Wikipedia:Reference desk/Archives/Computing/2012 July 22

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.

July 22

removal of antivirus

It turns out I have some rather nasty adware on my computer, which is rather strange, because it is the sort of thing that is supposed to save me from this sort of thing. It's known as McAfee, and it keeps prompting me to buy. It was, of course, shipped with Windows Vista, but now it has well and truly outlived its useful lifespan, because there is nothing so irritating as a constant popup ad on your desktop. It used to be just once when I fired up the computer, but now, just in case I didn't get the message the first 3 million times, it is telling me about 4 times a session to upgrade. I have the free version of Malwarebytes (which interestingly, is telling me my trial has expired, but I don't remember any trial, and my understanding is that this will revert to the free version anyway). With the free version of Malwarebytes, is it safe to remove the McAfee? I've really had enough of it, and I just need to know if there is anything it is doing in the background that is not provided by Malwarebytes. If so, is there a free alternative? I wouldn't mind paying, but it seems the paid stuff is the stuff that comes with advertising - imagine if cable tv was full of ads, and free to air had none - would you buy cable? IBE (talk) 02:14, 22 July 2012 (UTC)

I have always bought my computers from one of the best in the city. They have always installed http://free.avg.com/ca-en/homepage in them when I ask to have the best free one they know of. The free one has protected 3 of my systems well for close to 10 years.--Canoe1967 (talk) 04:27, 22 July 2012 (UTC)

I too suffer from a similar 'adware' problem but from the 'trial' version of Norton Internet Security that came on my laptop. I also use AVG Antivirus free version and Malwarebytes' Anti-Malware and they seem to keep the overwhelming majority of viruses and malware at bay. If you install AVG you should be safe removing the McAfee software. - 220 of ^Borg 10:15, 22 July 2012 (UTC)

Most things that come preinstalled with Windows are going to end up annoying you. ¦ Reisio (talk) 00:28, 23 July 2012 (UTC)

Including Windows. Thanks for the advice. IBE (talk) 01:43, 23 July 2012 (UTC)

Weekly menu planner

I need a program that will help me plan weekly menus for a child developement center. using my meals and not having any repeats for 2 weeks. breackfast, lunch and snack. is there one i can download for free or cheap? — Preceding unsigned comment added by Hughesmarni (talk • contribs) 04:36, 22 July 2012 (UTC)

Do you want it to do anything besides listing the meals, like counting calories, making sure each meal is balanced, etc. ? If not, I expect a spreadsheet would work. If you have a Windows operating system, the Microsoft Office suite includes a spreadsheet program. You will have to manually verify that there are no repeats. I imagine just having 2 weekly menus which alternate and repeat is the easiest way, with perhaps something special substituted in on holidays. StuRat (talk) 05:15, 22 July 2012 (UTC)

You could also just use a text editor, and set it up like this (adding weekends if you have meals then):

            +----+----+----+----+----+
  Dates     | M  | Tu |  W | Th |  F |
+-----------+----+----+----+----+----+
| BREAKFAST |    |    |    |    |    |
+-----------+----+----+----+----+----+
| LUNCH     |    |    |    |    |    |
+-----------+----+----+----+----+----+
| SNACK     |    |    |    |    |    |
+-----------+----+----+----+----+----+

Your version will be larger, of course, and you would need to use a fixed size font (mono-spaced), like courier, to get the columns to work out right. StuRat (talk) 05:30, 22 July 2012 (UTC)

jpg pictures

I have around 4,000 jpg pictures taken by my digital camera. They're around 2MB each, and thus consuming quite a lot of hard drive space, and as I keep taking more pictures this is only increasing.

I did some tests and found that the pictures still look okay when saved as low as 800KB (with the same dimensions).

I am therefore considering reducing the quality of all the pictures to this level to save disk space.

Some questions;

1. There is almost no noticable difference (but obviously some) to me when looking at the original and the compressed versions. Is there anything I am missing here, something I haven't considered that would warrent keeping the originals? — Preceding unsigned comment added by 75.87.101.44 (talk) 13:56, 22 July 2012 (UTC)

No, you're not missing anything, other than that hard drive space is ridiculously cheap.

As you probably know (and as our JPEG article explains), "lossy" compression algorithms like JPEG can do a very good job of compressing pictures, since (a) there's a lot of order and redundancy in most pictures, and (b) it's really not a problem if you lose a little bit of information (indeed, as you've seen, one tends not to notice at all). So you can dial down the image quality, i.e. dial up the compression, i.e. dial down the file size, and save a lot of disk space, and not lose much of anything in return.

Mostly what you'll lose is time, doing all the re-compressing and re-saving! (Although there are programs which can automate this task for you, on lots of images at once.) Also, if you ever decide to print the pictures out on a high-resolution photo printer, or expand them to full screen, they may then look noticeably worse, and you may wish you'd kept the original, less-compressed versions.

Finally, when you say "a lot of hard drive space", are you speaking in absolute or relative terms? Yes, 4000 x 2 meg is 8 gig. But a modern personal computer is likely to have hundreds of gigabytes available (the one I'm typing on, which is about a year old, has 465), so you might be talking less than one percent of your available space. —Steve Summit (talk) 14:08, 22 July 2012 (UTC)

(ec)And 8Gb of hard drive storage costs, at desktop scales, less than $1.50. So spending the time to downscale all the pictures, and check them, will save 90 cents. That's a lot of work (even if you automate it, you need to set up the automation and check the results) to save a tiny amount of money. -- Finlay McWalterჷTalk 14:16, 22 July 2012 (UTC)

Did you mean "Gb" (Gigabits) or "GB" (GigaBytes) ? StuRat (talk) 18:49, 22 July 2012 (UTC)

Yes, bytes. -- Finlay McWalterჷTalk 18:51, 22 July 2012 (UTC)

Something else to consider is storing them on an external device, like a USB flash drive. This costs slightly more per GB but is much simpler, as it doesn't require you to open up your computer to add a hard drive (or, even worse, replace the hard drive, which involves moving everything else from the old one to the new one). This will free up the entire 8GB on your hard disk. You can also use multiple flash drives, if you prefer to separate your pics in some way. If so, be sure to label each one. One advantage to this method is the portability. If you don't need portability, you could go with an intermediate solution, by getting an external hard drive. However, while those cost less per GB, they come in larger capacities than USB flash drives, such that the total cost may be higher. StuRat (talk) 18:43, 22 July 2012 (UTC)

Note that an external hard drive (3.5" based one) can actually be cheaper per GB then internal drives in places like the US and NZ when on sale at bigbox stores. Nil Einne (talk) 06:14, 23 July 2012 (UTC)

"hard drive space is ridiculously cheap" What he said. ¦ Reisio (talk) 00:30, 23 July 2012 (UTC)

Another alternative to downsizing them is compressing them with something like 7-Zip. I'm not entirely sure what kind of ratio you'd get on jpgs but it's worth a try My Ubuntu (talk) 04:13, 23 July 2012 (UTC)

You won't get any noticeable compression with 7-Zip. You will with recent versions of WinZip and StuffIt if you use their JPEG-specific algorithms, but the savings are not all that large (less than 25%, I think) and it's very slow both to compress and decompress, so it's really not worth it. -- BenRG (talk) 04:39, 23 July 2012 (UTC)

Using generic non-lossy compression on already compressed sources (like mp3, jpeg, mpeg) won't give you any discernible improvement and may add size due to overhead. Shadowjams (talk) 21:31, 23 July 2012 (UTC)

The Connection Machines

Is it possible to do a quick comparison of the Connection Machines to modern day technology? Is a laptop or smart phone able to match these devices from the 80's? — Preceding unsigned comment added by 92.30.155.250 (talk) 14:23, 22 July 2012 (UTC)

Benchmarks aren't easy to meaningfully compare over generations. But we can see that FROSTBURG could perform around 65.5 gigaflops; for a modern comparison, check the GFLOPs column of the table in GeForce 500 Series. -- Finlay McWalterჷTalk 14:50, 22 July 2012 (UTC)

Gigaflops#Hardware costs contains some interesting metrics of cost-per-GFLOP. It doesn't have a CM, but does have a broadly comparable Cray X-MP. -- Finlay McWalterჷTalk 15:14, 22 July 2012 (UTC)

Also worth saying: although there were compilers for other languages, the only meaningful language Connection Machines understood was a weird variant of Lisp. So, really, unless your task is well-suited to be implemented in Lisp, it's not really even fair to compare gigaflops to gigaflops. I would surmise that a Connection Machine would perform very poorly at decoding an H.264 video, such as a Youtube clip, in real-time - even though that task is both parallelizable and highly limited by numerical calculation performance; whereas your modern mobile digital computer can do that on battery power. You could, in theory, try to code up a modern video toolchain for benchmarking on such a machine - it could even be ported to Lisp... but that is prohibitively difficult for a casual comparison. The thing to keep in mind is that for many specialized applications, modern computers get a lot of their performance from application-specific CISC-like hardware acceleration; so gigaflops of video decoding aren't comparable to gigaflops of wave equation calculation. Nimur (talk) 21:26, 22 July 2012 (UTC)

ring device from script

I have a "Voip Voice V653SK" [1] device with no manual or software. When I plug it into the computer via USB it recognizes it as a microphone. I want to make it ring loudly (using the devices built in ringer) on command from a batch script, but I cannot work out how to do this. Any help would be greatly appreciated. Thank you 92.233.64.26 (talk) 15:41, 22 July 2012 (UTC)

Here's the quick start manual: [2]. It requires the CD, though, so you will need to find that. StuRat (talk) 18:58, 22 July 2012 (UTC)

Thanks but that only explains how to set it up to take calls, which isn't want I want to do. The drivers for it have already been automatically installed so the CD isn't needed. 92.233.64.26 (talk) 19:10, 22 July 2012 (UTC)

That's not necessarily so. The phone advertises itself to the OS as a "composite device": a collection of audio devices and a single human-input device (the keypad, which I imagine also includes the on-hook detector as a "key"). When you insert it into a modern Windows machine, Windows recognises the speaker and microphone on it and adds generic sound drivers to these (there's plenty of info in the USB device enumeration for this). But the keypad (a "HID" device) isn't a generic device (the different keys map differently on different phone keypads) and the concept of a "ringing" channel isn't generic either (some devices have a "ring" command which buzzes a piezoelectric sounder, others have another audio channel into which a "ring ring" tone must be played, and one or several MIXER and SELECTOR units which control where the audio goes). So with the generic driver you should get basic phone function (speak and hear) but the keyboard will be dead and the phone won't ring. Ideally you'll find a decent driver that supports both - but my experience (a few years ago) with a similar model from the same manufacturer was that their driver was poor. If memory serves (it's been 4 years since I plugged it into a Windows computer) the driver could get the keypad working and the ring-lights flashing, but I don't remember it actually making a ring noise. -- Finlay McWalterჷTalk 19:44, 22 July 2012 (UTC)

The keypad seems to work with the generic drivers, the top lights up blue when I press the answer key and I can turn the microphone on and off via it (maybe that's an internal function of the device though, and not being processed by the computer). It doesn't seem like there is an easy way to achieve what I want, so I guess I'll just give up with this idea. Thanks for the replies anyway 92.233.64.26 (talk) 19:54, 22 July 2012 (UTC)

How can I quickly and easily get to the first comments posted on this 'New Yorker' article (or others like it)?

The obvious link is to the newest comment but I would like to follow the debate from the oldest.

80.90.168.166 (talk) 20:34, 22 July 2012 (UTC)

The abomination that is http://www.newyorker.com/js/pluck.min.js injects ten comments at a time. If you figure out the function that's being called, you can do pretty much whatever you want, including swapping "Sort":"TimeStampDescending" for "Sort":"TimeStampAscending" and start with the oldest comments first. ¦ Reisio (talk) 02:05, 23 July 2012 (UTC)

I don't know how to even start doing that, but at least I've learned there isn't a way in the normal user interface. I presume your assessment of the 'New Yorker' is unrelated.80.90.168.223 (talk) 04:19, 23 July 2012 (UTC)

Merely commenting on the terrible JS they use. ¦ Reisio (talk) 04:34, 23 July 2012 (UTC)

Learning C++

I am nearing the end of my first book on C++ (C++: A Beginner's Guide, Herbert Schildt) and wish to advance my studies still further. Can anyone recommend a suitable second book? Thanks. asyndeton talk 21:52, 22 July 2012 (UTC)

Good third books are Koenig and Moo, Rumnations on C++, and Cline et al., C++ FAQs. I'm not sure about a second book.

Oh, wait: how about Meyers, Effective C++? —Steve Summit (talk) 00:48, 23 July 2012 (UTC)

You might consider starting over with another first book, based on how well known Schildt is at being wrong. ¦ Reisio (talk) 00:55, 23 July 2012 (UTC)

Yes, Schildt's books are terrible. Please try to forget everything you "learned" from this book, as a large fraction of it is not true at all. I think Steve Summit's recommendations are decent. Even if you don't like the style of those books, the authors do at least know C++. -- BenRG (talk) 04:22, 23 July 2012 (UTC)

I find this news to be rather unfortunate. I have just started learning C++ from Schildt's book. I'm pretty near the beginning though and this is something I would very much like to get right the first time round if possible; could someone please recommend an appropriate first book on C++ (for an almost complete novice programmer)? Thanks. meromorphic [talk to me] 21:33, 23 July 2012 (UTC)

• Well, the only way to advance your studies that really works is to start writing actual code. Looie496 (talk) 01:55, 23 July 2012 (UTC)

I highly recommend Code Complete ed. 2 by Steve McConnell (Microsoft Press ISBN 978-0-7356-1967-8). It is an amazing book, it is not language-specific, and assumes you already know how to program. It teaches you how to program properly, which in my personal experience is a rare skill. I've read it cover-to-cover, and it's worth re-reading every so often. Our article on it does it no justice, but it's a fantastic book which changes the way you think about programming. I really can't emphasize enough how much it's helped me improve as a programmer. Beg, borrow or steal a copy so you can see for yourself, and then go buy it. After that, if you want to deepen your programmer's toolbox, I recommend learning Design Patterns by the gang of four. It's on a more sophisticated level, but it makes an excellent reference even if you don't get a lot of immediate value from it. Pretty much anything by the GoF is worth looking at. BigNate37_(T) 04:45, 23 July 2012 (UTC)

Let me make clear that my recommendations, while written on topics which are important for any seasoned programmer to know, are not C++ references. If you're not finished learning the ropes, McConnell is still worth a read but Design Patterns won't be helpful until you've got some object-oriented programming experience already. BigNate37_(T) 04:52, 23 July 2012 (UTC)