A thread for boring computer questions.

When you choose random shuffle playback it seems that it takes a while for the randomness to cover the full range of content. Is this the way random number generators work or is this just a problem that I got wrong on a problem set?

"Shuffle" as most commonly implemented will just play all the songs in a playlist in a random order, without any weighting or bias to play an "even distribution" of the songs inside. If you have, to use an extreme example, 99 songs by The Beatles and one song by the Mormon Tabernacle Choir in your library, the Mormon Tabernacle Choir song is equally likely to be in any one of the 100 play order positions, so you wouldn't be surprised if you have to listen to over half your Beatles songs before you hear the Mormon Tabernacle Choir (since you'd expect that about half the time). So in a more realistic scenario, yeah, it's entirely possible that you'll hear multiple songs from one artist or album before you hear anything from another. If you have more of one thing than another thing, there are more orderings where you hear a bunch of the first thing in a row than a bunch of the second thing in a row. Generally though any pattern you discern in your shuffle mix is confirmation bias. (Humans are pretty bad at coming up with or identifying random numbers, or making random choices. There's literature out there on this.)

Ahh -- I should have realized there were too many Pooh Sticks songs ... but even within the same album I have the feeling that I can predict how things will jump around from some seed number and that it takes a while for the trace of the seed to go away ... if only I had a very long album or boxed set even ... sorry for speaking so unscientifically about this!!

Also, with short link generators, how do they guarantee uniqueness each time, and can you always make them some fixed length?

00001
00002...

That's 100,000 right there

Allow upper and lower case alphabetics and you get 62 to the power 5 combinations, about 916 million using a length of exactly 5 (like bit.ly)

They don't dish them out in order, no. There's probably a database behind the scenes (needs to be to store the target URL) and they pick a random unsigned value each time.

Unassigned

I work at a library and I want to get the IT department to implement an independent identifier system not dependent on any other system. I was told various things by various people, but the simplicity and effectiveness of what you've described seems obvious. I can't imagine 916 million not being enough. Is a database of that size common and manageable on typical hardware?

Sure. Especially as each row probably only has three columns in it.

That reminds me: Google has a book scanning service that is scans out-of-copyright books and adds them to archive.org. it uses a similar scheme, uppers, lowers, digits. But if you try and Google one of their own ids it doesn't work because their search engine treats uppers and lowers the same.

https://archive.org/details/ladandlassastor00reevgoog - AjQOAAAAYAAJ

https://www.google.com/search?q=AjQOAAAAYAAJ&ie=utf-8&oe=utf-8&aq=t

So the program to fill in the first column would be something like nested for statements for each position that iterate over the 26 lower case and 26 upper case letters and 10 digits? Is such a program included in libraries of common programming languages? Would it take a reasonable amount of time to fill in the first column? And if you did not want your identifiers assigned in sequence, you would use a random number generator to pick an unassigned identifier between allowable values in the first column, then once assigned add the status to the second column and the URL to third?

It is good to know that users want to search on these identifiers ... (Thanks!!)

competing ways of doing this, as usual - do the work upfront or do it when you need it. the first is like a bag of lottery balls. as you remove them the bag gets emptier and emptier and every ball you pull out is guaranteed to be a new number. but you've needed to create 916 million lottery balls ahead of time.

the second case, just think of a random number. then check to see if you've already used that number. if you have, pick another number and repeat. if not, add it to your list. the check gets harder as time goes by (more hits, so this works better if your keyspace is much larger than your number of items), but you're spared the pre-creation task.

as for creating the random strings, it's pretty trivial in any programming language:

String str = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789";
String s = "";
for (int i = 0 ; i < 5 ; i++) {
  s += charAt(Math.random() * str.length);
}

(yes, StringBuilder(), i know)

(there's generally a base 64 encoder but that uses 0-9 a-z A-Z and two punctuation characters (_+/- it differs) which complicates things.)

Ah -- it makes much more sense to make the character selection the random part, particularly if you're not going to be generating the numbers beforehand, which should be fine. This is great. Now I think I have enough information to ask about it and to respond constructively if I'm told it can't be done. Thanks!

Never forget the adage that if a software person tells you something can't be done what they mean is it's not fun to do.

the generation side of this is probably tiny compared with marking all the books with the generated numbers.

Is multi-threading generally more efficient than multi-tasking (running multiple instances of the same process)? I am running a program that in the latest version is supposed to run multi-threaded if multiple cores are available. A version that does not run multi-threaded was already installed on a server that I was eventually allowed to use for the task. I am wondering if it is worth asking the system administrator if the latest version could be installed.

Generally a program which can take advantage of multiple cores will be able to do more in less time (assuming that the program can run multiple tasks in parallel), so yes, multi-threading is better.

I have both an iPhone and am Android tablet. The phone has 4G data; the tablet does not.

Is there any way to transfer files directly from iOS to Android without using the Internet, possibly via WiFi or Bluetooth? I mean, there must be SOME way... Right?

Also, the phone does not tether because AT&T blows

You'll need to download an app but apparently: http://www.techrepublic.com/blog/smartphones/send-files-between-android-and-ios-with-fast-file-transfer/

I saw that but that's only Android -> iOS

I've started using the free Microsoft Office Online: it has Word, Excel, and a bunch of other worthy stuff, but all I care about is music storage (15 GB free). Thinking about paid (the cheapest, $1.99 monthly for 100 GB), but will see how it goes for a while. Anybody use it for that, any problems? How does it compare with Google's Drive re music storage?

I hope it won't think promo package zips are boots.

Is there any way to transfer files directly from iOS to Android without using the Internet, possibly via WiFi or Bluetooth? I mean, there must be SOME way... Right?

I think that if you use Dropbox across two devices on the same wi-fi network, it skips the internet when syncing, but I'm not 100% on that.

Oh, yeah, here you go:

https://www.dropbox.com/help/137

sry i meant like wifi directly btwn the two devices, not using a separate wifi network

like if I'm on a bus or something

Ah. This seems to do that:

https://itunes.apple.com/gb/app/send-anywhere-file-transfer/id596642855?mt=8

Actually, I'm not sure it does. I will end my unhelpful Googling now.

xp when you say music storage, do you mean streaming OTG? Or cloud archiving? If the former, then Google Music is fantastic. You can upload 20k songs into the cloud and stream them anywhere, for free. I use it all the time, can't recommend it highly enough,

Thanks for the tip on GM! But mainly I meant cloud archiving.

dumb computer question: would it be possible/reasonable to build a laptop with two power sockets, one on each side? because this would revolutionize everything for me

power jack probably the better choice of word there

I've never heard of such a thing, no idea if it's possible, ever been done before, etc.

yeah i don't think it exists (yet!!) just wondering if it would be an awful idea hardware-wise

I don't know enough about electricity to tell you it'd be dangerous or difficult, so I assume the reason it doesn't exist is that you're the first person to ever want it.

it would make things much easier for people like me who spend 90% of their time lying down on different couches and beds

get an iPad tbh

Wait a couple years for wireless power transmission

hope you like that wireless power when it's burning a hole thru yr abdomen

dumb computer question: would it be possible/reasonable to build a laptop with two power sockets, one on each side? because this would revolutionize everything for me

― linda cardellini (zachlyon), Friday, October 3, 2014 4:00 PM Bookmark Flag Post

No

xp

Pretty sure I've irradiated most of my insides from being a computer jock for so long.

You know how on a PC you can upload an image found online to Facebook, etc. by copying the image's URL and pasting that URL in the address bar when searching for a file to upload? Can that be done w/ a Mac somehow?

not generally. save the image by dragging it to your desktop or whatever then upload it like a regular file.

I am looking at ILX using Chrome on Mac OSX, updated my Flash today and now there's just white space instead of a YT embed, can anybody advise on how to get them back?

1. Somebody help MaresNest (or maybe try this thread would be better for that Q Ideas for ILX)

2. Recommendation for a good file search tool for a Windows 7 computer?

3. Is there a "favorite hoodjadoojas" (sp??) thread for windows? (I guess I could search but I'm using Zing so I can't right now.)

Thanks!

PS: a 3-legged dog just woofed at me like it wanted to play.

Jesse the built in search function in Win 7 is p decent esp if your drive is indexed. Are you having probs with it?

Don't know if this is related, but Youtube just changed its default display platform from Flash to HTML5

http://www.theverge.com/2015/1/27/7926001/youtube-drops-flash-for-html5-video-default

chrome uses its own built-in copy of flash so i don't think updating flash should fuck with it? i also think youtube has maybe been serving HTML5 video to chrome as default for a while now

Thanks for your help guys.

I figured it out, I typed 'about;plugins' in Chrome and got the list up in screen, a quick disable/enable of the Flash did the trick.

Damn I could have answered that one, had that problem recently as well.

Handy to bookmark that plug in location, I can't see it in the options.

loved http://www.voidtools.com/ for years re win filesearch

be careful of having two things indexing and being a drag if ur set up's getting on

Looking for a solution to this

Somehow u-torrent seems to have reconfigured itself in such a way that instead of downloading audio files in folders thus separating the downloads into whatever live recording they came as they just come down as files stripped of the folders. I can't see what you would need to reconfigure to get it back to the way it was up until yesterday where it had the separating folders.
Subsequently I have several live sets in a larger folder all jumbled up, sorted by the numerical part of their title so I have a load of 1s from different sets, a load of 2s from different sets etc.

Looks like i somehow had automatically update u-torrent ticked on the preferences box and maybe this has the thing needing to be reconfigured since it is possibly a new version of u-torrent but I've never come across an update leading to this. I also don't think I would normally leave an automatically update box ticked if I was aware of it. So am further wondering if I did leave it unticked or bowed to continual notification previously that there were new updates which would probably bug me too, especially if they were happenning with any great frequency.

Jesse the built in search function in Win 7 is p decent esp if your drive is indexed. Are you having probs with it?

IIRC, it was not feasible to index the voluminous files on my work's networked drives, which is where I do most of my searching. Plus it doesn't do very advanced searching.

Also it arbitrarily returns "no results" sometimes but then later does find results.

Thanks, rtc. I will check that out.