Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[methylatedspirit]

The motherboard's PCIe 16x slot doesn't seem to be fried; when I inserted the WiFi card into that slot, it did detect just fine. I tried another video card (a GTS 450), but that one seemed to be dead as well. Oh well, I guess that PC will be running on integrated graphics for a while. I'm waiting for Ampere and Big Navi to release later this year before buying another graphics card.

[Reelya]

Two different video cards both not working on the slot could point to something being wrong with the board. I wouldn't just assume that slot is working for graphics because it detected a Wifi card in the slot. So, buyer beware before spending money on a whole new video card in the hopes that that one works.

If you have a damaged video card there are several ways it can manifest. Firstly, the machine might just not boot. Second, it could boot but not register any card there. Third, it could identify the card but there are any of a number of things that go wrong with it. For example, Windows starts up but you get a bluescreen of death, since it tries to access the hardware for the additional screen but fails.

if both cards are doing the exact same thing with the exact same symptoms then that makes it a little less likely it's the cards and a little more likely it's the board.

[IonMatrix]

I just bought FarCry Primal on Steam, and Uplay is asking for a key, but steam said nothing about any keys, and all the solutions I searched don't work. I never futzed around with Ubisoft Games before, so, uh, solutions, anybody?

[Schmaven]

Ever since Ubisoft abandoned Heroes of Might and Magic 7 with a multiplayer that crashes 100% of the time, I've been boycotting all Ubisoft games. 

The key might be buried in one of the purchase confirmation emails they sent you.  Otherwise, I'd bet it's in your steam account somewhere. 

[wierd]

In the steam library, right click on the game, choose manage, then cd-keys.

It should display your key there.

[IonMatrix]

Hm, well I tried that and there was no cd keys option, but I already solved it on my own. Apparently if I launch it from the god damn uplay launcher directly instead of from steam then it's perfectly fine. HUH.

Even more interestingly, even though I have already played it, it STILL asks for a god damn key if I launch it from steam.

[Starver]

Here's an interesting one for people, I hope.

I'm using a regexp to match (and substitute out) quote-text-quote from the start of a string (in Perl), that in trivial use is:

Code: [Select]

$text=~s/^"(.*?)"//(I actually use s||| instead of s///, here, for reasons related to nearby bits of code, but it's the same.)
i.e from the start of string, match a quote, a group of as few as possible characters, then another quote. Replace with nothing. The group I found, between the quotes, is used elsewhere after the (valid) search and replace.

It needs to be an ungreedy *? because there's (usually) later ones to find and standard * would gobble far too much up. Could be further matched/replaced to stitch back in to the original, but that'd be wasteful.

Some if the quoted values are something like:

Code: [Select]

"Quoted text with \"escaped quotes\" within"Right now, I have no use for the escaped quotes, so I just pretreat $text to s/\\"//g, i.e. nuke them all before I start, but maybe I want them in the future. Or whoever else might pick up my data in the future.

As I just realised, by inserting [^\\] after the (grouping) and before the proper-close-quote I perpetually strip the last character off, and it cannot find "".
If I put it at the end of the group (or give it a (grouping2) to reconcatenate it back onto the end of where (grouping) is stored, it would save the character but still would not validly find "" whenever that pops up.

So, maybe a negative look-behind, then. (?<!\\) inserted before the true quote matcher. Except now I'm worrying how to deal with the (hypothetical) situation of an escaped-\ as the true last character in the true quote. Or indeed any number of escaped-\s (an even number of raw \s), which would mean that only an odd number of \s before a quote should disqualify it from ending the pattern, and {m,n} isn't that versatile. It looks like I'd need to zero-width-pre-not-match (\\\\)*\\ or similar, i.e. \\ times 0..n plus another \. Which (if it works as I think it should, haven't tested yet) would be slightly unweildy to put into the (?<!...)" bit and hardly good for future readability.

Anyone have any better ideas?

I'm almost tempted to slice out each character in turn (or, slightly quicker, in ungreedy blocks ending positively with [\\"] and flipping an when that m/\\$/ I then slice out the next character on a free-pass (whether \, " or whatever, doesn't matter), only if it's m/"$/ do I consider it finished (or unmatchable, which might mean restoring all the sliced-off partial-matchings just to show what didn't match the original intention).

There's also switching to greedy-grouping and coming back in from the /end$/ so long as that matches in the exact right way, but that sounds even more awkward (and prone to enbuggability).

But I'm sure there's a regexp (or, ideally, perlre) guru out there who has a perfecly wonderful alternative in mind.


(Until then, I'm prematching for \escapes of all kinds I expect to encounter, including \uunicode, to nuke or replace with best appropriate de-escaped and/or raw replacement version according to a separate lookup table, until it doesn't m/\\/ at all, then using the original matcher in perfect safety!)

[bloop_bleep]

This sounds like Programming Thread material.

EDIT: But anyway, you can just do this:

Code: [Select]

"((.*?)[^\\])"

and just extract the large group. (I don't know how Perl orders groups, but this would be group 1 in Python since its opening parenthesis is first.)

However, now if you escape the backslash, your regex will not recognize the end of string, as in this case:

"abc\\"

So you can do

Code: [Select]

"((.*?)[^\\](\\\\)*)"

allowing for an even number of backslashes before the end quote symbol.

What's lol is that if you were to write this regular expression as a string literal in Python without using raw strings it would look like this:

Code: [Select]

"\"((.*?)[^\\\\](\\\\\\\\)*)\""

That's 8 leaning toothpicks in a row, a truly rare sight. If you were to allow both single and double quotes in your input string, then you would not be able to use raw strings in Python, since you need to escape both Python string delimiters inside the string, so these would be necessary.

I wonder, what's the worst case of leaning toothpick syndrome you guys have seen? Is there perhaps 16 toothpicks running around somewhere in the wild? Maybe even 32?

[Starver]

This sounds like Programming Thread material.

Whoops, yeah. That's a thread I've got on permanent "must catch up on" mode, in my head, being horribly behind but unwilling to miss anything juicy, so not contributing either...

EDIT: But anyway, you can just do this:

Code: [Select]

"((.*?)[^\\])"


That doesn't match (quote)""(unquote), though. I did play with (quote)[^\\]?(unquote) after the sole ungreedy-anything group, briefly (I forget how it failed, but it did, and for 'obvious' reasons[1]), before I solved it the way I stated, maybe it'd work in the nested groups format. Or pre-test for ""s before I try the other (and the \\-proof extension)

PPE: [1] I remember now. It failed because the ungreedy .*? was allowed to be so ungreedy that the []? possible-single-match acted permanently as a [] always-single-matcher. So doesn't that mean it should have done exactly what I wanted it to do within the single, un-nested group? Straight after the ungreedy, where it doesn't matter that it's permanent except when not disallowing "" examples? And only still tripping over foo\\" examples to which we also may have the definitive answer by extension... Yes, that might be (the start of) the hack for this that I like most. Cheers. I knew there'd be something.

For the same reason I try to avoid excessive \\ing, I shy away from groups-within-groups, so I'd have to check for precedence/submembership conventions especially, after too long deciding I should avoid (foo((bar){3,4}baz)?)-type stuff where at all possible. But should be trivial to do, once I'm back on the Perl capable machine.

I wonder, what's the worst case of leaning toothpick syndrome you guys have seen? Is there perhaps 16 toothpicks running around somewhere in the wild? Maybe even 32?

Code: [Select]

  if ($testingText=~m/
         # ???On this line put the necessary regexp to uniquely match the known worst case usage, with {m} not allowed for 'reasons'???
      /xg) { print "Profit!!!\n")} #

[bloop_bleep]

Well, you can just add a question mark:

Code: [Select]

"(((.*?)[^\\](\\\\)*)?)"

Also, if you don't want groups to be captured you can just add ?: after the open parenthesis.

[AzyWng]

My laptop (which I’ve owned for a few years but isn’t that old) is having issues with the internet - occasionally the icon for wifi will just change to a globe with a no symbol, and say I’m conmected to the network but don’t have internet access.

None of the other devices my family own have this issue.

Anything I can do to fix this problem?

[Schmaven]

Does proximity to the router affect that at all?  If your family has less walls between their laptops and the router, that could explain it.

[AzyWng]

They’re all in the same room, and the router is fairly close by - no walls blocking me, at any rate.

[Reelya]

Try disabling then re-enabling the wifi in device manager, see if that helps. I have an external Wifi on my PC and it's recently begun to occassionally stop working. Unplugging the USB from it and plugging it back in again restarts it, but second best to that is turning off the device in Device Manager.

[AzyWng]

Disabling and re-enabling seems to work, but the device still stops working from time to time.

If there's some way for me to stop these issues, please let me know.