Wikipedia:Reference desk/Archives/Computing/2012 July 8

Computing desk
< July 7	<< Jun | July | Aug >>	July 9 >

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 8

What Android rooting & data-transferring app would you most recommend?

Hi, I have an issue with my Xperia Play only having 380 MB of internal space. I was told that to move EVERY LAST BYTE to the SD card, including phone OS information, I would first have to "root" the phone.

It would void a warranty, but the standard warranty (at least through Verizon and Best Bought phones, is just one year, and I obtained it in May 2011.

Now, I need to find out what phone-rooting and data-transferring app (both functions in one, preferably) you would most recommend.

How well has it helped you?

Moreover, I may pay for it if no other app does it better, but free ones take precedence. Thanks in advance. --70.179.170.114 (talk) 00:09, 8 July 2012 (UTC)

Have a look at this guide. As for copying the file over, I tend to use Remote Web Desktop. There are a number of apps like this. But I'm not sure if you need to have a rooted device to simply copy the files over. You certainly won't need it to copy the SD card contents over. I would try using the Remote Web app to see if you can copy everything over before rooting. - Akamad (talk) 22:56, 8 July 2012 (UTC)

Note that you may want to look in to what you are doing first. AFAIK, the norm with most Android phones is simply to put the internal data partition (/data or sometimes even just /data/apps) on the SD card. The actual OS etc is still stored on the phone's internal flash. Since it doesn't change much, except if you upgrade firmware and anyone making firmware will generally make it fit on the flash, I can't see much advantage to what you have proposed and while removing the SD card while the phone is on even if it's only the data partition is likely a bad idea, what you're proposing seems to guarantee a crash if you ever do remove the card while the phone is on and perhaps could even create problems after you put the card back. At most, perhaps it will make sense to move the swap partition as well. Nil Einne (talk) 20:02, 9 July 2012 (UTC)

'Replay' function on YouTube

YouTube has a button (play/pause/'rewind') on the left of the 'toolbar' that allows a replay of a just viewed video. This used to mean that after it had downloaded, on pause for example, (which I often do because my wireless download (only≈30 kilobytes/sec) is not fast enough to watch most vids in real time) I could click that button and it would play/replay without downloading again.

Now it seem that if I try this, then instead of replaying immediately (from a copy apparently cached on my PC) it starts downloading the entire vid again. I was also able to move the 'progress cursor' back to re-watch certain parts, now it also tends to reload again rather than playing what I have already downloaded.

I have plenty of download capacity, it is just that the waiting (& re-waiting) is annoying (5 min. video takes about 15 min. (or more?) to download)

My laptop PC OS is Windows 7 Home Premium 32 bit. About 10 days ago I downloaded/installed a lot of Windows updates. 2.74 Gb of usable RAM (900Mb available), 226 Gb HDD space. Intel Core i3-2310M CPU. Browser I was using was Firefox 13.0.1

• So my question is: has YouTube changed something so that I if I let the entire video load and play, then I must download again to watch it again? (I just watched a relatively short vid. (1:38) and it replayed without trouble, using Firefox browser. A longer vid (4 mins) would not replay w/o downloading again. Caching problem on my PC?) - 220 of ^Borg 06:06, 8 July 2012 (UTC)

YouTube actually has two different implementations at the moment, one using Flash, which has always had its problems, and one using native browser video playback support, which has always had its problems (but is less tested). If you want reliability, you should probably investigate using a 3rd party plugin (like VLC's maybe), or use something like DownloadHelper to simply download the video normally. ¦ Reisio (talk) 15:33, 8 July 2012 (UTC)

Based on some tests, I think it's YouTube's Flash video player. More information below.

In the past, YouTube's Flash video player downloaded one long video file in the background. You could visit a video page, start playback, then immediately pause the video and wait for the playback bar to indicate the video had completely loaded. Currently, the player loads several background files in segments. If I go to a video page, start playback, then immediately pause the video and wait for the video to load, only a segment of the video gets loaded. The next segment is only loaded when the play position is near the end of the current segment.

After reading this question, I performed some tests to see if a video is replayed from the browser cache. I played a 6-minute test video, then clicked the restart button at the end to play it a second and third time. I monitored my browser cache and kept track of the URLs of each segment downloaded as well as keeping a copy of each segment. On the first play, 13 segments of 1,781,760 bytes or less were downloaded. The second and third plays downloaded 7 segments each, and seemed to use other segments from the cache.

Each segment of the video is loaded from a URL of the following form

http://really.long.subdomain.youtube.com/videoplayback?n1=v1&n2=v2&n3=v3&...&n21=v21

The part after the ? has 21 parameters, each one in a name=value format, separated by &s.

Unfortunately, YouTube's Flash video player seems to specify the parameters in a different order from time to time. The browser cache thinks the segments are different because the URLs don't match exactly, but the URL parameters and actual downloaded data usually match an existing segment in the cache. This seems to be the main reason the player doesn't reuse existing segments from the cache.

The main parameter that changes is range=begin-end, where begin and end seem to indicate the byte positions, inclusive, from a master video stream that YouTube has stored for the video. (The size of the downloaded segment is always end + 1 − start bytes.) The only other parameter that changed during my test was cm2=1 was used for the first segment of the first play, and cm2=0 was used for all other segments. I don't know what this parameter means. All of the other parameters were the same across all segments.

On the three plays, only 1 out of 13 segments had a slightly different range between the first and second play. The other segments had matching ranges. Even though the the ranges were the same, the first segments of each play had 9 bytes near the beginning of the downloaded data that were different each time. All of the other segments with the same ranges had identically downloaded bytes.

So the main reason the segments are redownloaded seems to be YouTube's Flash video player specifies the URL parameters in a different order from time to time. Other less frequent causes might be the player specifies slightly different segment ranges on a subsequent play, or requests other parameters with different values.

I mainly tested with Internet Explorer. I tested again with Firefox and the progress bar behaved similarly so I suspect it's doing something similar, but I don't have a good tool to monitor Firefox's cache. As an alternative, I disabled Firefox's Flash plug in, and then YouTube used the HTML player. The HTML player downloaded the video in one background file. When I restarted the video the progress bar still indicated the video was downloaded, suggesting it reused the cached file.

--Bavi H (talk) 01:31, 10 July 2012 (UTC)

I'm Really curious how did you monitored the cache. 65.49.68.173 (talk) 16:27, 10 July 2012 (UTC)

I used NirSoft IECacheView.

Basic procedure: In IECacheView, I sorted the cache by Last Accessed time. In Internet Explorer, I cleared the cache (Tools, Delete Browsing History, uncheck everything, only check Temporary Internet Files, click Delete.) Then I went to the test video page, played the video and periodically refreshed the IECacheView window. I could occasionally see new items in the cache as the new segments were downloaded. In the IECacheView Edit menu, you can copy the text data from the table of items to further analyze the URLs, and you can copy the actual cached file (Explorer Copy) and paste it to another folder to analyze if certain segments are the same.

NirSoft also has a MozillaCacheView, but it seems to only see changes to Firefox's cache after you close Firefox, so I couldn't use it to track the cache while Firefox was open. --Bavi H (talk) 03:31, 11 July 2012 (UTC)

Thanks!

Thanks to "Reisio", and Double+ thanks to "Bavi H" for making extraordinary effort to investigate the issue. {I still have to fully read and digest your posts Bavi H :-)
The life of a technical support person is not an easy one. Should I expect a bill now? ;-) }

• Re. Reisios' suggestion, I already have a downloader installed. Oddly the 'download' button has been appearing and disappearing!.
  
• I should add that as an alternative to specially downloading, like after watching a video and then wanting to keep a copy, it appears that viewed videos are stored as *.tmp files in a 'Temp' directory. I have found that all you need to do (after locating the file) is change the filename from 'foo'.temp to 'foo'.flv and it can be played with FLV player. Not sure if this applies to, or only to, Youtube as I have been finding 'cached' content from other sites I have viewed videos on.
  

Anyone with more to add are still welcome to chime in. - 220 of ^Borg 08:50, 11 July 2012 (UTC)

Iphone 4

I hear this (and a few other mobiles) are about to be obsolete. Is this true? Kittybrewster ☎ 23:18, 8 July 2012 (UTC)

Yes, Steve Ballmer has spoken: a windows phone, is now, the only phone to have. Almost as good as two paper cups strung between a string. </sarcasm>--Aspro (talk) 23:48, 8 July 2012 (UTC)

The ref desk is not the place for speculation. There are always rumors of new phones "coming soon." RudolfRed (talk) 00:20, 9 July 2012 (UTC)

Very silly people with more money than sense will call a gadget "obsolete" the instant there's a newer version available. To their mind, the iPhone 4 was obsolete the moment the 4S, or possibly the Samsung Galaxy, etc, was released. As far as actual functional obsolescence goes, I imagine the network technology which the iPhone 4 connects to will be around for at least a couple more decades. FiggyBee (talk) 03:56, 9 July 2012 (UTC)