steve_v

Perhaps I should have said "everything else that I have seen". My current arch nemesis (and chosen example) is AVG free, which certain people keep reinstalling after I remove it... Resource hog? Check. Nagware popups? Check. Scareware popups? Check. Browser toolbar: Check. Doesn't have ads though... Except ads for the paid version of AVG. So, adware: Check. As for creating security issues, if you keep up with the news on emerging exploits, there have been several nasty issues found in both free and paid AV products recently. Such as unpacking potential malware in kernel space with ring0 privileges, and laughably insecure update mechanisms. Then of course there's the utterly idiotic "feature" of intercepting SSL and messing with certificates. if you can point me at a free AV product with no nag (upgrade to premium) popups, no scareware (everyone can see what you do on the internet!) "alerts", no "PC tuneup" interruptions, and no browser toolbars - essentially nothing but unobtrusive antivirus, and one that doesn't do moronic things like flog the disk to death scanning a 100G VM image while I'm trying to work... Maybe I'll recommend that instead. AV is the ambulance at the bottom of the cliff anyway. You shouldn't be acquiring infected files to begin with.

RCS Build Aid all the way. Not 1.3 compatible yet though, AFAIK.

At least we're getting somewhere now. output.log.txt? Also, installing under C:\Users has been known to cause issues (Or so I have heard from those that use that Windows thing). Might be worth moving it elsewhere (not Program Files either) in case Windows is being stupid about permissions.

And all those mods are compatible with 1.3, are they? Incompatible mods will crash the game... The first step is to see if it still crashes with no mods installed. I looked around that site for about 15 seconds, didn't see an "add DMP" button, closed the window. If you can't be bothered listing your mods, why should I go hunting around for them? You are the one wanting help with a crash, no? Go read the "how to get support" sticky at the top of this subforum, then upload your output.log.txt (to e.g. Dropbox) and post a link.

Version? Mod list? output.log.txt? Crystal ball?

Must be all that coffee you Ubuntites drink over there. I'll stick to my calming fermented beverages.

Not so fast buddy. You'll need to change your sources.list to point at the new release first, or not a whole lot will happen... But don't do that anyway, read the post I linked on upgrading Mint, the procedure (mintupdate) appears very different to Debian. The alternative is to go hunting for backporting instructions or some kind soul that fired up a GCC 4.9 repo for trusty. You might get lucky here, but I don't know enough about Mint / Ubuntu to provide either. <grumble> There's no reason for Squad to build their library against GCC 4.9 libs anyway, it's not using anything new. If they had compiled it with 4.8 this problem would not exist, and it would work on a bunch more distros. </grumble>

That's a version mismatch, the library exists, but doesn't support the ABI version the application was compiled with. ABI version 1.3.8 comes in with GCC 4.9. Nothing, I would think, as nothing was installed. You just asked to install GCC 4.8, which is already installed and is the latest version available in Trusty / Rosa. You need GCC 4.9.

From that error message, it looks to me like we need GCC 4.9 libs here... They're not available in Trusty (and therefore Rosa), so if one really wants to stick with that release the packages will have to be backported. Otherwise, upgrade to Sara or Serena is the simplest answer.

It exists for Yakkety & Xenial, but not Trusty, upon which Mint Rosa is built, IIRC. Might be time to upgrade to Serena (Mint 18.1), you won't find those packages for Rosa. Looks like Mint mainline, not Mint DE. So nope, Debian package names won't work. Ubuntu ones should.

I play with much bigger electricities all day, 12v is pretty friendly really. Gotta take prospective fault current into account though, a 750W PSU will do ~60A on the +12v rail, and that's going to destroy something if you short it. I start to get nervous above 20KA PFC... Flameproof gear, long handle, all that jazz. It does surprise me how many people think electricity is some kind of voodoo, the "plug it in and magic happens" attitude is everywhere. Then again, I just can't stand not knowing how something works. Can't fathom how anyone could not be at least a little curious. From curiosity comes understanding, and from understanding, respect. Most of the time. Nah, a bit of magic smoke got out, nothing to worry about. It's still working fine, so there must be plenty more where that came from. Well, you certainly can weld with 12v @ 60A... but in this case it was more burning. Nasty smelly burning plastic. This I know well. A mains boot is nothing to what you'll get off the HT lead in a CRT. And those exposed tags in old ratsnest wired valve amps are pure evil.

I'd buy that.

Distrowatch down? Not for me.

Specifically? I'm not seeing anything noticeable... Pretty much the same as what I'm seeing, and exactly what I expected. I would still be running 1.1.3 for the performance, but can't because of the constant crashes... How's 1.2.2 vs 1.3? I'm seeing much of a sameness here. Pre-release: Me complaining about prioritising translations over performance improvements. Post-release: Me complaining about prioritising translations over performance improvements.

I see you have a video describing how to give microsoft like 20 dollars. Can I direct you to this website? I have tried it, and they are actually free...

Steam, Squad store or GOG? Procedure varies. For steam, it's under 'betas' IIUC. (I don't use steam), for the other two: Download the 1.2.2 archive, unpack it somewhere, and copy your saves over. There is no install / uninstall.

This thread is in unmodded support... Yet you clearly have a bunch of mods installed. Are they all 1.3 compatible? Anything that includes a .dll and is not compiled for 1.3 will almost certainly crash the game. Install 1.3 without any mods and try again. Or stick with 1.2.2, 1.3 is the most boring update I've seen yet. Aside, please upload log files and post a link, rather than posting inline, it makes it hard to read. Dropbox is the standard suggestion for a host.

I'd use something like 'rsync -a --delete --exclude 'saves' --exclude 'screenshots' <steam install> <non-steam install>'. Backup important stuff first, reinstall mods later (as they probably won't be compatible with 1.3) Because: Copy-paste is a clumsy tool. Or, if you're feeling brave, use --include / --exclude to keep your mods...

A file you will not be able to find, as you are not running Windows. From the modded support sticky (why this is not also in unmodded support I don't know): That said, that crash dump looks a little more serious than something easily diagnosed with Player.log... I'm not particularly familiar with MacOS, but it appears to be catching a SIGABRT - every one of those I've seen (on GNU/Linux) has been before the game loads far enough to create a useful Player.log. Upload it anyway, you might get lucky. If not, hopefully there's someone around who can read Mac crash dumps... But that's not me, unfortunately. On Windows, sure. You won't find that filename anywhere on *nix. You won't get a "crash folder" either, whatever that is. You might get a core dump though, if you have those enabled. Indeed. A useful test to see if it's something new in 1.3. No Unity upgrade this time, but AFAICT it does include a new native library.

For the moment I can't postmortem it properly, as it's in use. Another unit has been ordered, as not having a spare is silly, so I'll likely investigate the root cause when I have a go at fixing that board. First step is patching up the tracks it blew off the PCB... then it can be tested.

Sounds a lot like paging to me. Check your RAM / pagefile usage when this happens.

Fun story, my PC started rebooting unexpectedly today. Opened case, smelled dead electronics from the (very hot) PSU. Uh Oh... Opened PSU, saw huge burn mark on inside of case. Thought "Damn, it's Sunday and I don't have another PSU". Closer inspection reveals it's the fan controller (switching transistors & sense resistors had gotten hot, melted into plastic insulator, shorted to case). Ripped fan controller board out, brought fan wires out of case & jammed into FDD power plug... Powered it back up, and here we are. Yay for modular power supply designs and thermal protection. Boo for lousy sheet plastic case insulators. One little design flaw really messes up your day. Now for the old "do I fix properly, or just replace it?" question...

ckan.exe upgrade KIS=1.4.4 This really needs to be sorted, modders fault or no. CKAN is unusable if it is going to bork my install every time I allow an update. I can't think of an easily implemented technical solution, so it'll likely have to be: find incompatible update -> poke author. What would be nice is some workaround functionality in CKAN e.g. force version and hold version in the GUI. Installing the correct version manually from the CLI is fine (though upgrade <identifier>=<version> is still not in the manual), but there's currently no way at all to hold mod versions (i.e. ignore upgrades).

I have only one question for this thread: Why do we need mod-packs at all? I see no advantages to installing via a mod pack, individual downloads aren't difficult, and if you want ridiculously easy installation just distribute a .ckan file. If this were a game where inter-mod compatibility requires a lot of tweaking, I'd see a need for mod packs with all the kinks worked out, but for KSP it's repackaging and nothing more. What purpose do KSP mod-packs actually serve, besides recognition for the packer on the back of other peoples work? Why would anyone want one of these things?

Dunno 'bout AMD, but I recently went from a 4 core to a 6 core CPU with otherwise identical specs (besides 1/3 more cache, as expected), and I'm seeing exactly the same single core performance. So long as you're comparing apples with apples (same product line, clock speed and generation) and the manufacturer isn't using "moar cores" as marketing to distract from poor per-core performance (AMD) or skimping on added cache for the extra cores, I really don't see how putting more of the same core elements on the die would make any difference to per-core performance. Granted most 8-core CPUs clock the cores lower, but that's a thermal thing, it's exactly the same silicon for each core, just more of them in the package. Those cores will share things like data bus bandwidth, but that's not relevant for single-core performance either. Hell, many 4 or 6 core CPUs are 8 core packages, with some cores disabled due to defects or to fill quotas (i.e. binning). Got any benchmarks to back up that "will more than likely" bit?