click menu to create any directory at all directly at the system-root. Rocketshiporion♫ 06:30, 17 September 2010 (UTC)[reply]

What's stopping you from opening a shell and typing mkdir /K: as root? JIP | Talk 06:52, 17 September 2010 (UTC)[reply]

I'm using Ubuntu Desktop Edition, and the shell cannot be opened - going to Applications->Accessories->Terminal has no effect. Rocketshiporion♫ 08:49, 17 September 2010 (UTC)[reply]

OK, I didn't know that, I use Fedora myself. But I cannot help but wonder what good Ubuntu thinks not being able to use the shell would do. JIP | Talk 09:16, 17 September 2010 (UTC)[reply]

Not being able to access the Terminal application is a clear bug, something seems to be wrong with your Ubuntu installation. Since I'm using Kubuntu, I'm not too familiar with Gnome, the standard desktop of Ubuntu, which you are using, so I'm hoping someone else will jump in and tell you how to get that Terminal application running.

A workaround would be to press [Ctrl]-[Alt]-[F1], which should switch to the text screen and show you a login prompt. Log in there with your regular user account and corresponding password, then type the following commands:

sudo mkdir /K: #this will prompt you for your user password again, enter it
#the following line copies user and group names for the ownership information from your home directory
sudo $(ls -l /home/ | grep yourusernamehere | head -1 | awk '{ print "chown "$3"."$4" /K:" }') 
sudo chmod 700 /K: #if you want other users to be able to access your files, apply a more relaxed setting here
logout

Now you can switch back to the GUI by pressing [Alt]-[F7]. -- 78.43.71.155 (talk) 10:37, 17 September 2010 (UTC)[reply]

If you already know and remember your user and group names, then there's no need to write all that complicated stuff on the third line. Suppose your user name is rocketshiporion and your group name is users, you can just write:

sudo chown rocketshiporion.users /K:

Of course, the line provided by 78.43.71.155 above will work whatever your user name and group name are. JIP | Talk 12:53, 17 September 2010 (UTC)[reply]

And that's why I used it - While I'm sure Rocketshiporion knows his user name, he might be unfamiliar with Linux groups. -- 78.43.71.155 (talk) 21:36, 17 September 2010 (UTC)[reply]

It's worth noting that, under Unix-like systems, the filename ~ is an abbreviation for your home directory. This can be used in paths; i.e., the file ~/asdf refers to a file named asdf in the home directory. It cooperates nicely with tab-completion, also. Paul (Stansifer) 11:50, 17 September 2010 (UTC)[reply]

Ah, it works! I have successfully accessed the command-line using [Ctrl]+[Alt]+[F1], and created the desired folders. Guess I won't be needing an open-source multi-root operating-system after all. Rocketshiporion♫ 01:22, 18 September 2010 (UTC)[reply]

Since I was the one suggesting that particular solution, may I ask for something in exchange? Please make sure your rocket ship, if you ever get around to building one IRL, doesn't crash in (or anywhere near) my backyard, "near" being defined as "within blast/fallout radius". ;-) -- 78.43.71.155 (talk) 01:43, 18 September 2010 (UTC)[reply]

In the unlikely event that I manage to get the U.S. Congress to approve the spending of a few tens of billions of dollars (could well reach a hundred billion dollars), I will have the nuclear-propelled rocketship built (and launched from) one of the following places; Devon Island, Jarvis Island and Howland Island. As long as you're not living in any of these places, you should be just fine, because the rocketship will do one of the following three things.

If all goes well, it will reach orbit, and continue with its journey (hopefully with me aboard it).
It might suffer a mid-flight malfunction, and crash into the sea (hopefully without me aboard it)!
In the worst-case scenario, it could explode on the launchpad, and wipe the whole island off the face of the Earth (and I'll probably be either aboard the rocketship or on the on that island)!

Rocketshiporion♫ 06:40, 18 September 2010 (UTC)[reply]

Umm, what about #4, a catastrophic failure of your nuclear propulsion system that occurs when you're in the one of the upper atmospheric layers, which exceeds the estimated Design Basis Accident and causes uncontrolled nuclear fission

and some nasty fallout?

I'm not sure how old you are, but I'm old enough to remember the Chernobyl incident and what an ugly mess that was, see these charts of the fallout:

and [1] - and mind you, that thing went off on the ground and still contaminated such a huge area. A detonation in an upper atmospheric layer will distribute the stuff even further, so you might be riding on top of the world's largest dirty bomb. -- 78.43.71.155 (talk) 12:39, 18 September 2010 (UTC)[reply]

This discussion no longer has anything to do with directories on operating systems, or computers at all for that matter. Please take it to Wikipedia:Reference desk/Science. JIP | Talk 14:04, 18 September 2010 (UTC)[reply]

Will do, assuming Rocketshiporion wishes to reply. Is there a standard procedure to copy/move content between the desks, or is it a simple copy and paste? -- 78.43.71.155 (talk) 16:02, 18 September 2010 (UTC)[reply]

MediaWiki only supports move operations for entire pages, so you'll have to copy and paste. JIP | Talk 16:46, 18 September 2010 (UTC)[reply]

I will notify Rocketshiporion on ~~his~~her talk page and we'll see where to take it (WP:RD/S or one of our talk pages) if heshe wishes to go on. Kiitos & Qapla! -- 78.43.71.155 (talk) 21:27, 18 September 2010 (UTC)[reply]

Cont'd on my talk page. -- 78.43.71.155 (talk) 09:43, 20 September 2010 (UTC)[reply]

Resolved

Vista Themes[edit]

I've just installed some important Windows updates, and they've messed my computer up. The theme has changed to classic and the old Vista theme has been removed from my themes. I can't seem to find any Vista themes on the Microsoft website either; they're all Windows 7 themes. Does anyone know of an official Microsoft source, or at least a safe one to get the theme back? 81.102.190.237 (talk) 09:45, 16 September 2010 (UTC)[reply]

When all else fails, use System Restore under Accessories/System Tools. Restore your computer to a point in time to before the upgrade had run, then make sure you turn off the automatic upgrade feature. Don't worry, this only changes system settings and does not affect your personal data. Sandman30s (talk) 10:47, 16 September 2010 (UTC)[reply]

Strange Icons[edit]

The files on my desktop not have a solid red disk containing a white exclamation mark over them. What does it mean, and why have they started to appear? I googled it and some people mentioned Mozy Home, or something. But I've never heard of Mozy Home, it doesn't show on my software list in control panel, and I've not accepted a download from them. 81.102.190.237 (talk) 09:45, 16 September 2010 (UTC)[reply]

Someone over here says McAfee did it. If that's not it, try taking a screenshot and uploading it to http://imgur.com/ so we can see what they look like. Lots of different systems (TortoiseSVN, for example) can decorate your files like that. --Sean 15:05, 16 September 2010 (UTC)[reply]

Could also be Tortoise SVN, but presumably you'd know if you were using that. APL (talk) 02:16, 17 September 2010 (UTC)[reply]

Dropbox does this Quadrupedaldiprotodont (talk) 15:16, 17 September 2010 (UTC)[reply]

Programming question[edit]

I am trying to program a label (C#, GDI+) to display a line of text with word-wraping and to have one character flashing. I need the entire character to flash, and possibly the background also. What is the best way to go about programming this? So far I have split the text into words (and spaces), measured the space of the words, and stored this information in a list. Next, it looks like I will have to split one of the words up, so that one of the letters can be drawn in a different color. I just have this nagging felling that I am going about this all wrong.

Thanks, 86.41.187.20 (talk) 14:59, 16 September 2010 (UTC).[reply]

I don't know that API, but it sounds like a reasonable solution to me. If you're new to programming, don't worry about doing it "right", just git 'r done! --Sean 15:07, 16 September 2010 (UTC)[reply]

and copiously comment to yourself how you did it!!! If it's a hack, place comments before and after it, clearly showing the exact part of the code that you aren't happy with, what a simple, elegant replacement would be (if you knew how to do it) and how you're going about it instead of in that elegant way. If you simplify or change the hackish code, change the comments to reflect it! 84.153.224.118 (talk) 15:10, 16 September 2010 (UTC)[reply]

I would split the string into 3 parts: the part before the flashing character; the flashing character; and the part after the flashing character. Use DrawString to draw the first bit of text. Then move the x co-ord the correct amount using MeasureString. Then use a Font with one different colour to draw the single character, and then update x, etc. Then use a Timer to alternate the colour of the middle character. --Phil Holmes (talk) 16:46, 16 September 2010 (UTC)[reply]

This may not be sufficient, depending on the workings of GDI+'s font anti-aliasing engine. On many anti-aliasing systems, overpainting a character in one colour with a copy of itself in another can produce a weird halo around the character, as the anti-aliasing of the second draw blends with that of the first (etc.). The safe thing to do is to fill the character space with background colour before drawing a fresh copy. -- Finlay McWalter ☻ Talk 17:13, 16 September 2010 (UTC)[reply]

Hey - we have to leave something as an exercise for the student :-) --Phil Holmes (talk) 17:16, 16 September 2010 (UTC)[reply]

You can calculate the font-width of the string before the flashing character. Then, you know the left coordinate of the flashing character. Draw the label with plain text. Then, on a timer, draw the character in the calculated coordinate in plain text, then in background text, then in plain text, then in background text... If you are interested in how this used to be done eons ago... You had a color map. You'd set two indexes to the same color. Then, you write the string using one index for the text and another index just for that one flashing character. At first, they look the same. Then, you change the color at the index used for the flashing character and the character changes color. Since computer graphics are not indexed that way anymore, that trick doesn't work. -- kainaw ™ 16:57, 16 September 2010 (UTC)[reply]

Splitting the text into 3 parts won't work, I need word wrapping. I have the string "Hello World, How are you today?" (Note the three spaces between Hello and World), split into 13 boxes (6 words + 7 spaces). The boxes are layed out on a line until space has run out, then it moves onto the next line. This part of the code works fine. Next it looks like I will need to split one of the boxes into 3 parts, with the middle part containing the flashing character. I can't split the flashing character into it's own box as it would then word-wrap in the middle of a word. I just need to make sure that there isn't a better way of doing this before I continue. I know other programs have had to solve this problem before, so how do they do it? Also, if anyone knows of any tutorials on the net about this topic I would love to see some links. Thanks, 86.41.181.177 (talk) 22:39, 16 September 2010 (UTC)[reply]

In practice, once you start treating a text as a collection of parts, where some are handled differently to others, you can no longer rely solely on the in-built behaviour of components that implicitly treat their texts as a single item. So in this case, to do what you suggest, you indeed cannot rely on the underlying control's text wrapping behaviour. If you go this path, you'd have to do the text layout yourself, figuring out the sizes of the various component strings using MeasureString, calculating where each should be, and drawing them yourself. Such is an inevitable consequence of going beyond the behaviour implemented by a simple control. In essence your unusual requirement means you have to implement what amounts to your own GUI control, which is at least an order of magnitude more work than simply using the features built into an existing one. You may wish to rethink your approach; if you're implementing some kind of text-based entry (like a terminal) then maybe the TextBox object (which already has a built-in cursor) will be a more suitable thing to use. -- Finlay McWalter ☻ Talk 22:47, 16 September 2010 (UTC)[reply]

I am creating a custom GUI control. (It inherits from Control.) The unusual requirements is exactly why I'm not using a standard Windows control such as a Textbox. The control will be used for the display of text, and not for text entry. And an entire character (or more) and/or it's background needs to flash, a flashing line between letters is not enough. 86.45.182.35 (talk) 23:02, 16 September 2010 (UTC)[reply]

Another suggestion: perhaps you could use some kind of HTML-rendering widget and just put a blink tag at the right place. --Sean 15:42, 17 September 2010 (UTC)[reply]

Then you'll need to implement the word wrapping yourself. For speed, implementers of multi-line text controls sometimes cache the widths of each word, so they don't have to make dozens or hundreds of calls to the underlying font-metrics code every time they re-layout the page. Whether that's a worthwhile optimisation for you depends on how frequently, if at all, the text in your control changes, and in what manner. -- Finlay McWalter ☻ Talk 23:09, 16 September 2010 (UTC)[reply]

I suspect that word-wrapping is one of those things that everyone under 30 hates and everyone over 30 despises from college. Shadowjams (talk) 08:35, 17 September 2010 (UTC)[reply]

Why? Nil Einne (talk) 09:55, 17 September 2010 (UTC)[reply]

I would recommend trying to reuse an OS implementation for word wrapping if at all possible. Latin based scripts are fairly simple, but Japanese (禁則処理) and many other scripts are significantly more complex. If still needed, then I suggest referring to UAX #14. 180.11.188.56 (talk) 08:51, 17 September 2010 (UTC)[reply]

That's an astute observation. Detecting appropriate positions for breaks is generally the function of the i18n library. Java does it with BreakIterator, which detects both word and natural line-break points. The equivalent Windows API seems to be TextElementEnumerator, but that seems to reports breaks only on grapheme boundaries. This guy has written a word breaker, but I'm surprised (hopefully I just can't find it) that the Windows platform doesn't come with a pretty comprehensive set of word and line breakers (in so far as "word" breaker means anything for CJK). -- Finlay McWalter ☻ Talk 22:24, 17 September 2010 (UTC)[reply]

Fonts[edit]

What typeface font gives the impression of prestige and grandeur? —Preceding unsigned comment added by 116.15.154.193 (talk) 15:01, 16 September 2010 (UTC)[reply]

Well, there are plenty. One way to figure out such a thing is to look at the typefaces used by places that try to give such impression. The Waldorf Astoria website, for example, has about three very different typefaces that all work pretty well for that (the sans-serif one used on the logo, which looks like some variation on Century Gothic; the "singular experience" font, which looks like Caslon; and the ornate script one in the lower right hand corner). As the page also makes clear, it's not just the font that matters, it's the overall presentation -- the white on black, for example. --Mr.98 (talk) 15:44, 16 September 2010 (UTC)[reply]

Comic Sans is well known for working in every situation but is particularly good when you want prestige and grandeur. In terms of the overall presentation, I would suggest very bright and garrish colours. I believe Steve Baker has remarked before that crank websites are an excellent style guide for this sort of thing. I would also take a look at old websites. Nil Einne (talk) 16:18, 16 September 2010 (UTC)[reply]

Your brief would be an important nudge in the right direction. If you're targeting wealthy retirees to invest in a golf course that would be different to designing wedding invites for your mate. Current fashion is for light to medium font weights and sans serif, but depending on your brief you may need something like a script, small caps (I'm still partial to Trajan Pro even though it's in common use) or a serif! Or have a look at the main typeface libraries and if you see something you like google for a free alternative. Spoonfulsofsheep (talk) 17:22, 16 September 2010 (UTC)[reply]

Nobody would expect a typewriter. Shadowjams (talk) 08:29, 17 September 2010 (UTC)[reply]

For a somewhat tongue-in-cheek answer to the question, see So You Need a Typeface. — QuantumEleven 13:44, 17 September 2010 (UTC)[reply]

I'd have laughed but I spotted a good few of my fave fonts in that chart :-) Spoonfulsofsheep (talk) 22:49, 17 September 2010 (UTC) [reply]

Somehow Computer Modern is not mentioned at all. So what good does the flowchart do me in reconstructing my personality? --Trovatore (talk) 02:50, 19 September 2010 (UTC)[reply]

Times New Roman Quadrupedaldiprotodont (talk) 15:15, 17 September 2010 (UTC)[reply]

How about Trajan (font)? -- AnonMoos (talk) 14:24, 21 September 2010 (UTC)[reply]

View Options—Mac OS X[edit]

In Mac OS X I search through a folder containing files created over many months. I can choose "View options" including "Date modified" and "Date created". But it is only noting the time of the day. I'm interested in the date on which it was created. How do I get it to show me that "date" on which a file was created or modified? Bus stop (talk) 17:06, 16 September 2010 (UTC)[reply]

You propably have the column width for dates set too narrow. click and drag on the column separator in the column header row - the dates should appear once you've expanded it enough. --Ludwigs2 22:51, 16 September 2010 (UTC)[reply]

That's not it. No matter how much space I give it all it displays is the time of the day. It seems pretty useless—why would anyone want to know the time of day in absence from the month and day if one were looking for a file? Thanks for the suggestion. Bus stop (talk) 00:43, 17 September 2010 (UTC)[reply]

Well, this isn't a helpful response entirely, but on mine (running 10.6.4), checking "date created" shows me the full date plus time. So there's something funny on your end; it can certainly do what you want it to do. --Mr.98 (talk) 01:13, 17 September 2010 (UTC)[reply]

Mine too. interesting... the normal scattershot fixes for odd Finder problems like this would be:

Restart the computer
Delete the Finder preferences (~/Library/com.apple.Finder.plist) and restart the Finder
Check that you don't have some kind of funky date preference set in the Formats section of the Languages & Text preference pane

If none of those fix it, create a new user account (just a temporary diagnostic account) to see if the problem occurs there as well.

do the proper dates for the files show if you switch to column view or use get info on a file? --Ludwigs2 02:02, 17 September 2010 (UTC)[reply]

I don't know about Mac OSX, but in some programs that list date-and-time stamps, showing only the time is an abbreviation that means today at the time shown. There may or may not be a way force it to always show the date. --70.253.38.181 (talk) 02:12, 17 September 2010 (UTC)[reply]

Yes, well the only thing it shows is "Today" or "Yesterday." All the other (past) dates are only identified by time of day. It is pretty funny. I am not using that computer right now, so I can't try some of these diagnostic suggestions. Tomorrow I will be back at it. It is pretty funny—it sort of stands reality on its head. It makes it seem that it is important that I know that a file created 6 months ago was created at 3 PM. Bus stop (talk) 03:22, 17 September 2010 (UTC)[reply]

I have restarted the computer. It is still only showing times of day. I don't know what it means to, "Delete the Finder preferences (~/Library/com.apple.Finder.plist) and restart the Finder," as Ludwigs suggested above.

In response to Ludwigs' question, "do the proper dates for the files show if you switch to column view or use get info on a file?," this is in column view that I am talking about. I should have said that. It is in column view that the columns for date created and date modified are not showing dates at all, but rather only times of day. Very strange. And in answer to the second part of his question—concerning "Get info"— same problem—only times of day! "Get info" is also displaying the identical problem.

As I say—it is funny. For a date it is telling me nonsense like "12:17 PM"!

I'm not going to have too much time to work on this today. I just thought I'd keep you all updated. And thanks for your help. Bus stop (talk) 14:23, 17 September 2010 (UTC)[reply]

Update: I've tried Ludwigs' interesting suggestion to create a new user account. I created that account. But of course it had no old files. So I plugged in a thumb drive with a lot of old files on it. They are reading properly! They are showing the complete date, in column view. I then logged out of that account and logged back into my usual account. I plugged the thumb drive in again. The old files on the thumb drive are only indicating time of day. So the problem seems to be confined to this account, which is my usual account. One thing I can do is just move all my stuff to the new account. Oh well, there is no urgency to do so. Bus stop (talk) 14:51, 17 September 2010 (UTC)[reply]

No, no need to move all your files. as I said, that was just diagnostic. what it tells me is that you have (a) a corrupt preference file, or (b) a system addition that's causing your problem.

so, go to your user preferences folder (the Preferences folder in the Library folder of your home folder) and delete the file called 'com.apple.Finder.plist'. then log out and log in again. see if that fixes the problem. --Ludwigs2 15:30, 17 September 2010 (UTC)[reply]

Ludwigs—I think I've done what you've suggested, and it hasn't helped. I've dragged 'com.apple.Finder.plist' to the trash, emptied trash, and restarted the computer, and the time of day is all that is being displayed. Thanks for trying. Bus stop (talk) 18:22, 17 September 2010 (UTC)[reply]

fascinating. just out of curiosity, is the 'use relative dates' checkbox checked in the View Options panel? --Ludwigs2 19:40, 17 September 2010 (UTC)[reply]

I've tried it both ways—with "Use relative dates" checked and unchecked—no difference. Thanks for your help. At least I now understand that I have the option of abandoning this account and setting up shop in another account. Funny stuff. Thanks again. Bus stop (talk) 22:02, 17 September 2010 (UTC)[reply]

I've gotten back the full date. I fiddled around with the "Language and Text" "Formats" preferences, as suggested by Ludwigs, and also the "Date and Time" preferences. I'm not aware that I changed anything. But I just tried various alternatives, always returning the setting to what it had been. Not sure how that might have helped, or even if that is what did it, but that's my best guess. Thanks all. Bus stop (talk) 12:03, 19 September 2010 (UTC)[reply]

PHP imagecreate to imagick[edit]

I can do two things easily in PHP, but I want to combine them and I don't see a way from one to the other. I can create images on the fly using imagecreate and all the image functions. I can load an image file into imagick and alter it. I want to take an image creates with imagecreate and directly use that as an imagick image. I don't want to save the image to the harddrive and then reload it. Is it possible to transfer directly from imagecreate to imagick. I would be nice if this worked: $frame = new Imagick(imagepng($image)); -- kainaw ™ 17:56, 16 September 2010 (UTC)[reply]

it's probably best just to create the image in imagick - see Imagick::newImage. Is there a reason you need to do it using imagecreate? --Ludwigs2 22:50, 16 September 2010 (UTC)[reply]

Unless I'm being selectively blind, imagick doesn't have the ability to create images by drawing lines, rectangles, and such. So, I have to draw the image with imagecreate. However, imagecreate cannot place multiple frames into an animation. So, I'm stuck trying to jump from one to the other without saving files on the drive. I'm considering making a virtual drive in RAM, saving images there, animating them, and then deleting the virtual drive. -- kainaw ™ 13:59, 17 September 2010 (UTC)[reply]

Isn't Imagick Draw what you need, then? This seems like an example of roughly what you're doing. -- Finlay McWalter ☻ Talk 18:19, 17 September 2010 (UTC)[reply]

Thanks! That is exactly what I'm looking for. -- kainaw ™ 22:43, 17 September 2010 (UTC)[reply]

Outlook rule on outgoing messages[edit]

Suppose I have two different sets of clients, to whom I send e-mail many times per day in Outlook 2007. Recently, when typing into the "To:" line the names of various people at Client 1, I'll find myself accidentally and automatically typing the name of one particular person at Client 2. Actually sending the e-mail would be a major mistake. I tried setting up a rule in Outlook in order to try to detect whether this one particular person at Client 2 is on the "To:" line along with a particular person at Client 1, so that when trying to send an e-mail to both, Outlook would stop and ask me if I'm certain I know what the hell I'm doing.

Does anyone know how to do this? The farthest I got was setting up a rule to act after a mail has been sent, if person 1 or person 2 was on the "To:" line. (That first point I can work with because my e-mail is set to connect only every 5 minutes, so an after-the-fact alert is actually going to be OK for me most of the time.) I also didn't see a great action to cause Outlook to take even if I could define the trigger points for this rule. Comet Tuttle (talk) 20:42, 16 September 2010 (UTC)[reply]

Not that this helps directly, but you might Google for some sites to explain how to duplicate the functionality in Outlook: Gmail has a Labs feature called "Got the wrong Bob?" which identifies when you've put an odd name into a list of people you normally email. It also has another feature called "Don't forget Bob" which suggest missing recipients from a list of people you'd normally email together. Zunaid 21:30, 16 September 2010 (UTC)[reply]

I'm not sure how Outlook 2007 works as I found it bloody hard work to use but I've set-up separate identities in Outlook Express so I can swap between clients. You can do the same as per this link which also means you can have emails from 1 set of clients going to 1 account and the others to your 2nd account which makes life much easier. 87.112.106.143 (talk) 22:44, 16 September 2010 (UTC)[reply]

You may need to use VBA. I've never used VBA in Outlook, but I expect you'd hook into the Send event, pop up a message box asking you if you're sure you want to send the message when it meets a certain condition, and set Cancel to True if you decide to abort. I'm sure someone proficient in VBA could whip it up in 5 minutes. Riggr Mortis (talk) 23:29, 16 September 2010 (UTC)[reply]

It depends on a few factors but if you are mostly sending emails to similar groups of people you could set up distribution lists instead. Say you're frequently sending emails to the same 3 people in the accounts section, or 5 people in management. You could set up distribution lists called client 1 accounts, client 1 management. Then you don't have to type individual names, and you'll never accidentally send to anyone at client 2. Vespine (talk) 00:29, 17 September 2010 (UTC)[reply]

Thanks for the suggestions. "Got the wrong Bob?" sounds spot-on, though I won't use GMail for work e-mail. Given time constraints, I think I may create a mailing list and train myself to use that. Thanks! Comet Tuttle (talk) 20:59, 17 September 2010 (UTC)[reply]

You could use Google Apps if you have your own domain name Nil Einne (talk) 13:50, 18 September 2010 (UTC)[reply]

I don't know if this will fit your case, but what about putting your different clients into different personal address books? Then you could train yourself to do a two-step process: select the right address book, and then one or more people from it to email to.

rc (talk) 04:13, 19 September 2010 (UTC)[reply]

Software to plan PhD project?[edit]

I've just/am just about to begin a PhD project and I need to begin planning it. I'd like to use software to keep track of my plan, since I'll need to modify it and present it to different people in different ways, as well as use it as a guide. I'll want the software to be able to summarise or give details, showing information in textual and graphical forms, include time periods. Can anyone recommend software for this purpose? I don't think I need all the features of MS Project. Does anyone here think there are alternatives that will suit my pretty basic needs? --178.98.218.244 (talk) 21:54, 16 September 2010 (UTC)[reply]

Have you checked out this list of software?. I think I used something from VCS or VSC but that was many years ago.. if I remember the name I'll post it. Spoonfulsofsheep (talk) 23:21, 16 September 2010 (UTC)[reply]

Wikipedia:Reference desk/Archives/Computing/2010 September 16