Also review audit. HALF THE QUESTIONS I REVIEWED WERE AUDITS. HOW AM I SUPPOSED TO CULL THE CLOSE QUEUE!!! :/
You should be backing up before failure occurs. Waiting for a device to die, then backing up is too late. Don't forget, that is only one failure mode. You may have controller failure, or random ferret attacks making your SSD unusable as well — Journeyman Geek ♦54 secs ago
Today is saraswathi pooja (so, I popped a few forensics books, rather than what I most recently studied. In a few years though, I wonder if people will just stick a tablet there ;p)
> Error listening: dial tcp xxxxx An operation on a socket could not be performed because the system lacked sufficient buffer space or because a que e was full.
oh. well you know i have had that teeny tiny ram corruption problem (that f--s up everything) I got a new board, let it test for 2 days (with cheap chip) and all was fine. Assembled it in my present computer, everything was fine. installed all my stufff, and Well Gol Damn it , it is back.
So the regular hardware monitor that came with the board, says the voltages are ok (the ones it shows). but none of the monitors show a proper -12 and -5
but they have not done that for years now anyways.
It really sucks cause i have proved it is not the memory (or did i) I have replaced the board, although every board made seems to be pressed out by the same factory in china
What is left is the memory, the CPU (which is where the ram controller is) the ram, the video card, and the PSU. Well the f everything i have not yet replaced.
everything had gone along fine for about 1 year, the error(s) are so freaking tiny i could sell it and nobody would even know, but because all storage routes through it , it is endless problem doing things like who backups of TBs
I shot off to the local tech store to grab a null modem cable as that would work as well... but they were just closing when I got there and didn't want to let me in.
@JourneymanGeek I was reluctant to go for 7 as well but later, I was happy to use it, am having a licensed copy of 8 but am still using 7, cuz as a dev, 8 is like pia
am feeling that by releasing w8, they forgot about the serious users and went funky when os is first meant for work instead of enjoying, you cant have a common platform for mobiles, tables and computers
somone should run out and snag the WindowsX-Files domains :-) Windows X-Men (man) and Windows XXX (yea for porn)
with so many people running off to linux the EX-windows , and windows-EX might be valuable too
spanish site MeX and MSeX
Well i replaced the cip with the cheap "test" chip, and the memory is still buggered, one more down. Thinking about replacing the PSU , eventually i will replace the memory :-) after all it is a memory problem (although the memory all tests out)
Yeah, I just have a quick question as I've specified a veeery specific OU in the query, but it appears to be ignoring that and pulling machines and users alike from every OU which is a bit annoying.
superuser.com/a/819748/98855 <--- reading this answer , half way through I am thinking WTF does that have to do with it :-) Then i see it is a bit spammy.
@allquixotic now you have until their next attempt!
Oh yea, how did the optimisation go?
> IF YOU LIVE IN THE UNITED STATES, SECTION 13 CONTAINS A BINDING ARBITRATION CLAUSE AND CLASS ACTION WAIVER. IT AFFECTS YOUR RIGHTS ABOUT HOW TO RESOLVE ANY DISPUTE WITH MICROSOFT. PLEASE READ IT.
@Bob I nicked a DoublyLinkedList implementation from a gem I found (and read the source; it wasn't implemented using an array), and it cut down the runtime by about 5 minutes, but that's it
the change proved functionally correct but not immensely faster
at this point I'm probably just going to bite the bullet and accept the 25-minute runtime... this program doesn't need to be run frequently enough for it to matter
currently the plan is to run it every week at 5 PM on a Friday (everyone should be gone by then, since everyone leaves early on Friday)
and when I run it on a smaller dataset (which is going to be necessary a bit more often than weekly), the runtime is about 2-3 minutes
there seems to be some nonlinear factor involved in running Tarjan's iteratively after each add to the tree, it's probably quadratic
@allquixotic when I bought a phone like, 3 generations ago, Motorola provided a included aka complimentary but not free 32 GB MicroSD card in the phone (removable) (source)
@Bob now that you mention it, there is; but it would be too labor-intensive and brittle to make that relationship in the code, so I'd have to come up with a way to iteratively run the program on a bunch of small datasets
the way the data works is, there are several hundred test cases that are "Common", which are called by (aka linked to in the digraph) thousands upon thousands of other test cases... but the non-Common test cases rarely, if ever, link to one another
what I could do is run the tool on little bits and pieces of the non-Common scripts (concatenated with the Common scripts, of course), producing a digraph with around 600 test cases at a time, instead of a digraph with 7000+ test cases at a time
if it's O(quadratic or cubic or exponential or something distinctly non-linear), running it 25 times on digraphs with <1000 nodes will be way faster than running it once on a digraph with 7000+ nodes
in fact most of the digraphs in that case would be <600 nodes
if it's quadratic, adding node 6999, doing Tarjan's, then adding node 7000 and doing Tarjan's again would probably be as much computation as running Tarjan's 600 times to build up the first 600 nodes
@Bob splitting it up would allow me to use multiple threads, yes. currently I don't think it's possible to break up my existing algorithm with multiple threads.
actually, the quadratic behavior comes from the fact that I'm iteratively running the linear-time Tarjan's algorithm... duh; why didn't I think of that?!
the runtime would be something like little-o(Summation from 1 to N(|V| + |E|)) which would be an arithmetic series: (N/2) * ((|V1| + |E1|) + (|VN| + |EN|)), where |V| is the number of nodes, |E| is the number of edges (which in my graph is always equal to |V|-1), and |VN| is the number of vertices when adding node N
@Bob: few reasons, firstly, they probably 1. Know they can do that in the US 2. Are more likely to be sued in the US than elsewhere. 3. May get in trouble if they did it elsewhere, which is precisely what they want to avoid
> In the dawn of time, the original Windows console was created. For millenia, geeks and developers (typically both) steeled their nerves and leveraged the console’s utility. After a (very) short time, a great lacking was noticed. And so, legions of ever intrepid command line mavens migrated to other platforms and other consoles, ‘til but a stalwart few remained. Now, today, with a small rumbling, the ancient, weary console heaves a great sigh and rises. The journey to legitimacy has begun. WE’RE BACK!
I get that binaries can be in different places. What I don't understand is why when I run whereis, which and type it says vim is at /usr/bin/vim, but when I try to run it using sudo it says I can't run /bin/vim.
[user@host conf.d]$ sudo vim test.conf
[sudo] password for user:
Sorry, user user i...
But Linux is ignoring /usr/bin/vim and preferring /bin/vim (which I have no permission to sudo). Despite which AND whereis AND type saying vim resides at /usr/bin/vim!
I get that binaries can be in different places. What I don't understand is why when I run whereis, which and type it says vim is at /usr/bin/vim, but when I try to run it using sudo it says I can't run /bin/vim.
[user@host conf.d]$ sudo vim test.conf
[sudo] password for user:
Sorry, user user i...
