A short survey



  • This is what Asus wrote to me:

    <font id="yui_3_7_2_1_1368123334610_2039" face="Arial">This survey aims to understand your level of satisfaction and expectations towards ASUS and its products. </font>

    <font id="yui_3_7_2_1_1368123334610_2035" face="Arial">Your opinions will help shape the future of ASUS and you'll stand a chance to receive <font id="yui_3_7_2_1_1368123334610_2036" face="Arial">one Asus 500G USB 2.5inch Hard </font>Drive upon completion of this short 25-30 minute survey. </font>

     This just made me think WTF!

     



  • Compared to the 80+ question Microsoft surveys I've seen, that is short.

    Incidentally, in my last job we had a survey product that we sold to companies. It was always a huge challenge to talk them down to 6-8 questions instead of the 60-80 they invariably wanted to ask.



  • @renewest said:

    <font id="yui_3_7_2_1_1368123334610_2035" face="Arial">Your opinions will help shape the future of ASUS and you'll stand a chance to receive <font id="yui_3_7_2_1_1368123334610_2036" face="Arial">one Asus 500G USB 2.5inch Hard </font>Drive upon completion of this short 25-30 minute survey. </font>

     This just made me think WTF!

     

    I know. If your opinion is so valuable, they can offer you something better than a hard drive that doesn't even have a terabyte of storage space. What is this, 2010? Who even USES Gigabytes as a unit of storage anymore?

    Of course this is exactly what I'd expect from a backwater company whose flagship product is STILL [url="http://www.apple.com/ca/ipod-touch/"]capped at 64GB[/url]

     



  • I would write a survey where the first question is "What is your name?" and the remaining 249 questions are "Why did you choose this answer?" and then I'd give anyone who filled it out correctly a photocopy of a cookie.


  • Winner of the 2016 Presidential Election

    @Ben L. said:

    I would write a survey where the first question is "What is your name?" and the remaining 249 questions are "Why did you choose this answer?" and then I'd give anyone who filled it out correctly a photocopy of a cookie.

    1. Joe
    2. I chose that answer because I want a photocopy of a cookie.
    3. I chose that answer because I want a photocopy of a cookie.

      ...
    4. I chose that answer because I want a photocopy of a cookie.


  • @Lorne Kates said:

    Who even USES Gigabytes as a unit of storage anymore?

    ben@loads ~$ df --si --total
    Filesystem                    Size  Used Avail Use% Mounted on
    devtmpfs                      1.7G     0  1.7G   0% /dev
    tmpfs                         1.8G   11M  1.7G   1% /dev/shm
    tmpfs                         1.8G  182M  1.6G  11% /run
    tmpfs                         1.8G     0  1.8G   0% /sys/fs/cgroup
    /dev/mapper/vg_loads-lv_root   53G   26G   25G  52% /
    tmpfs                         1.8G   24M  1.7G   2% /tmp
    /dev/sdb2                     500M   95M  379M  20% /boot
    /dev/mapper/vg_loads-lv_home   99G   66G   29G  70% /home
    192.168.1.79:/opt             152G   37G  108G  26% /australium
    /dev/sdg1                     1.5T  273G  1.2T  20% /mnt/you_ess_bee_one
    total                         1.8T  401G  1.3T  24%
    

    So the total of my two computers is 401GiB out of 1.8TiB used. (The Used column counts the entire drive, even on network shares.)



  • @blakeyrat said:

    It was always a huge challenge to talk them down to 6-8 questions instead of the 60-80 they invariably wanted to ask.

    Question 57: Is your wife attracted to doughy, middle-aged marketing managers?

      If yes -> Question 57b: What is your home address?



  • @Ben L. said:

    tmpfs 1.8G 24M 1.7G 2% /tmp

    No! Bad! hits with rolled-up newspaper Where did you learn this?



  •  @morbiuswilters said:

    @Ben L. said:
    tmpfs 1.8G 24M 1.7G 2% /tmp

    No! Bad! hits with rolled-up newspaper Where did you learn this?

    It probably came that way. Somebody recently noticed that "If I put /tmp in memory, my program runs faster! herp! derp!"

    And instead of thinking "I guess my program should have been using memory instead of disk" they thought "Well OBVIOUSLY /tmp belongs in memory for EVERYONE!"



  • @Ben L. said:

    192.168.1.79:/opt 152G 37G 108G 26% /australium

    we has teh IP address and a hostname now, precious ... we hax thee!



  • @zelmak said:

    @Ben L. said:
    192.168.1.79:/opt 152G 37G 108G 26% /australium

    we has teh IP address and a hostname now, precious ... we hax thee!

    Huh, that IP just loads up a hardcore German shit fetish site for me.

    Oh wait..



  • @superjer said:

    It probably came that way. Somebody recently noticed that "If I put /tmp in memory, my program runs faster! herp! derp!"

    And instead of thinking "I guess my program should have been using memory instead of disk" they thought "Well OBVIOUSLY /tmp belongs in memory for EVERYONE!"

    The real scary thing is if you have some application that expects to be able to drop a 2gb turd in /tmp. And it should only be faster if /tmp can't fit in the disk cache in the first place, which is exactly when don't want it in memory.



  • @morbiuswilters said:

    @superjer said:

    It probably came that way. Somebody recently noticed that "If I put /tmp in memory, my program runs faster! herp! derp!"

    And instead of thinking "I guess my program should have been using memory instead of disk" they thought "Well OBVIOUSLY /tmp belongs in memory for EVERYONE!"

    The real scary thing is if you have some application that expects to be able to drop a 2gb turd in /tmp. And it should only be faster if /tmp can't fit in the disk cache in the first place, which is exactly when don't want it in memory.

    It's in memory. It swaps out if needed. It's automatically blank when I reboot, no deleting required. What's not to like?



  • @Ben L. said:

    It's in memory. It swaps out if needed.

    So basically it's just a less-efficient form of the disk cache. Awesome.

    @Ben L. said:

    It's automatically blank when I reboot, no deleting required.

    Virtually every distro I've seen wipes /tmp on boot.

    @Ben L. said:

    What's not to like?

    You're limited to 1.8gb in /tmp. You're using tmpfs + swap instead of just letting the kernel manage pages for you. Heck, you're expecting tmpfs to possibly swap, which is a WTF in its own right. You're doing something that's worse than the built-in functionality. All I can figure is somebody saw "tmpfs" and thought "oh, that's for mounting /tmp on".



  • @morbiuswilters said:

    Where did you learn this?
     

    Sun Microsystems?



  • @Cassidy said:

    @morbiuswilters said:

    Where did you learn this?
     

    Sun Microsystems?

    You mean whoracle? Did they used to do this? I haven't touched a Solaris box in nearly a decade.. I don't remember ever seeing it in Linux or BSD.



  • @zelmak said:

    @Ben L. said:
    192.168.1.79:/opt 152G 37G 108G 26% /australium

    we has teh IP address and a hostname now, precious ... we hax thee!

     

    Not sure if serious...



  • @morbiuswilters said:

    Did they used to do this?

    /tmp in SlowLoris was mounted in RAM since 2.4.

    ISTR a story of a DBA hastily constructing an instance using /tmp, and this friday-job quickly became production (as it never does). One day the sysadmin bounced the box, losing the DB.  Who got sacked? Well, it was a toss-up between the DBA that didn't choose appropriate storage locations and never scheduled any backup jobs versus the sysadmin who never backed up /tmp because it was .. well.. temporary.

    Yeah, the sysadmin got fired.

    @morbiuswilters said:

    You mean whoracle?
     

    No, design decision by Stunning Micropenisus. Whoracle took a look at a long-broken thing and decided to finally put it ri-- WHAT THE FUCK AM I SAYING?

    @morbiuswilters said:

    I don't remember ever seeing it in Linux or BSD.

    Not experienced much BSD so can't remember on this front. All Linux distros I've managed have /tmp as a separate slice mounted noexec. I didn't like the idea of allocating something expensive as memory to something trivial like temporary storage.



  • Since nobody's explained this in plain English am I understanding correctly that he put his Temp folder on a RAM Disk?



  • @blakeyrat said:

    Since nobody's explained this in plain English am I understanding correctly that he put his Temp folder on a RAM Disk?
    Yes.



  • I love the Rackspace one-question survey;

    "Based on your recent experience with Rackspace, how likely would you be to recommend them to a friend or colleague?" with a 1...10 slider - perfect - sums up your entire opinion of that company/product in a single five-second survey!



  • @Cassidy said:

    Not experienced much BSD so can't remember on this front. All Linux distros I've managed have /tmp as a separate slice mounted noexec. I didn't like the idea of allocating something expensive as memory to something trivial like temporary storage.
     

    As I understand it Fedora, Debian, Arch and Ubuntu have all either done this by default or made it trivially easy to enable over the past few years, although Debian at least was wise enough to back it out once someone pointed out that using a large tmpfs partition in memory to speed up access to files was taking away memory that would otherwise be used as a large filesystem cache in memory to speed up access to files.

    The most commonly cited reason for abusing tmpfs in this way is the sense of eldritch horror associated with writing to solid state disks.  Apparently every time you create a new file on an SSD you have to fill out TPS report in triplicate and sign each copy with the unicorn blood so it's somehow worthwhile to go to extreme and painful lengths to avoid it, even if it means that you end up running out of memory and swapping everything out to the same disk that you didn't want to put /tmp on.

     



  • No, this time I don't agree with whatever morbiuswilters just said.

    @morbiuswilters said:

    So basically it's just a less-efficient form of the disk cache.

    No, it's a faster disk cache because the OS doesn't need to recover anything in there if you turn your computer off, so it doesn't actualy have to write anything at the disk, but will if needed.

     @morbiuswilters said:

    You're limited to 1.8gb in /tmp.

    What's great, because no process will fill up the disk. After you get into those 1.8GB the biggest process is killed, and your computer continues working.Things wouldn't happen that way if you filled the partition that has /var, and if you created another partition just for /tmp, well, you are limited again, and can't use the extra space there for anything else.

    @morbiuswilters said:

    You're using tmpfs + swap instead of just letting the kernel manage pages for you.

    Wait, what?  You are using tmpfs, what means that the kernel is managing pages for you.

    @morbiuswilters said:

    Heck, you're expecting tmpfs to possibly swap, which is a WTF in its own right.

    Yes, stale data will swap to free the memory, and data with lots of access will not. How is that a problem?

     @morbiuswilters said:

    You're doing something that's worse than the built-in functionality.

    Again, what? That's the builtin functionality.

     @morbiuswilters said:

    All I can figure is somebody saw "tmpfs" and thought "oh, that's for mounting /tmp on".

    Well, it was the other way around. Somebody said "hey, we don't need persistence at /tmp, let's exploit that to gain some speed", and created tmpfs.

     



  • @blakeyrat said:

    Since nobody's explained this in plain English am I understanding correctly that he put his Temp folder on a RAM Disk?

    Yes.



  • @DCRoss said:

    The most commonly cited reason for abusing tmpfs in this way is the sense of eldritch horror associated with writing to solid state disks.  Apparently every time you create a new file on an SSD you have to fill out TPS report in triplicate and sign each copy with the unicorn blood so it's somehow worthwhile to go to extreme and painful lengths to avoid it, even if it means that you end up running out of memory and swapping everything out to the same disk that you didn't want to put /tmp on.

    And that's silly; I've used SSDs in my laptops for years and they have no problem with this. It's not like there's that much shit being written to /tmp in the first place, and modern SSDs can handle lots of write cycles. Also, I don't use swap because: 1) it's lame; and 2) I don't want to mess with setting up another encrypted partition just to put swap on.



  • @Mcoder said:

    No, it's a faster disk cache because the OS doesn't need to recover anything in there if you turn your computer off, so it doesn't actualy have to write anything at the disk, but will if needed.

    So it marginally speeds up shutdown times? That's the argument in favor?

    @Mcoder said:

    What's great, because no process will fill up the disk. After you get into those 1.8GB the biggest process is killed, and your computer continues working.Things wouldn't happen that way if you filled the partition that has /var, and if you created another partition just for /tmp, well, you are limited again, and can't use the extra space there for anything else.

    What? Where are you getting this from? The 1.8gb is just the limit of the fs. If he fills it up, the writes block until space is freed up. Now, it's possible he might trigger the OOM killer on his way to 1.8gb, but the OOM killer doesn't even necessarily kill the "biggest process"; I've often see it pick a small, process to sacrifice. And your computer will not continue to work if the process it kills is Xorg or init or a few dozen others. Also, it's not likely it's going to kill the process that's filling up the tmpfs, that's not even how it's even supposed to work.

    @Mcoder said:

    Wait, what?  You are using tmpfs, what means that the kernel is managing pages for you.

    I meant letting the kernel manage disk pages for you. Basically this is a really broken setup which seems to show no understanding of how the kernel manages memory.

    @Mcoder said:

    Yes, stale data will swap to free the memory, and data with lots of access will not. How is that a problem?

    Because that's not what tmpfs is meant for. If you want that behavior, just write to disk like normal; it's easier and the kernel will handle it better. tmpfs is meant for stuff that shouldn't be written to disk (either for security reasons or if it's a very small set of data of known size which changes very frequently.) /tmp meets neither of those requirements. (If data on /tmp is actually sensitive, then you should encrypt the volume--or preferably the entire root filesystem.)

    @Mcoder said:

    Again, what? That's the builtin functionality.

    By built-in functionality I mean "writing to disk like normal". I thought that was somewhat clear.

    @Mcoder said:

    Well, it was the other way around. Somebody said "hey, we don't need persistence at /tmp, let's exploit that to gain some speed", and created tmpfs.

    What? No, you're wrong. It wasn't created as a backing-store for /tmp. And the only way you're gaining speed is if you have a lot of small files that change very rapidly, which shouldn't be happening in /tmp in the first place.



  • @morbiuswilters said:



    @Mcoder said:

    Well, it was the other way around. Somebody said "hey, we don't need persistence at /tmp, let's exploit that to gain some speed", and created tmpfs.

    What? No, you're wrong. It wasn't created as a backing-store for /tmp. And the only way you're gaining speed is if you have a lot of small files that change very rapidly, which shouldn't be happening in /tmp in the first place.

     

    Well, technically the argument is sound; some files simply don't need persistence. However, the /tmp was meant to house non-persistent files that were larger than your RAM. If you want to store temporary files to tmpfs, then you have to KNOW that those files are smaller in total than your RAM. 

    Personally, I put into tmpfs the /var partition.

    Also, I put to tmpfs object files on my build server when the build has to be a clean full build. And moc-files on Qt projects. Because I don't want those files  to take disk-I/O that would be better spent reading the source and writing the resulting binaries.

    The kernel would cache those files for quick re-use in any case, but it would also try to write them to disk while the compilation process was still reading in source files.



  • Why doesn't the OS just do this bullshit for you without you having to think about it?



  • @OldCrow said:

    Well, technically the argument is sound; some files simply don't need persistence.

    I don't dispute that tmpfs has its legitimate uses, just that slapping /tmp on there makes any sense.

    @OldCrow said:

    Personally, I put into tmpfs the /var partition.

    Huh. Most of my var is persistent, so this would be bad. Also, my var is several GB, so also bad. It also seems like a lot of hassle to get performance that is probably the same (and possibly worse) than just letting the kernel handle disk cache for you.

    @OldCrow said:

    Also, I put to tmpfs object files on my build server when the build has to be a clean full build. And moc-files on Qt projects. Because I don't want those files  to take disk-I/O that would be better spent reading the source and writing the resulting binaries.

    The kernel would cache those files for quick re-use in any case, but it would also try to write them to disk while the compilation process was still reading in source files.

    That actually seems like a legit usage because you have truly transient data that you don't want getting synced to disk during a large, IO-bound task.



  • @blakeyrat said:

    Why doesn't the OS just do this bullshit for you without you having to think about it?

    That was my point: the kernel does do this for you, and it does it better than most people will do on their own. Sticking files on tmpfs is an "enhancement" that often results in the same or worse performance than you'd get normally, usually with extra work and extra chance something's going to go wrong.



  • @OldCrow said:

    Personally, I put into tmpfs the /var partition.
     

    The WHAT? You mount /var as a temporary filesystem?

    What the hell do you use /var for - or rather, what don't you use it for? I hope I've misunderstood you and /var isn't tmpfs.



  • @Cassidy said:

    @OldCrow said:

    Personally, I put into tmpfs the /var partition.
     

    The WHAT? You mount /var as a temporary filesystem?

    What the hell do you use /var for - or rather, what don't you use it for? I hope I've misunderstood you and /var isn't tmpfs.

    Oh, don't be such a wuss. System logs are really easy to regenerate.



  • @Ben L. said:

    @Cassidy said:

    @OldCrow said:

    Personally, I put into tmpfs the /var partition.
     

    The WHAT? You mount /var as a temporary filesystem?

    What the hell do you use /var for - or rather, what don't you use it for? I hope I've misunderstood you and /var isn't tmpfs.

    Oh, don't be such a wuss. System logs are really easy to regenerate.

    I have a lot more than logs on /var. I have mail spools, working files for dozens of programs, caches for tons of stuff (which technically can be regenerated, but to do it on every single boot?)



  • @Cassidy said:

    @OldCrow said:

    Personally, I put into tmpfs the /var partition.
     

    The WHAT? You mount /var as a temporary filesystem?

    What the hell do you use /var for - or rather, what don't you use it for? I hope I've misunderstood you and /var isn't tmpfs.

     

    Oops. Meant to say that /run was mounted as tmpfs. Not /var. That would indeed be crazy.

    Maybe I shouldn't write to forums at 02:00AM.

     

    Edit:

    P.S.:

    I have contemplated moving /var/log to tmpfs, but decided that it would defeat the purpose of having logs in the first place. Then again, some logs you don't need to preserve through reboots. But if something writes a log often enough that this makes a difference, then that something is broken already.

     



  • Debian (at least the version I have running atm) doesn't put /tmp on a separate partition, just a part of it's main partition, mounted on /.

    @morbiuswilters said:

    I have a lot more than logs on /var. I have mail spools, working files for dozens of programs, caches for tons of stuff (which technically can be regenerated, but to do it on every single boot?)

    Hmm now if only I had a way to prevent the mail spool from being written to in the first place. It's full of noise that cron put there.

    Just make sure you don't have anything along the lines of Apache serving from /var/www/html. I'm glad Ubuntu's new Click package format is a bold step towards not spewing crap all over the file system (each application in it's own independent directory).



  • @MiffTheFox said:

    Debian (at least the version I have running atm) doesn't put /tmp on a separate partition, just a part of it's main partition, mounted on /.

    @morbiuswilters said:

    I have a lot more than logs on /var. I have mail spools, working files for dozens of programs, caches for tons of stuff (which technically can be regenerated, but to do it on every single boot?)

    Hmm now if only I had a way to prevent the mail spool from being written to in the first place. It's full of noise that cron put there.

    Just make sure you don't have anything along the lines of Apache serving from /var/www/html. I'm glad Ubuntu's new Click package format is a bold step towards not spewing crap all over the file system (each application in it's own independent directory).

    My HTML files are in /usr/share/nginx/html. AFAIK YMMV HTH HAND

  • Discourse touched me in a no-no place

    @blakeyrat said:

    Why doesn't the OS just do this bullshit for you without you having to think about it?
    Depends on whether the OS came set up that way (in which case you've just got exposure of detail for the obsessed) or whether he did it himself (an attack of excessive tinkering disease). In any case, meh. Any serious software that wants to write very large temporary files will let you tell it where to do so.



  • @MiffTheFox said:

    Hmm now if only I had a way to prevent the mail spool from being written to in the first place. It's full of noise that cron put there.

    cron only sends out mail when there's output. So you can just add >/dev/null 2>&1 to the end of all of your cron entries and they won't send email.

    @MiffTheFox said:

    I'm glad Ubuntu's new Click package format is a bold step towards not spewing crap all over the file system (each application in it's own independent directory).

    This actually sounds like a half-way decent idea from Ubuntu, for once. I hate how packages scatter shit all over the place. Especially when they have hard-coded paths and you want to run multiple instances or versions side-by-side, but then you have to copy files all over the place.. ugh.



  • @Ben L. said:

    My HTML files are in /usr/share/nginx/html. AFAIK YMMV HTH HAND

    Mine are in /www/$DOMAIN_NAME/ Less typing than a deeply-nested directory.



  • @renewest said:

    This is what Asus wrote to me:

    <font id="yui_3_7_2_1_1368123334610_2039" face="Arial">This survey aims to understand your level of satisfaction and expectations towards ASUS and its products. </font>

    <font id="yui_3_7_2_1_1368123334610_2035" face="Arial">Your opinions will help shape the future of ASUS and you'll stand a chance to receive <font id="yui_3_7_2_1_1368123334610_2036" face="Arial">one Asus 500G USB 2.5inch Hard </font>Drive upon completion of this short 25-30 minute survey. </font>

     This just made me think WTF!

     

     Taking the price at $40, this is $80-$96 per hour - .nearly $200K per year if full time and with a 99% chance of being tax free......not bad for spare time filling in surveys..



  • @TheCPUWizard said:

    @renewest said:

    This is what Asus wrote to me:

    <font id="yui_3_7_2_1_1368123334610_2039" face="Arial">This survey aims to understand your level of satisfaction and expectations towards ASUS and its products. </font>

    <font id="yui_3_7_2_1_1368123334610_2035" face="Arial">Your opinions will help shape the future of ASUS and you'll stand a chance to receive <font id="yui_3_7_2_1_1368123334610_2036" face="Arial">one Asus 500G USB 2.5inch Hard </font>Drive upon completion of this short 25-30 minute survey. </font>

     This just made me think WTF!

     

     Taking the price at $40, this is $80-$96 per hour - .nearly $200K per year if full time and with a 99% chance of being tax free......not bad for spare time filling in surveys..

    Except, of course, that your chances of actually getting anything are probably 1 in 5000. So it works out to about $40 /year if you did this full time.



  • @TheCPUWizard said:

    @renewest said:

    This is what Asus wrote to me:

    <font id="yui_3_7_2_1_1368123334610_2039" face="Arial">This survey aims to understand your level of satisfaction and expectations towards ASUS and its products. </font>

    <font id="yui_3_7_2_1_1368123334610_2035" face="Arial">Your opinions will help shape the future of ASUS and you'll stand a chance to receive <font id="yui_3_7_2_1_1368123334610_2036" face="Arial">one Asus 500G USB 2.5inch Hard </font>Drive upon completion of this short 25-30 minute survey. </font>

     This just made me think WTF!

     

     Taking the price at $40, this is $80-$96 per hour - .nearly $200K per year if full time and with a 99% chance of being tax free......not bad for spare time filling in surveys..

    Yeah, except it's a CHANCE of getting the hard drive. You don't get the hard drive. You get the feeling that you MIGHT have gotten a hard drive.



  • Sure, I miss one little word.......

     You are all of course, correct....



  • @Ben L. said:

    My HTML files are in /usr/share/nginx/html. AFAIK YMMV HTH HAND
     

    That must be a pain to type when you want to get to that directory from the command line.  "cd /usr/share/nginx/html.\ AFAIK\ YMMV\ HTH\ HAND" seems like it'd be tiring to type every time you need to update "my first php project 1.0"

    I stick to a real man's path: C:\inetpub\filesroot.

     



  • @drurowin said:

    @Ben L. said:

    My HTML files are in /usr/share/nginx/html. AFAIK YMMV HTH HAND
     

    That must be a pain to type when you want to get to that directory from the command line.  "cd /usr/share/nginx/html.\ AFAIK\ YMMV\ HTH\ HAND" seems like it'd be tiring to type every time you need to update "my first php project 1.0"

    I stick to a real man's path: C:\inetpub\filesroot.

     

    Why would I install PHP? I write all my webapp in Go.



  •  I find at least 4 or 5 things wrong with that URL.

    Also, when I see people talking about the Go language, why do I think the name should expand to "Go fuck yourself"?



  • @Ben L. said:

    http://nightgunner5.is-a-geek.net:1337/l4d2

    Port 1337? Please don't tell me that Go doesn't follow Node.js into the "mature httpds are boring, let's write our own" cult.


  • Discourse touched me in a no-no place

    @morbiuswilters said:

    @MiffTheFox said:
    Hmm now if only I had a way to prevent the mail spool from being written to in the first place. It's full of noise that cron put there.

    cron only sends out mail when there's output. So you can just add >/dev/null 2>&1 to the end of all of your cron entries and they won't send email.

    Or you could simply tell cron to simply stop sending the mail in the first place if you're not interested in any of the output.



  • @morbiuswilters said:

    I hate how packages scatter shit all over the place. Especially when they have hard-coded paths and you want to run multiple instances or versions side-by-side, but then you have to copy files all over the place.. ugh.
    For this reason alone I wish GoboLinux had had more success, or at least had caused an impact on Filesystem Hierarchy Standard.



  • @MiffTheFox said:

    @Ben L. said:
    http://nightgunner5.is-a-geek.net:1337/l4d2

    Port 1337? Please don't tell me that Go doesn't follow Node.js into the "mature httpds are boring, let's write our own" cult.

    The Go server is running on port 8482. This is just a nginx server that handles logging, etc.


Log in to reply
 

Looks like your connection to What the Daily WTF? was lost, please wait while we try to reconnect.