« first day (2035 days earlier)      last day (2994 days later) » 

5:02 AM
have those cards in them. So I was able to grab one of the cards from a machine that died (processor was dead, but the 32 GB of DDR3 and a M6000 graphics card managed to survive the short-circuit) for around $900. Which, if you look at the normal price for the M6000 alone, was an absolutely great deal.
Her work's actually pretty cool about their old computer parts. Their contracts with NVidia, Dell, Intel, etc. dictate that they cannot re-sell the computers and must recycle them (Why this is necessary, I will never know), so my mom's boss sells any parts that I want from their old or broken (keep in mind, that just because a computer is "broken," doesn't mean that nothing works inside) at prices that are insanely cheap.
 
ooh
the M6000 iirc is basically one of the (older) titans
Different firmware, and performance is better at some things iirc
 
One of my servers (the one that hosts the main servers (FTP, file, web, etc.)) is running off a dual-socket server motherboard and two $1200 Intel Xeon processors (I find it cool how it actually has two separate processors!) that each have 8 cores (All together the board has 16 physical cores, but has 32 hyper-threaded cores). The server also has 20 Terabytes of storage (I will probably never fill up the drives). And I got all those parts for around $1,000 when they sell at $2600 for the >>
 
proccessors alone. Not to mention the costs of ten, 2 Terabyte drives that operate at 10,000 RPM's. So those are about $120 a piece times 10. I get some really deals from this, and they get to actually make some of their spent money back from machines that would've just been recycles. It's a pretty cool gig. The only bad thing is that I can really only get enterprise grade hardware which, while it's much, much more expensive, it doesn't run as fast as consumer grade processors such as theCore-i7
::Word Missing::
"...some really deals from this, ..." ----change--to----> "...some really good deals from this, ..."
 
Different optimisations
 
5:15 AM
Oh yeah, completely. The Core-i7 is meant to run super fast at high operating temperatures, whilst the Xeon series is meant to run 24/7 at stable operating temperatures and to use as little power as possible.
 
and more threads over per thread performance
 
Same thing goes with the NVidia GTX series vs the NVidia Quadro series. Almost identical cards, in some cases, but the firmware settings make the Quadro significantly slower at rendering while also being extremely fast as calculations, and extremely accurate at rendering (you almost never need more than 2x MSAA for anything).
 
Yeah
old workplace used a ton of em
 
Another thing about the Quadros, especially the newer ones: I've found that they tend to provide much more stable frame rates (and when placed against an "equal" card - eg. M6000 to GTX Titan - higher frame rates in total) when displaying a ton of objects on the screen.
 
Bob
@AlexSolon ...wow.
 
5:22 AM
They're terrible for gaming. Some of my games won't even run on my Quadro cards. For example: Counter Strike: Global Offense is, by all means, not a very demanding game, but the $4,500 Quadro M6000 card that should theoretically be 100% capable of running the game at above 200 fps, will barely run the game at 20 fps.
 
Bob
That's all I have to say :P
s/Offense/Offensive/
@AlexSolon What CPU are you running it with?
IIRC it's far more dependent on CPU performance than GPU.
 
Hmph, it is global offensive and not offense isn't it
 
Bob
Also, for some reason it just underperforms on some systems with identical hardware to far better performing systems :\
 
Oh, I was running it on a Core-i7 4770k machine that I borrowed from a friend (we basically put the M6000 in his desktop). None of the hardware should've been a limiting factor, and I did try multiple driver versions (including the NVidia "High Performance" drivers for the Quadro series), but no matter what hardware we tried running it with, the Quadro cards failed to perform
 
Bob
Interesting.
 
5:28 AM
That's true. And it's not even the people who make CS:GO's fault. It has to do with the very back-end of the Source engine not being optimized for a lot of hardware. But, in Valves defense, I do see their reasoning for not optimizing towards Quadro cards. I mean, around 1% (absolutely no scietific evidence behind that number. Extreme estimate) of the market actually has Quadro cards. Most of them are too expensive for most people to afford, and the people that can afford it will usually>>>
just get multiple Consumer cards
 
Bob
1%? Far too high.
 
Probably haha
it's probably closer to 0.1%
Or less...
 
Bob
It'd be lower even if you went by units shipped.
But if you actually wanted the number of people who use them for gaming...
Yea, that'd be minuscule.
 
You could always check the Steam database
 
Bob
Doesn't help that you could easily buy 20+ decent graphics cards for the price of a single Quadro
Or something like 6 really high end ones.
@AlexSolon hm. not really representative of total units, but it's close enough for Steam engine dev, I guess
@AlexSolon They're not even on the list
 
5:32 AM
I mean it's an accurate representation of how many people use them for gaming, because most people who play games on the PC at all use Steam.
 
Bob
So definitely below 0.35%
 
steam mis-detects my monitors
 
Bob
Huh. 70% of cards are "DirectX 8 GPUs and below"?!
 
Look up the Steam force monitor profile options
 
Bob
And this number went up by 1% in the last month?!
Wait. I think their detection broke.
 
5:34 AM
I run dual UHD, it detects dual 1440p
 
Bob
Last December it was 80% DX 11
Then in Jan that dropped to 14.5% DX 11 and DX 8 (and below) jumped from 2% to 68%.
 
Holy crap. We need to start a Kickstarter or some sort of fundraiser to help these poor, innocent souls who are stuck on DirectX 8 cards.
 
Bob
Yep, that's broken.
@AlexSolon I'm reasonably certain you can't even run Windows on those.
Windows 10 requirement:
> Graphics card:
DirectX 9 or later with WDDM 1.0 driver
 
You can, but from vista and up, you won't have any sort of native hardware acceleration. And that Windows 10 requirement is inaccurate. In order to use all of the graphical optimizations for the Windows UI, you need a DirectX 11_1 card or higher
Wow... It sounds like Steams detection API thought that DirectX 11_2 somehow got sent "back to the future."
 
@Bob I've gotten windows 10 running on one of these thinkwiki.org/wiki/Intel_Graphics_Media_Accelerator_950
in one of these thinkwiki.org/wiki/Category:R60 (it wouldn't even install)
 
Bob
5:38 AM
@AlexSolon Well, I think I was reading off the minimum requirement list. Not the recommended one.
@JourneymanGeek Didn't you hack around it? i.e. not a clean install?
 
@Bob I added more ram, installed, removed the ram
 
@JourneymanGeek Kill it with fire. That thing shouldn't even be allowed to be used!
 
Bob
There's DX 9 drivers available for the GMA 950 anyway.
> Updated Intel® Graphics WDDM drivers supporting the Intel® 945G Express Chipset are included in the Windows Vista* February CTP build (build 5308). These drivers are only compatible with this build and should not be used with other versions of Windows Vista.
o.O
 
Also needed a ssd for it to be usable
 
Bob
> With a powerful 400MHz core and DirectX* 9 3D hardware acceleration ...
 
5:40 AM
One of the King Dians in this case
 
Bob
So, yea. Yours is still within the min reqs.
 
.... I might have an r51
But no way I can get an SSD in it
 
@Bob I think you're right about the minimum vs. recommend requirements, but I think they may have updated them to say that they recommend a card that is DirectX 12 capable (surprisingly all DirectX 11_2 cards support 12 even if they weren't designed to. Not all aspects of the API, obviously, but those parts would be "emulated" almost)
@JourneymanGeek @Bob, a basic rule of thumb it, that when you're building a computer, you never use Intel graphics. Never. Using integrated graphics on a AMD CPU is just fine (especially if you have one of the APU's, which is essentially a less efficient GPU put on top of the CPU), but never Intel
 
@AlexSolon rofl
@AlexSolon My gaming boxen are nvidia. I prefer intel graphics on laptops
(I've been building system for years and a deft hand at bringing fairly old machines up to scratch)
 
My AMD FX-8350 Black Edition (8 cores, O/C'ed to 4.9 GHz, water-cooled) has integrated graphics that are capable of playing Battlefield 4 at medium settings (even high in some places). And I also should revise my statement: Unless you have one of the brand new Skylake processors (The ones that started support for DDR4) and it's a high-end Core-I7, don't use Intel graphics
 
Bob
5:46 AM
@AlexSolon Meh. I've considered Intel graphics decently good since Sandy Bridge.
Since it became integrated, in other words.
 
Bob
Not something I'd recommend for gaming, but it works for just about everything else.
Then again, I wouldn't recommend any iGPU for gaming (of the modern 3D variety). Not even AMD ones.
 
@JourneymanGeek I use Intel graphics for a lot of things, actually. Just not for gaming, graphics editing, or programming (opengl/dx api programming in particular). My MSI Stealth Pro GS70 has a GTX 980m discrete card, but, when not under load, it falls back to the Intel graphics.
 
5:49 AM
The only thing that differentiates Pre-Skylake Intel CPU's and Post-Skylake CPU's in their usability in gaming, is the fact that Skylake brought Intel graphics up to the newest OpenGL version 5ish (they've been at version 3.3 at the highest for many years. Most Intel graphics are either version 1.1 or 2.0)
The newest version is OpenGL 4.5
Well the Skylake CPU's are italic_slightly_italic behind at OpenGL 4.4. But, most games don't use anything newer than 3.3, but that will change very quickly, and very soon, because OpenGL 4.0 has been out for quite a while. And any game for the next-gen consoles, I guarantee will use DirectX 12 or OpenGL 4.0 or newer. And if they're not using those versions, then they're probably indie games.
I'm going to test something real quick, so don't get annoyed if you see some random text in chat
_italic_Italic Font Test_italic_
__bold__Bold Font Test__bold__
italicItalic Markdownitalic
italic Markdown italic
Okay. How the hell do you do italics with the word "italics" showing up?
 
Bob
@AlexSolon Do what now?
Can you give me an HTML example?
 
Like italics on the chat system. The FAQ says it's "italic text data italic"
Hang on, I think I got it. italic Test Markdown Italics italic
Dammit
 
Bob
@AlexSolon I mean, can you give me an example in HTML?
I'm not entirely sure what you're asking.
 
The equivlent of the <i> tag in HTML. The boards on stackexchange use something similar to HTML markdown (simplified html that's only used to manage how text, images, etc. looks without ever actually creating elements), and the FAQ for the StackExchange chat says that you can do italics with either italic or italic, and then bold with either bold (if this shows up bold, it's just ** (two stars) on both sides of the word bold
 
Bob
@AlexSolon Could you give me an actual full example of what you are trying to achieve, in HTML?
Because as far as I can see, you've already got italics there.
Give me an example of the end result you want.
Not a text description. An actual example I could dump into a webpage and render.
Then I can tell you how to reproduce it in Ghetto Chat Markdown (TM) or if it's not possible.
 
6:02 AM
Wait. Wow. I feel stupid now. In order to make a word bolded in chat you put the two asteriks on both sides of the word/sentence. But you never actually use the word "bold," so why the hell would they make it any different for italics. I guess you don't use the word "italic," but rather just single stars rather than the two stars for bold.
Hang on, I'll give you an example on my website, because I can't type it in here without it trying to format the markdown
 
Bob
@AlexSolon ...That's what I guessed you were trying to do, but it didn't really make sense so I wanted the example :P
@AlexSolon Just hit ctrl+k or click the fixed-font button.
 
@AlexSolon the "help" link covers most of the formatting
 
raknov.com/testmarkdown.html <--- This is an example, that I uploaded to my website, of what I mean. By markdown language
Yeah, sorry guys. I'm a little newer to the StackExchange chat concepts
FYI if you follow a link to my website currently, do NOT put https:// before it, because it will just reject your request for now. Use the standard http:// instead
 
Bob
@AlexSolon And ... that works as described.
So, there's no problem?
:S
@AlexSolon Heh. Set up HTTPS already! :P
 
6:24 AM
Actually, I'm working on that as we speak! But I'm also implementing a system the texts my phone to tell me there's updates avaible for my server, and then e-mails me the list. And then I'm going to try and make it so I can either text or e-mail "approve" in order to install the updates! It's a wicked concept and not hard at all. Linux has a built in, lightweight smtp-server (mail server) that can send mail via the "Sendmail" command. Now most phone >>>
providers offer you an email that corresponds with your phone. Like I have att so mine is 1234565566@txt.att.net (if I put my actual phone number in and send an e-mail to that address, my phone will get a text that's formatted like an e-mail (sender, subject, etc)
 
Anonymous
@AlexSolon so, logwatch :)
 
Bob
@AlexSolon "most" :P
 
Anonymous
but it's cool, your system does a couple of things logwatch wouldn't do by default, and it's cool you build things yourself
 
Also, despite the fact that I have a Core-I7 in my laptop (you can look up the specs for the MSI Stealth Pro GS70. It's the model with a 980m and not the 970m), google chrome still want's to start lagging with 50+ tabs open. It's wierd though, because I can have 500 tabs open (trust me. I've done this) in 25 different chrome windows and the laptop never bats an eye. If fact, I can still run games without even noticing that I have 500 windows of the nyan cat meme-gif open in the background. Wierd
It isn't *too* complicated to understand. Basically I have a script that will be set to run 2-3 times a day (I will probably make that configurable), and when it runs, if there's any updates aviable it will send a message to my phone saying
"Sender: #somerandom number thing that's werid
Subject: SkyboxServer Notif.
Data: X number of updates are avaiable to be downloaded and installed. An E-Mail listing these updates has been sent to your configured account. Text "approve-update" with you 4-digit pin number, or respond to the e-mail in a similar fashion"
And the automation part can be taken care of by the cron system in linux
Pro-Tip: If you're anything like me and you find that you sometimes have so many tabs open in Google Chrome that you can't even see the titles anymore, Ctrl+Tab will go forward one tab, Ctrl+Shift+Tab will go back one tab, and Ctrl+W will close the currently selected tab. And if you really feel like adding even more tabs to the mess, Ctrl+N opens a new tab, and Ctrl+Shift+N Opens an incognito tab
 
Bob
Eh.
 
6:35 AM
Just thought you guys might like those shortcuts, in-case you didn't know them all already! They're, by far, the most useful ones.
 
Bob
It's an interesting idea, but not a particularly good one.
Updates can break things.
Automating updates when you're not there (i.e. sitting at the terminal) can make breakages worse.
IMO you're better off just signing up to your distro's security mailing list.
That way you'll catch critical security updates fast, and feature/bugfix updates really don't matter much on a stable system.
@AlexSolon They're probably old news to most people who frequent this chat :P
Also, meh, I use FIrefox anyway. It doesn't hide tab titles from me.
 
@Bob, exactly why I'm implementing an approval system with a password-pin. For example: if I check the update list and see that there's a linux-kernel update, or an update to anyone of the server components (apache, mysql, php, etc.) then I'll wait until I get home to update the system
 
Bob
Only 360 tabs (in one window!) right now, but I've hit 2000 before.
@AlexSolon Then why do you need the response + pin? Why not just initiate the update from a computer?
Also, security mailing list will probably give you additional advice (e.g. restart recommended) or mitigation tips or potential impact.
Just seeing the update list won't provide that.
 
Oh, when my mech-keyboard decided to malfunction at one point, it did so right as I hit Ctrl+N in chrome. So it continiuously opened new tabs until it gave an "Low on memory" warning. Do you know how many tabs it takes to give that warning on a 16 GB system???
I could always do a combination of the two. And because I have school during the day, so I can't always get on my laptop (although I do carry it with me) all the time. And my phones not the best phone in the world, so it doesn't have the ability to even check e-mail. That's why I'm implementing the feature to text me and let me know that I need to check my e-mail and see what updates are avaiable. But I was thinking that if there's only 2-3 Updates avaiable to just go ahead and list them via SMS
And if I approved of it, then I could type "approve-update" and in another text "XXXX or password" where you could either configure a 4-digit pin number or a password that needs to be e-mailed or texted to the server in order to initiate the update sequence from an external location (which isn't the most secure practice, in general. Hence the need for a pin/passcode)
 
Bob
@AlexSolon Still risky, IMO.
There's just about never an update that requires action that immediate, especially on a personal server.
You're playing with the risk of downtime until you can go fix it, vs the risk of the lack of update for a few hours.
The first is far more likely. The second is probably more dangerous in the very very rare cases it occurs.
 
6:44 AM
I'm doing more for the ability for it to let me know that there are updates avaiable, and less for the ability to update externally. Because in reality, if I can get on my laptop, my UPS battery backups can turn on or off servers over the local network (which I can access via ssh to any one of the servers. I do keep the 48 character password on hand in my wallet. It's quite the password)
 
Bob
Eh. It's up to you. I'd still recommend the security mailing lists.
 
@Bob, I can restart anyone of my servers from the Network either by these special Ethernet ports that let me change BIOS settings and control the power state from over the Network. Or I can send an acpi shutdown/power-on signal (The power control unit for the UPS actually has cables the plug in where you would normally put the headers for your power/reset switches
And then for the computers that don't have either options(talking about my Raspberry Pi Model B micro-server here) the PCU (Power-Control-Unit) also has an option to cut power to any of the outlets on the back, and the turn it back on; which is exactly how you turn-on or hard-reset the Raspberry pi (by unplugging it basically)
 
Bob
@AlexSolon It's not a matter of restarting. It's a matter of updated potentially breaking the system, in ways a restart can't fix.
In fact, in such a situation, restarting would be a terrible idea.
 
It's actually a pretty cool system, but the password for managing power-states is over 100 characters long and composed of a-z, A-Z ,1-9, and your normal symbols. And, if I'm not mistaken, I actually have the ability to use a password at all disabled. It's setup to work like an ssh connection where's there is the server-key, public-key, and client-key systems
 
Bob
And I say all this because it's happened to me before.
As for your how many tabs, in Firefox it's ~1000 blank tabs per GB.
 
6:52 AM
Every server (excluding the raspbery pi) has what is called a three-part boot-loader. First it boots into the grub boot manager, then directly into a special micro-operating-system (I'll explain in a second), and then into Linux (for most servers. Some of them have the latest version of Windows server. Aptly obtained via ISO and KMSActivator ;))
 
@AlexSolon we don't typically admit to condone piracy on this room
 
If that's the size for firefox, then I would say Google Chrome is probably close to 700 tabs per Gigabyte, as it is extremely memory-intensive. So I had somewhere around (14.5 * ~700) tabs open
 
Bob
Probably more than 1000 actually, since I didn't account for the initial size.
1000 is the figure with the standard newtab page. If you set it to an actual about:blank you get more. Dunno how many more.
I'd guess ~1300.
But you'd also crash pretty damn fast unless you were running 64-bit FF
And the current e10s alpha hogs memory like no other
@AlexSolon eww, grub. eww, complex boot routines.
And here I am trying to simplify boot as much as possible. More robust that way.
Server goes straight to EFIstub and into the kernel.
 
Two things, it's covered under the US law for like backups or whatever, because I do have a Windows Server license. It just happens to be a PITA to constantly update that license with the official Microsoft servers (it doesn't always work right, which is bad for a server). And on top of that, the two servers using Windows Server are only accessible from my local network (I have
 
Bob
(that was also a necessity cause grub support for zfs fluctuates between "kinda broken" to "completely broken" to "are they just pretending to have implemented it?!")
 
6:58 AM
them locked down to a IP range of 192.168.0.2-254 and localhost of course. Technically, the server's are capable of being accessed for configuration or management outside the network, but you can't access the files/information on them. One hosts a file sharing server (Samba) along with a server for my old ps3 and stuff
 
Bob
fyi, you don't run "samba" on Windows
it's SMB
Samba is a specific open-source implementation that people run on *nix
If you actually ran a copy of Samba on Windows that'd be rather weird
 
It's a Media/File Sharing Server leveraging the Samba protocal.Actually the boot process isn't complicated at all,
 
Bob
Protocol is called SMB.
Samba is an implementation of it...
It ... annoys me when people call it the other name.
Hm... now you have me wondering if Samba can actually run on Cygwin. I assume it wouldn't run on native.
 
^ it goes straight from grub (which doesn't actually do anything other than configure some super low-level CPU options. Like how my server board want's to force my CPU's to throttle all cores, but with grub 2 I can set some kernel params. the override that on 4 of the 16 physical cores. So I essentially have 4 cores always running at full speed (not counting for the Intel SpeedStep factor, which take's them from like 4.3GHz down to 2.9GHz but almost always bring's them straight back to 3.7
Ghz under a light-load
 
Bob
Looks like someone ported the client via Cygwin, but the only real discussion of the server on Windows is from 2001...
@AlexSolon I can see the necessity of complex boot processes, sometimes, but I'd prefer to minimise steps where possible.
bootmgr can be a real nightmare in that sense
 
7:03 AM
Like you said, Samba is just an implementation of SMB (although it's usually referred to it being the implementation of SMB, due to it's durability and popularity), and there are other variations that follow the exact same protocols and stuff, but on a linux server. So as far as any outside computer is concered, it's talking to a windows media server (even though it's actually a linux server).
 
Bob
@AlexSolon Hm? Thought you were talking about Windows servers.
 
Well I never did finish saying that. Grub only sets some low-level config options (otherwise I wouldn't even have it enabled.), after that it boots into a custom linux Operating System that is accessible only from SSH or actually attaching a monitor and keyboard. Basically what that secondary layer lets me do is recover any server remotely, and if I really need to, I can actually re-image the server using a fresh copy of any OS
 
Bob
Eh... that's what OOB management is for.
If you're running on a server board, you almost certainly have a BMC on there.
 
My implementation of the SMB protocol is the Samba version, and it's running on a Windows Server (2012 I think?). But It is possible to run an exact copy of the Samba version of the SMB proct. So if a Windows computer (or mac osx, or linux, whatever) it sees the Samba server as being ran on a windows server platform (which allows for some trickery between linux and windows. Like how you normally wouldn't be able to steam a movie from a linux media server, but you could from a windows
 
Bob
@AlexSolon Hm? You're literally running a bopy of the Samba server software on Windows?
I'm reasonably sure it doesn't even support Windows...
 
7:11 AM
<<< server (an artificial limitation imposed by Windows). But if you run the copy-cat Samba on linux, Windows doesn't know the difference. That's how I used to run it before I bought a license for Windows Server 2012 (btw the true reason I use KMS is because my license includeds one physical license (not virtualized), and three licenese for the OS to run in a virtual machine. Now any of those virtual server licenses can be utilized on a linux-virtual machine running windows, but when you try
^ to run all the licenses on the same machine (I have the hardware license, and both virtual server licenses setup on one server. It also has a virtualized linux server that servers files using protocols unuiqe to Linux/Mac OSX. This is setup this way so they can share the same files but run every possible protocol for file sharing, no matter what platform)
Now the problem that arises is that the license-key activation for the two virtual servers will not work for me when I'm try to virtualize them on a physical machine with the same license-key (one key is shared between 1physical, and 2 virtual)
 
Bob
@AlexSolon In other words, you are running a Windows file share, on Windows, probably with SMB 3.0.
The only place Samba would be involved would be on the *nix side.
It's a terminology difference, yea, but I think I first started getting annoyed at it when people asked about "Samba" issues where their entire environment was Windows, on both ends. That just makes the issue harder to diagnose.
 
So I just use KMS Activator for the virtual machines. Technically it's a violation of the Microsoft EULA, but all that does is say that if anything happens, I'm not going to get any support for those two virtual servers (although I still could for the physical because its a "legit" copy of Windows Server). This concept is actually reinforced by a few DRM Laws that were created a while ago that state that you can make as many backups as you want of any digital data (movie, cd, windows, etc.) as
 
Bob
Also that the Samba version is distinct from the SMB protocol version.
 
^ long as it's for personal use and you own a legitamate copy in the first place.
 
Bob
It's like calling all DNS servers "Bind". Or all HTTP servers "Apache". Or all SQL servers "MySQL".
 
7:17 AM
 
Bob
It might be right a lot of the time, but the rest of the time it's just flat out wrong.
 
Now if I ran three Windows virtual servers on top of the physical windows server, that would be considered illegal due to the fact that I don't "own" that fourth license. Now Microsoft "fixed" this loophole by technically making the Windows 10 and newer Server licenses "rent/lease but not to own"
Lmao, I like the comic strip! And yeah, I'm using whatever SMB server that comes built into Windows Server. I don't remember the exact name, because all I did to set it up was give it some folders to share access to and tell it to not let anyone connect outside of my network.
 
Bob
@AlexSolon I think they just call it "File and Printer Sharing". Something like that.
But that's also a nicely ambiguous name... so it's easier to just call it "the Windows SMB server" sometimes.
 
Everything that is uploaded to the server itself (the physical server doesn't "share" the content, but hosts the content. The two media virtual servers (One linux and the other Windows), serve/share the content to the clients connecting. And what protocol you're attempting to use dictates what server you access (you actually don't have to worry about that part as it's mostly automatic due to the fact your "connecting" to the physical server, and it takes care of the routing to virtual servers >>
depending on what protocol is being requested
 
Bob
I've only ever run Windows servers virtual (except for a minor entanglement with 2003) so I've never had licensing issues there
 
7:24 AM
Nah, it's not "File and Printer Sharing," It's either a more complex version of that, or a version that's more aimed towards streaming stuff like music or movies across my network.
 
Bob
@AlexSolon DLNA?
In which case it has nothing to do with SMB at all.
Windows technically has an integrated DLNA server, but it's not very good :\
Heck, the DLNA protocol in general just plain sucks.
 
If I was trying to install the Windows virtual server on a linux host machine (or a windows server with a different key than the one used by the virtual server) it would authenticate just fine
 
Bob
I've used Serviio before, but they've progressively gotten worse.
Then I abandoned DLNA altogether.
 
I think there's around 10 media sharing protocols all together on that server. There's SFTP (ssh over ftp) which is restricted to the directories that are shared on the network (I could allow it file access to the whole server, but that's a security risk), theres the Linux file sharing protocol (Samba, I think). And then there's a few specialty ones for the Xbox systems and Playstations. Also, since I never use hulu, I actually have a network override that redirects the hulu domain to one of
 
Bob
@AlexSolon s/ssh over ftp/ftp over ssh/
which isn't really ftp at all but we pretend it is :P
afaik SFTP has little in common with FTP(S) apart from the name
 
7:29 AM
my "custom" protocols (I made a makeshift interface between standard SMB and the protocol the Streaming device was expecting). What this did was allow me to trick a device that would normally only allow streaming from services such as netflix, hulu, amazon, etc. and not from local servers, to allow me to stream from a local server by Hijacking the Hulu protocols (and I also had to fake an SSL certificate for it to allow that, which if I had done that outside of my network, would be extremely
 
Bob
@AlexSolon There's really no "the" Linux file sharing protocol, btw
I think at the moment it's a toss-up between SMB (implemented by Samba) and NFS (...implemented by BSD NFS? idk)
 
illegal, and I'm pretty sure it's also a federal crime. But it's legal and okay because it's only "faked" on my home network. All I had to do was intercept the SSL certificate info from the connection to Hulu on my laptop, copy that data into a "custom" SSL cert. And hand over the fake certificate when it asks for it (I originally tried not faking the cert. but all devices said the connection to Hulu could not be verified as the SSL cert. is no longer valid
@Bob, it's a mixture of a lot of different protocols for different devices. I don't mess with each one individually, but instead manage the two servers, and their individual protocol settings, via a web-interface called webmin. Basically all it does is provide a common GUI to editing all the config files for each and every protocol-server (long and tedious for 10+ servers/protocols for just File Sharing).
 
Bob
webmin... hand in your sysadmin card, please.
 
lol
kicks it oldschool
I do use a Gui for setting up kvm VMs but I don't do that often. Otherwise, most of my interaction with my linux boxen is over ssh
 
Bob
@JourneymanGeek I don't even do that.
virt-builder + virt-install + virsh ftw
 
7:37 AM
@Bob bit too lazy to learn the syntax + some VMs are windows VMs
 
Bob
(though, sometimes I wish I did...)
 
@Bob my current preferred BT client dosen't expose one feature I need over web, so I need a gui anyway
(rss feed downloads)
 
Oh come now. Never. There's two reasons I use webmin for it, there are more than 10 servers that all have the exact same settings, but each one of those servers has a different configuration setup. All webmin allows me to is to change a "config option" globally for all servers. Not webmin itself doesn't change it but it actually passes the information on to a script I made that changes the
configuration files. And the second reason for using webmin, is that, without opening up an ssh server on a different port than the other 6 ssh servers I have setup (I think there's actually closer to 25+ ports dedicated to ssh connections to each virtual machine and their corresponding physical machine management system (That "second layer" os I told you about)).
 
@AlexSolon might I suggest you take a look at puppet or salt instead?
 
Bob
@JourneymanGeek virt-builder takes care of that, when it works
 
7:41 AM
webmin's got a bit of a record for being insecure
"when it works" ;p
Yeah, I need to take a look the next time....
and I'll forget and do it the same way I have always done it ;p
 
Bob
@JourneymanGeek heh, you just need to avoid one broken version. Which happens to be the one in the jessie repos.
I can give you my standard setup command if you want
 
@Bob oh I run KVM/virt manager on KVM
@Bob I'll likely forget it ;p
 
Bob
@AlexSolon ...this is where doing things the standard way is probably a good idea...
 
(speaking of which, config management systems is one of those things I need to look at ._.)
 
@JourneymanGeek another advantage of using a GUI for a BT server is that you don't have to know the exact link in order to start a download. You can upload the meta-data file (.tor**nt obviously), and some sites don't even allow you to download the meta-data file directly and instead force you to use "magnet links", which you can't really access from a server at all. The only work
around I've found for that is to start the download on my laptop and let it download just the meta-data, and then copy that over to the server, after I write it to a file. It's pretty easy to do actually. All you need to do is make a fake bittorrent client that hijacks the meta-data (that normally is never written to a file. FYI this is only applicable for magnet links), and collects that meta-data, and in the
 
7:46 AM
@AlexSolon You assume I've not done this for years.
And why would you?
I just run my torrents on a small, low powered box
web ui for adding the latest linux live cds, rss for my entirely legit TV shows
I even have a system that sends a notification to my desktop over gntp + growl, and I move downloads over to my desktop with BTsync
 
@JourneymanGeek as far as Webmin being insecure, I have it setup on ports 12046-12080 (more than needed to reserve ports for future servers). Those ports are blocked via my Router firewall, a hardware firewall from cisco (btw the expensive "hardware firewalls" are nothing more than a normal software firewall than you can't change or modify as much. So it's just dummy proof. Not nesacarily more secure)
 
My old system used a seperate rss feed reader that would spit the torrent files into a folder, which a cli bt client would pick up
but that wouldn't let me filter what to download as easily
 
And then each server individualy is set up to make sure the connection is a local connection by ensuring A) The ip address is in the range of 192.168.0.2-254 and B) The connecting computer has a verified MAC address and C) that the connecting computer has the "public-key" nesacary to connect. You can't even put a conventional password in, you have to have the file that can never be re-generated or cracked. And on top of all that, you still do need to log in with a username and password of a
 
 
1 hour later…
woooohoooo!!!! 2 hour drive to work :-|
 
@Psycogeek Convertible 33" umbrella is good. Has fixed white fabric and removable fabric with silver on its inner side and black on top. I thought I can remove white fabric and replace with silver but it's permanent.

I only wish I bought 135W light versus 115W...
 
9:38 AM
@Psycogeek When I get low shutter does that mean I lack lighting? Do these photos lack lighting? I find that both are not as sharp as I want them to be. Is it because of insufficient lighting or bad focusing?
 
yes they lack lighting
are you using a tripod?
have a large aperture (f2 or something) and increase the ISO rating
the downside to the higher ISO is that you get grainy pictures but it does mean you can have more reasonable shutter speeds
if you are using a tripod you might be making it wobble when you press the trigger leading to the blurry pics
a little technique i discovered was to set the auto timer then you can step away from the camera and prevent trigger wobble
 
9:55 AM
unix user account that's configured to be an "admin" . Oh and did I mention that any admin account for all my servers (defined as any account that has more ability than others) has a 100 character password involving Captial, lower-case, numerical, and every symbol above your 1-9 keys (I was too lazy to include the rest of the symbols into the random generator). So each admin account has a different password for every single computer. So to, in review/conclusion, you have had to have access to the server in the past 72 hours in order to hold down a button on that individual server (Connects
 
10:20 AM
0_0
WHoa. MASSIVE wall of text
 
There's more... I'm still typing
 
Its chat. not essay writing hour ;)
 
@AlexSolon That wall of text should be in a blog ... :/
 
And weeeeel
It feels like you're trying too hard to impress folk here. There's no need to XD
 
he gets paid by the character
 
10:31 AM
I only try to impress folkettes.
DisplayPort 1.4! Yay!
 
Wow, I did not mean to post the entire concepts wrapping around my server security and furture plans, and a possible innovation in the field of hardware security (protecting data stored on a medium not by preventing software from accessing/changing it - hardware security is completely unrelated to the software security of data (although they are sometimes related in one way or another. Ex. The Apple iPhone had a very small microprocessor on it's operating board, this chip generates hash-keys (a specific stream of data that's usually capped by power of eights bit measurements (16-bit, 32-bit
 
was that meant to be an email?
 
Yeah, this is chat support, not email.
 
Bob
s/support//
 
/s
 
10:40 AM
@JourneymanGeek not really. Just thought I would post something I've been thinking about. There is no better way to better the world than to share knowledge
 
Bob
> 100 character password
y'know, I'm just gonna stop reading there
 
17 mins ago, by DavidPostill
@AlexSolon That wall of text should be in a blog ... :/
 
Bob
Not only would that be unnecessary, it's also utterly impractical
 
hehehe
I assume you tried to read the card and can't? Depending on what was damaged, a portion might still be readable. If not, the only thing I found after extensive research involved leaving goat entrails under a full moon and chanting. I'd be surprised if that actually works, though. — fixer1234 4 hours ago
 
@AlexSolon "I did not mean to post the entire concepts wrapping around my server security and furture plans" - you just did.
 
Bob
10:44 AM
You're talking typing in an alphanum password that's almost an order of magnitude longer than a 128-bit key.
 
@qasdfdsaq I saw that. Might take a while for things to actually support it tho
You can get your UHD at 120hz tho ;p
Unrelatedly
theverge.com/2016/3/2/11145148/… "Kanye West? Hell, what I wonder is Kanye Pay?"
 
@JourneymanGeek Yeah, DP1.3 which was finalized in late 2014 is just reaching the market now. :-(
 
But I want it nao!!
 
I think I'm 'slumming' it with DP 1.2 ;)
 
10:52 AM
Heh I'm on 1.2 for most things but I suspect my next monitor will have 1.3
 
Thankfully I'm unlikely to need another 'first line' monitor for a while
 
Yeah that's cause you have 4K already :-P
 
I might buy a shitty one for the old system, but ironically, that's waiting on $$
@qasdfdsaq and no real regrets
even the TN screen + IPS combo works well for home use
 
that seems handy
 
@HackToHell o_0
 
> wolfe
of course the dog thinks it's handy
me? I hate snakes (python) and dogs just smell funny and try to lick me
*hisses at the snake*
 
!!baroo
 
11:42 AM
*baps JMG right on the nose with a claws-retracted smack* Bad dog! Don't ally with the snakes! If you want to use your nose to good effect, sniff out where Owner keeps the cat treats jar.
Cats are superior
 
Bob
!!tell 27959119 foxno
 
Bob
(sorry, couldn't resist)
 
You stay out of this, vulpine! :P This isn't your fight.
that's a hilarious vid though
 
@allquixotic chances are a cat would have a better chance. This wauzer can't really jump
 
11:49 AM
my RL meow kitty is looking at me, I think that means I'm her treats slave
 
That works.
 
funny thing is that cats are absolutely ruthless and manipulative; they relish in power.... whereas "pet" dog breeds have had those traits bred out of them
 
my cat acts all sweet and cute until I give her treats, then she walks away like "pff, you're nothing to me, you fat loser"
 
11:54 AM
That's all cats
 
Humans have pet dogs. Cats have human slaves ;)
3
 
a fairer canid comparison to cats would be wild wolves
they are probably pretty damn fierce and ruthless
pet dogs are wusses :P
 
Very fierce and ruthless.
 
awww goggy so CUTE!
nevermind then ;p
 
> These two are not 100% wolf.
 
11:59 AM
@qasdfdsaq first google hit.
ALso bears
 

« first day (2035 days earlier)      last day (2994 days later) »