Wikipedia:Reference desk/Archives/Computing/2010 October 14

Computing desk
< October 13 << Sep | October | Nov >> October 15 >
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.


October 14

edit

How to fix Windows 7 font

edit

I am having bitter experience in reading fonts in my W7 OS. After installing Microsoft Net Framework 4.0 (64 bit version)and some other software, the font shape suddenly changed and now it looks blur. Even fonts in different website are behaving in same manner. I checked Font folder in "C" drive and also in browser's option (Mozilla and IE8). All of them are in normal state. I couldn't fix them yet. How can I rectify this? I am thinking to reinstalling the OS. Thanks--180.234.32.4 (talk) 05:19, 14 October 2010 (UTC)[reply]

Try to disable ClearType. 180.11.188.56 (talk) 10:17, 14 October 2010 (UTC)[reply]

Update: I think the problem's been solved. I have found a solution here. Though, I had to turn on clear type font which brought font in normal state. Thanks--180.234.41.149 (talk) 17:25, 14 October 2010 (UTC)[reply]

Choice of data archive format

edit

I have an application that involves a large number of data files, which are grouped into sets. Each set of data files consists of about 1000 files, each of which is about 2 MB. At any time, about 20000 such file sets need to be easily accessible. When a file set is generated, the individual files are generated one by one until the full set is collected. What would be a good archive format for combining the individual files in a set into a single file object? Are some archive formats better than others? Since each file set is built up incrementally, adding a file to an archive should not be an expensive operation. The archive format should also allow easy random access to the files in a set. Any suggestions? My thanks in advance. --173.49.77.140 (talk) 11:33, 14 October 2010 (UTC)[reply]

By my calculations, that is 40 terabytes... which is very high end indeed. I don't think standard file archiving is going to help you very much. What is your budget for implementing a solution? An RDBMS like Oracle can handle that kind of volume with storing those files in LOBs (Large OBjects, or the 'single object' you are referring to)... but it's going to take pretty long to direct-load (incrementally and with simple compression) and your licensing is going to very VERY expensive. Oracle is excellent at reading high volume data with appropriate SQL commands. Then again maybe you meant 40 gigabytes which opens a whole lot of other cheaper and free options. Sandman30s (talk) 13:19, 14 October 2010 (UTC)[reply]
Most trivial thing is concatenate all the files together from each set (i.e. 1000*2MB=2GB), remembering the length and offset of the individual files within the set. So you have 20k sets of 2GB each. Depending on the data, you might benefit from compressing the individual segments within each set with libgzip or the like. It might also be possible to append to a tar archive without having to rewrite it. I'm not sure of that though. 72.89.106.242 (talk) 04:13, 17 October 2010 (UTC)[reply]

'Site Build It' / www.SiteSell.com

edit

What is "Site Build It" ?( company name = www.SiteSell.com ) Found Wikipedia page for Ken Evoy (founder of company / product) but nothing to explain "Site Build It". It is NOT 'software'- NOT a 'download' ( is IS a 'Subscription' e-Learning "Process" I think ?!?? ) It is NOT in your CMS 'computer management system' List - it is NOT an "entry" in ANY of your many 'Lists' or 'Comparables' - It IS a mystery ! Maybe, it IS a "Business Building System" or maybe a "Website Builder" or maybe a "Blog Maker". WYSIWYG - Yes - But also needs SOME HTML knowledge ? Cannot find it in any of your Lists of 'Editors'. Is there nothing COMPARABLE with it ? Is it a Web 2.0 application ? Or something else ? Would be very grateful indeed for your kind assistance, if you can help. —Preceding unsigned comment added by 83.67.57.245 (talk) 11:35, 14 October 2010 (UTC)[reply]

It looks like a small-scale consulting company to help people start websites for small businesses. Wikipedia has a guideline on notability, and is NOT a directory of every existing company, service or product. If you feel that this particular website or company is worth inclusion, you should read our notability guidelines and then you can add the content yourself. Make a strong case for notability, and write in encyclopedic style; otherwise, your additions may be reverted by another editor. Nimur (talk) 15:51, 14 October 2010 (UTC)[reply]

web cache

edit

I read the web cache article but I don't fully understand. Would sending all internet requests through something like Polipo on my own computer reduce bandwidth and increase speed, or are the benefits from web caches only noticeable on large scale implementations? 82.44.55.25 (talk) 14:02, 14 October 2010 (UTC)[reply]

The portion of content that is fetched more than once is going be the portion of bandwidth that is saved. This is a good application for Amdahl's law. You have P as the portion that is fetched more than once and S as the speedup you get from using cache.   is a measure of your overall speedup. By estimating values for S and P, you can see how much of a benefit you would get. -- kainaw 14:18, 14 October 2010 (UTC)[reply]
Thank you for the answer but I don't understand maths or equations. Could simplify the answer for me 82.44.55.25 (talk) 14:32, 14 October 2010 (UTC)[reply]
Basically, that equation (a very important formula in studying optimizations and speedups) says that the total speedup is limited by the frequency of the task you're speeding up. Or, more informally, "if you do something a lot, then improving the speed makes a big difference; but if you do something rarely, it doesn't matter how much you speed it up, because it won't make much difference." In your case, caching on your local computer will only make a big difference if you regularly re-load the same data. The cache can only deliver data is has already fetched - so if you request the same data again, the next request is faster. This isn't really very helpful for most single web users. Furthermore, a lot of ISPs are already performing caching for you, so you'd be duplicating their effort. Finally, keep in mind the difference between bandwidth and latency; both can be optimized by a cache, but in different ways. One advantage of a smart cache or web proxy is that it can prefetch - in other words, it starts guessing what pages or links you might click on next, and begins downloading them before you manually request them. This will noticeably improve your browsing latency (but adversely affect your bandwidth). I don't know if Polipo can do pre-fetching, but Squid with Prefetching definitely can. Squid (software) is a much more powerful, industrial-strength program; it is running in many enterprise-scale environments. It can be a little bit more intimidating for a novice, but is really the program to learn/operate if you want a web proxy or web cache. Nimur (talk) 16:00, 14 October 2010 (UTC)[reply]
Thanks for explaining :) 82.44.55.25 (talk) 19:06, 14 October 2010 (UTC)[reply]

outlook envelope on the clockbar of winxp pro

edit

I right-clicked on it and selected 'hide'. Is there a way to have it appeared back? t.i.a. --217.194.34.103 (talk) 15:33, 14 October 2010 (UTC)[reply]

Microsoft's Support site provides instructions - it is a preference setting. Consider using Microsoft's Bing Search with a query like "outlook show icon in task area" or checking the Outlook online help tool by pressing F1. Nimur (talk) 16:10, 14 October 2010 (UTC)[reply]
For future reference (and web-searching) the icons are in the system tray of the task bar. CS Miller (talk) 18:51, 14 October 2010 (UTC)[reply]
(Surprisingly, even though I used the wrong terminology, my search worked out pointing to the right place anyway - but good tip! Nimur (talk) 22:05, 14 October 2010 (UTC))[reply]
(The area of the taskbar that shows the small icons is correctly called the notification area but is often called the system tray so searching for either should be fruitful.) --Bavi H (talk) 00:43, 15 October 2010 (UTC)[reply]

Yet another C++ question.

edit

I'm sorry for all these annoying questions, but this one is quick. What's the difference between "Coordinates operator+(const &Coordinates);" and "Coordinates operator+(const &Coordinates) const;" ? KyuubiSeal (talk) 18:32, 14 October 2010 (UTC)[reply]

the const keyword after any member function (not just operators) will make *this a constant. Thus the function can't modify the object it works on. CS Miller (talk) 18:48, 14 October 2010 (UTC)[reply]
Ok, thanks! KyuubiSeal (talk) 19:21, 14 October 2010 (UTC)[reply]
The concept is explained in (excruciating) detail in our article on const-correctness :). Ideally, however, it should be neither. The canonical form would be
class Coordinates
{
  // ...
};

Coordinates operator+(const Coordinates& a, const Coordinates& b)
{
  Coordinates tmp(a);
  return tmp += b;
}

// or alternatively:
Coordinates operator+(Coordinates a, const Coordinates& b)
{
  return a += b;
}

In other words, implemented not as a member of Coordinates, but as a freestanding function. This enables implicit conversions for the left-hand argument. decltype (talk) 11:55, 18 October 2010 (UTC)[reply]