Wikipedia:Reference desk/Archives/Computing/2014 November 3

= November 3 =

re-create your Youtube account
Is there a way that you can re-create your Youtube account so you can enjoy commenting and liking videos after years of your Youtube account being closed down by Google? — Preceding unsigned comment added by 70.29.32.180 (talk) 01:09, 3 November 2014 (UTC)
 * It would likely depend on why your account was closed. Nil Einne (talk) 01:34, 3 November 2014 (UTC)


 * http://www.wikihow.com/Reopen-Your-YouTube-Account ...has a description of how to re-open your account - but, as Nil Einne says, it all depends on why they shut your account down in the first place.   If you'd been violating their terms of service in some egregious manner, they may decide to forgive and forget - or they may not.   But if it was for some mundane reason - they probably would.   But at least there is a mechanism. SteveBaker (talk) 14:58, 3 November 2014 (UTC)

Linux software raid 5, 6 (with mdadm), stripe cache management and access to "updated" data
Hello everyone, after a bit of analysis i realized that one major factor to increase the write performance of a raid (5 or 6) system is the write cache (this because in most of the case the parity is computed by a cpu / specialized cpu way faster that the IO access on the disk, so the CPU should not be the bottleneck), in particular on linux software raid configuration, using mdadm, the values are set through /sys/block/md*/md/stripe_cache_size. If i understood it correctly, the cache allows to absorb "burst" of write operations in a transparent way for client system that uses the raid array, increasing the performances. Of course if the stream of writes is continuous and too big the advantage is gone (because the cache will be always full).

So far i recognized two risks: one is in case of power outage, the data in the cache will be lost, but this for now is an acceptable risk (since i have UPS units); the other risk would be less acceptable, would be the case that an application is writing a certain data and, immediately after, is reading it. If the writing data is still in the write cache, could be possible that the data read is an old value not updated since the "new value" is in the cache?

Thanks for the answers 91.66.224.104 (talk) 12:26, 3 November 2014 (UTC)
 * after a little search i guess that mdadm provides the writeback policy (that is, allows to read updated values from the cache) by default. Could someone confirm (my sources are forum discussions)? 91.66.224.104 (talk) 13:01, 3 November 2014 (UTC)

Bookmarking someones profile page on facebook
If I login to my facebook account from Chrome browser and bookmark profile page of a member of public from my Chrome browser will that guy somehow come to know that i kave bookmarked him and by what method or prcess will he come to know that.Also below the below the blue bar sometimes in some profile pages a white bar containing captions like Timemine occur sometimes permanently and sometimes intermittently why. Thanks. 117.194.244.68 (talk) 17:21, 3 November 2014 (UTC)

No. Facebook doesn't register profile views (or if it does it doesn't release it to the user) and similarly it doesn't advise when a page has been favourited. ny156uk (talk) 19:08, 3 November 2014 (UTC)


 * I'm not sure its even technically possible for a server to know when a page has been bookmarked. It may be able to make a reasonable guess when a user has visited a page from their bookmarks (based on the HTTP referer, or more accurately the lack thereof) but this would not be conclusive.   d avid p rior  t/c 19:41, 3 November 2014 (UTC)

Basic things that every computer scientist should know
I want to know what things are really basic like floating-point arithmetics, string representation and things like that. — Preceding unsigned comment added by Senteni (talk • contribs) 20:47, 3 November 2014 (UTC)
 * Have a look at Outline of computer science—there is a section on "Basic concepts". --Canley (talk) 00:13, 4 November 2014 (UTC)