Computing desk
< September 26	<< Aug \| September \| Oct >>	September 28 >

Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

September 27

Javascript on the server side

Does the server side of a web application ever use JavaScript? If so, what percentage of web apps use JavaScript on the server side? 1%? Comet Tuttle (talk) 00:46, 27 September 2011 (UTC)[reply]

Yes, it is known as... wait for it... server-side JavaScript! I've no idea how prevalent it is. I would expect not very, because Javascript is in many ways a pain in the neck to use and not very optimal in many cases, but the article indicates that it is at least enjoying some enthusiasm with faster implementations. --Mr.98 (talk) 01:08, 27 September 2011 (UTC)[reply]

Don't sell JavaScript short! This article discusses some of its advantages for (non-web) programming (though the author only mentions it by name elsewhere). Server-side, it's certainly not popular, and it has its disadvantages, but I'd gladly take it over Java. Paul (Stansifer) 02:05, 27 September 2011 (UTC)[reply]

Server-side scripting is almost always limited to what the management has purchased. If the management purchases machines that require J2EE NetBeans for scripting, that will be used. If the management purchases a LAMP system, PHP will be used. The popular packages do not run server-side JavaScript, so management doesn't purchase them and they aren't used. If, instead, you are a hobbyist and building your own system, you can do whatever you like. I used a lot of different server-side things. Recently, I've been messing around with server-side CLisp. That's something I certainly don't expect to see on many servers. -- kainaw ™ 13:23, 27 September 2011 (UTC)[reply]

It is definitely possible, and have no sources, but I would put the percentage of web apps that use server-side javascript at much less than 1%. Probably 1 in 5,000.TheGrimme (talk) 13:51, 28 September 2011 (UTC)[reply]

Wireless network on Win 2000 PC

I have a Windows 2000, Service Pack 4 PC, with a network card (Ethernet), and 2 USB ports, (also a phone modem port, if that matters). I'd like to connect to the AT&T U-verse network in the house. The cable enters in the basement, and the PC is two flights straight up. So, how can I get both transmitter/receivers, at minimal cost ? I found a PC-end device that support Windows 2000, that's the Netgear WNCE2001. It seems to cost at least US$60: [1]. This price seems a bit steep, to me. Does the cable-end device (wireless router) care about the O/S, or can I just get any ? And please, nobody tell me to get a newer computer or switch to Linux (I have both, but want to get this computer to work online, too). StuRat (talk) 02:43, 27 September 2011 (UTC)[reply]

It's not completely clear to me whether you already have WiFi from your U-verse service, but if you do, I would expect that wireless device to work with any computer with an Ethernet connection. It appears to use USB for power, so you should either have working USB or an alternative power source. --Phil Holmes (talk) 12:21, 27 September 2011 (UTC)[reply]

No, I don't currently have WiFi. I have an Ethernet plug on the cable box in the basement, and want to connect from there to either the Ethernet port or USB port on the Windows 2000 computer 2 stories straight up. So, what WiFi hardware (other than the device listed above) supports Windows 2000, SP4 ? StuRat (talk) 15:47, 27 September 2011 (UTC)[reply]

query

what is SOFTWARE SECURITY SYSTEMS project Conclusion ? — Preceding unsigned comment added by 210.212.228.91 (talk) 10:47, 27 September 2011 (UTC)[reply]

iPod Touch Problem

My iPod has an extremely annoying habit of making this odd beeping sound whilst I'm playing a track and then switching to another track, even when the iPod is securely docked and on hold. It's pretty damn frustrating because it does it completely of its own accord. Does anyone know how to stop this? Thanks. 92.19.238.196 (talk) 10:56, 27 September 2011 (UTC)[reply]

It's the "shake to shuffle" feature; you're inadvertently moving the iPod in such a way that it counts as a "shake" (perhaps you're running or something). To disable this neat but fundamentally exasperating feature, go to "settings->music->shake_to_shuffle" -- 2.122.75.122 (talk) 12:20, 27 September 2011 (UTC)[reply]

File system to optimize SSD/HDD combo?

I have an old 500GB 7200rpm hard drive and am thinking of getting a smaller, faster solid-state disk drive to use alongside it. Based on two articles from ZDNet, I'm thinking the optimal solution for performance would be to have most of my swap space and smaller, more frequently-used files on the SSD, and most of my larger or rarely-used files on the HDD. I'm also thinking when two files are frequently accessed together, it'd be good to have one on one volume and one on the other to provide some parallelism. Are there any file systems out there that can accomplish these goals automatically while virtually merging the drives into one? CPU-cycle overhead probably won't be an issue. Neon Merlin 11:57, 27 September 2011 (UTC)[reply]

On which operating system do you plan to access the disk(s) most frequently?

One option is to set the entire high-performance disk (the SSD) as your swap file or page file (on Windows). This will allow the operating system scheduler to decide when a file or other data should be placed on the SSD; the entire process will be completely transparent to you, except as a measurable performance increase compared to accessing the file from the hard-disk drive. Modern operating systems cache commonly-used files in their swap / page space, so even file-access will appear to be accelerated. Nimur (talk) 16:12, 27 September 2011 (UTC)[reply]

I'm skeptical that the OS would keep files in swap space (but feel free to prove me wrong). Unless the user is in exactly the situation above (main FS is on a magnetic disk, and swap space is on an SSD), putting files in swap space is going to provide only a tiny bit of performance improvement in the aggregate. The OS designers would probably prefer to spend limited swap space on actual applications, rather than on duplicating files that already exist on disk. Paul (Stansifer) 17:33, 27 September 2011 (UTC)[reply]

The OS precaches files in physical memory; and if physical memory is needed in a hurry, the disk pre-cache is either (1) flushed back to the file-system; or (2) swapped. Either can happen, as long as the result is guaranteed to be consistent.

Here are some technical articles: A Tour of the Linux VFS; File Caching (Windows). Any time you have unused physical memory, (e.g., RAM that is not 100.0% utilized), many operating system prefetchers will predictively read random files out of the file-system and cache them in unused physical memory "just in case" you want those files later. On most Windows, this is performed by the System Cache Manager; part of the NT kernel. On Linux, this is handled by VFS, part of kernel kernel/fs/. Other major modern operating systems use similar strategies. Some operating systems "differ;" but it's all the same conceptual idea. RAM caches the file system; swap backs up the RAM; sometimes, due to inversion, swap (on disk) may be used to store contents of RAM that was already being used for caching disk files. This is not even "undesired" behavior. In the case where /swap is faster than another file system, this is preferred behavior; proper, bug-free kernels avoid priority inversion-esque thrashing.

When physical memory ("RAM") is full, on Windows, the System Cache will be paged to the pagefile. On Linux, the VFS is part of the kernel, so parts of it will never swap out of physical memory; but disk cache will swap out as soon as it's needed. Depending on the timing of events, and depending on your kernel, and your file system driver, this swap-out could fall back to .. /swap... or it could just drop to the file-system driver. Your behavior may "differ."

An SSD is not always a better/faster disk than a magnetic disk. The performance depends on your access patterns - sequential access may be faster on a high-performance magnetic disk; but random access is often faster on an SSD. Nimur (talk) 18:53, 27 September 2011 (UTC)[reply]

Caching doesn't work that way, at least not in a well designed modern OS. Every page of paged RAM is backed by some file (or device), either the swap file or some ordinary filesystem file. The swap file doesn't back every page, only pages that have nowhere else to go. Raw file contents might end up in the swap file if, for example, a user-level process implemented its own caching instead of using the OS caching (which is a bad idea for this very reason). But it won't happen as a matter of course.

ReadyBoost is designed to take advantage of a HD+flash combo by caching file data on the flash drive, but it seems to be primarily for smaller-capacity flash drives (~4GB). I don't know how well it works with a 40+GB SSD. -- BenRG (talk) 21:41, 27 September 2011 (UTC)[reply]

2000.00 investment in Roth IRA

If a $2000.00 initial investment in a roth IRA. What would it return in 2021,2031,2041,and 2051. With no additional add ons. — Preceding unsigned comment added by 98.211.11.106 (talk) 18:37, 27 September 2011 (UTC)[reply]

That would depend entirely on the rate of return. If we assume a 5% rate of return, compounded and accrued annually, you would use the formula 2000.00×1.05^Y, where Y is the number of years since you bought it. For 2051, that would give you $2000.00×1.05⁴⁰ = $2000.00×7.04 = $14079.98. (On a standard calculator, type 2000 × 1.05 x^y 40, then hit =.)

Of course, you really should consider inflation, which means that $14079.98 won't be worth anywhere near what it is now. The way to account for inflation is to subtract it from your interest rate. So, if you figure a 5% return, less a 3% inflation rate, that gives us 2%, and we plug that into the same formula: $2000.00×1.02⁴⁰ = $2000.00×2.208 = $4416.08. This means that your $14079.98 will only buy as much then as $4416.08 buys now. What the future inflation rate will be is conjecture, at best. The same may be true of your annual rate of return, unless you have an investment with a guaranteed rate of return. StuRat (talk) 19:18, 27 September 2011 (UTC)[reply]

StuRat - I think you just did someone's homework for him. Inter change able|talk to me 23:07, 27 September 2011 (UTC)[reply]

Note that I only did 1 of the 4 time periods, to show them the method, and left the other 3 for them to solve. StuRat (talk) 17:11, 29 September 2011 (UTC)[reply]