• srtjjk (unregistered) in reply to QJo
    QJo:
    The real WTF is naturally to expect veteran experienced developers to be instantly familiar with all aspects of the IT world the very moment it becomes available.

    "Master developers" with "years of experience" may not even know that they don't know what they're doing with a particular aspect of a technology -- but have enough of a pragmatic approach to get something working (which is why they have reputations for being "masters").

    It's a matter of learning opportunities. Having found that line in the .bashrc file, it is then incumbent upon Alex to communicate to the master developers why that is the wrong thing to do, and what is the right thing to do.

    So tell me (I know enough about Linux to be able to hack together bash scripts if I need to, and I have a working knowledge of what the various directories are for, and I know how to use man), why is it wrong to configure a network redirect in the .bashrc file?

    No, the real WTF is: "Alex died a little that day". Would it not be a wholly more appropriate response to come down off his know-it-all high horse and understand that not everybody has had the opportunity of an expensive education in computer science in an institution of learning? Some of us pick up our knowledge through on-the-job experience.

    I'm guessing the person who "joked" about it is the person who did it....

  • (cs) in reply to Jeremy
    Jeremy:
    QJo:
    Jeremy:
    QJo:
    The real WTF is naturally to expect veteran experienced developers to be instantly familiar with all aspects of the IT world the very moment it becomes available.

    "Master developers" with "years of experience" may not even know that they don't know what they're doing with a particular aspect of a technology -- but have enough of a pragmatic approach to get something working (which is why they have reputations for being "masters").

    It's a matter of learning opportunities. Having found that line in the .bashrc file, it is then incumbent upon Alex to communicate to the master developers why that is the wrong thing to do, and what is the right thing to do.

    So tell me (I know enough about Linux to be able to hack together bash scripts if I need to, and I have a working knowledge of what the various directories are for, and I know how to use man), why is it wrong to configure a network redirect in the .bashrc file?

    No, the real WTF is: "Alex died a little that day". Would it not be a wholly more appropriate response to come down off his know-it-all high horse and understand that not everybody has had the opportunity of an expensive education in computer science in an institution of learning? Some of us pick up our knowledge through on-the-job experience.

    You think those of us with the highfalutin CS degrees earned them by learning about ip configurations and bash scripting?

    You'd be surprised. I watched over the (figurative) shoulder of a participant in an IT course once, and one of the exercises was "Write a Unix shell script to do an admin-ish sort of function. Then make a menu to run several useful admin-ish sort of functions. Then do the same thing in DOS."

    Well, it was never my implication that those things were NEVER taught, and certainly never learned in passing, but if Mr. Highhorse there thinks a CS degree is "learn how to configure a server 201" and so on then he's wrong, at least if my university is any guide.

    Far be it from me to be rude about a person I've never met, but Jeremy is a stupid fucking prick.

    The way I read this is that QJo is pointing out that, even if you've been in the business some time, you can't know everything. On the other hand, people who are new to the business, who have just finished a course of learning (who gives a fucking shithead bugger if that fucking course is called "computer science" or "IT" or "plugging in boxes") fact is these little shithead brats come flouncing in thinking they know every fucking thing. And maybe they do know every fucking thing that was presented to them neatly on a little fucking (metaphorical) plate, and so they've got a job where they can exercise this stuff they've learned.

    And, haw! haw! haw! Those developers who think they're sooooooo fucking clever and sooooo fucking important can't even edit Unix code properly! What a bunch of fucking dicks!

  • Angered Dev (unregistered) in reply to Mark
    Mark:
    QJo:
    hobbes:
    This is why I hate supporting developers. They thing that because they're good at programming, they're good at everything with a keyboard. Oh, and that they shouldn't be subject to any company policy related to security.

    Think how wonderful life would be if there weren't any pesky developers filling up our nice shiny hardware with those disgusting applications. Let's just fire the lot and then we will be certain that the servers will continue to run just as efficiently as they did on the day they were bought.

    Yep, the real WTF is the company paying the salaries of people whose job it is to build applications to provide a service to customers. Eww.

    Your comment makes no sense. No where did hobbes say that all developers should be fired. He made an, in my experience justified, assertion that some developers will assume that their experience as programmers makes them better sysadmins than the actual sysadmins. Also in my experience, this assumption is disastrously false.

    It is called separate domains of knowledge. Don't assume that knowledge in one (programming) automatically bestows knowledge in another (sysadmin) just because both involve computers. From what I've seen, programmers are second only to academics in this arrogance.

    Wrong! Helpdesk stuff are far, far, far more arrogant and assume they have expertise in absolutely everything. Lawyers are pretty bad for think they're experts in everything too.....

  • Angered Dev (unregistered) in reply to QJo
    QJo:
    Mark:
    QJo:
    hobbes:
    This is why I hate supporting developers. They thing that because they're good at programming, they're good at everything with a keyboard. Oh, and that they shouldn't be subject to any company policy related to security.

    Think how wonderful life would be if there weren't any pesky developers filling up our nice shiny hardware with those disgusting applications. Let's just fire the lot and then we will be certain that the servers will continue to run just as efficiently as they did on the day they were bought.

    Yep, the real WTF is the company paying the salaries of people whose job it is to build applications to provide a service to customers. Eww.

    Your comment makes no sense. No where did hobbes say that all developers should be fired. He made an, in my experience justified, assertion that some developers will assume that their experience as programmers makes them better sysadmins than the actual sysadmins. Also in my experience, this assumption is disastrously false.

    It is called separate domains of knowledge. Don't assume that knowledge in one (programming) automatically bestows knowledge in another (sysadmin) just because both involve computers. From what I've seen, programmers are second only to academics in this arrogance.

    In which case you ought to offer some slack to those developers who, in the interest of Getting The Job Done, implement solutions which to you are cringeingly bad, because they are not experts in your field.

    The boundary between configuring the application to work in the environment it is placed and performing sysadmin tasks is not well-defined. You frequently find that developers who are uncertain about what they may or may not control are more likely to stray into the area that they are not "supposed" to because (being self-starting, learn-on-the-job, achieve-the-results kind of guys) they find it's often quicker and easier to do the job themselves than rely on an outside agency (to them, internal IT is an outside agency), especially when they know the IT guys consider them with distaste (some even go so far as to call developers "arrogant" in public fora), and are less than likely to make more than a nominal effort to provide the appropriate level of support.

    In this situation, the ones of fault are of course the IT guys who haven't configured the appropriate level of authoriations so as to demark the IT tasks to the IT team, leaving the developers to work on the tasks which they have been specifically assigned.

    Also (rightly or wrongly) dev's often take the view that (almost) everything is a program written by dev's, and by extension devs must therefore understand how it works.....it sort of almost nearly actually seems to make sense.
    NB: I'm not syaing they're right.

  • (cs) in reply to Angered Dev
    Matt Westwood:
    Far be it from me to be rude about a person I've never met

    haha

    Angered Dev:
    Wrong! Helpdesk stuff are far, far, far more arrogant and assume they have expertise in absolutely everything. Lawyers are pretty bad for think they're experts in everything too.....

    Don't forget about car salesmen.

  • Mick (unregistered) in reply to Anon
    Anon:
    hobbes:
    This is why I hate supporting developers. They thing that because they're good at programming, they're good at everything with a keyboard. Oh, and that they shouldn't be subject to any company policy related to security.

    Not all are like this. There are some who appreciate the concept of separate spheres of knowledge. But most? Ya.

    Devs shouldn't be subject to MOST company policy related to security.

    The car-mechanic equivalent would be requiring that all your mechanics use metric wrenches, no matter the car on which they're working.

    You can either have effective software developers, who have full control over their local system, or you can have devs who quit for ANY other pastures at the first opportunity.

    No one of any decent level of talent wants to work at a company where they're constantly hamstrung by "security."

    Especially when a solid 99% of the security at most orgs is a set of utterly ineffective, inefficient, counter-productive 'rules,' rather than real, inexpensive, effective physical security devices.

    Disagree. There may be elements of security that devs are exempt from, but I wouldn't think it's MOST. A dev environment (and perhaps a Dev Test one) should probably have minimal security, allowing devs to set up any element of the system how they want, but other environments need to be kept secure - if nothing else to ensure that only the right groups of people are fixing the problem. If a Support Dev believes they have a fix that they have insufficient priviliges to fix a problem, then they should be engaging the appropriate group (irrespective how tedious that is). If they need to change DB schema (in a prod environment) they need to use a DBA; if they need to modify system settings, they need a SysAdmin; etc.

    The more dev's are allowed to be cowboys in a working (prod)environment, the more issues there will be down the track when noone has any records of how a particular issue was circumvented - and perhaps today's WTF is a refelction of exactly that....

  • didjabringyagrogalong (unregistered) in reply to Snooder
    Snooder:
    QJo:
    Jeremy:
    QJo:
    The real WTF is naturally to expect veteran experienced developers to be instantly familiar with all aspects of the IT world the very moment it becomes available.

    "Master developers" with "years of experience" may not even know that they don't know what they're doing with a particular aspect of a technology -- but have enough of a pragmatic approach to get something working (which is why they have reputations for being "masters").

    It's a matter of learning opportunities. Having found that line in the .bashrc file, it is then incumbent upon Alex to communicate to the master developers why that is the wrong thing to do, and what is the right thing to do.

    So tell me (I know enough about Linux to be able to hack together bash scripts if I need to, and I have a working knowledge of what the various directories are for, and I know how to use man), why is it wrong to configure a network redirect in the .bashrc file?

    No, the real WTF is: "Alex died a little that day". Would it not be a wholly more appropriate response to come down off his know-it-all high horse and understand that not everybody has had the opportunity of an expensive education in computer science in an institution of learning? Some of us pick up our knowledge through on-the-job experience.

    You think those of us with the highfalutin CS degrees earned them by learning about ip configurations and bash scripting?

    You'd be surprised. I watched over the (figurative) shoulder of a participant in an IT course once, and one of the exercises was "Write a Unix shell script to do an admin-ish sort of function. Then make a menu to run several useful admin-ish sort of functions. Then do the same thing in DOS."

    CS != IT.

    Genius!

  • Norman Diamond (unregistered) in reply to QJo
    QJo:
    The real WTF is naturally to expect veteran experienced developers to be instantly familiar with all aspects of the IT world the very moment it becomes available.
    Bingo. I'd say an expert brain surgeon might be a master in addition to being a doctor, and an expert heart surgeon might be a master to being a doctor, but I wouldn't try making them do each other's job.

    But... but...

    QJo:
    So tell me (I know enough about Linux to be able to hack together bash scripts if I need to, and I have a working knowledge of what the various directories are for, and I know how to use man), why is it wrong to configure a network redirect in the .bashrc file?
    Because maybe the servers should provide services to clients when the servers get booted, and they shouldn't have to wait for someone to log in ... but, um, what kind of network redirection are you talking about?
  • Boat Glower (unregistered) in reply to Matt Westwood
    Matt Westwood:
    Jeremy:
    QJo:
    Jeremy:
    QJo:
    The real WTF is naturally to expect veteran experienced developers to be instantly familiar with all aspects of the IT world the very moment it becomes available.

    "Master developers" with "years of experience" may not even know that they don't know what they're doing with a particular aspect of a technology -- but have enough of a pragmatic approach to get something working (which is why they have reputations for being "masters").

    It's a matter of learning opportunities. Having found that line in the .bashrc file, it is then incumbent upon Alex to communicate to the master developers why that is the wrong thing to do, and what is the right thing to do.

    So tell me (I know enough about Linux to be able to hack together bash scripts if I need to, and I have a working knowledge of what the various directories are for, and I know how to use man), why is it wrong to configure a network redirect in the .bashrc file?

    No, the real WTF is: "Alex died a little that day". Would it not be a wholly more appropriate response to come down off his know-it-all high horse and understand that not everybody has had the opportunity of an expensive education in computer science in an institution of learning? Some of us pick up our knowledge through on-the-job experience.

    You think those of us with the highfalutin CS degrees earned them by learning about ip configurations and bash scripting?

    You'd be surprised. I watched over the (figurative) shoulder of a participant in an IT course once, and one of the exercises was "Write a Unix shell script to do an admin-ish sort of function. Then make a menu to run several useful admin-ish sort of functions. Then do the same thing in DOS."

    Well, it was never my implication that those things were NEVER taught, and certainly never learned in passing, but if Mr. Highhorse there thinks a CS degree is "learn how to configure a server 201" and so on then he's wrong, at least if my university is any guide.

    Far be it from me to be rude about a person I've never met, but Jeremy is a stupid fucking prick.

    The way I read this is that QJo is pointing out that, even if you've been in the business some time, you can't know everything. On the other hand, people who are new to the business, who have just finished a course of learning (who gives a fucking shithead bugger if that fucking course is called "computer science" or "IT" or "plugging in boxes") fact is these little shithead brats come flouncing in thinking they know every fucking thing. And maybe they do know every fucking thing that was presented to them neatly on a little fucking (metaphorical) plate, and so they've got a job where they can exercise this stuff they've learned.

    And, haw! haw! haw! Those developers who think they're sooooooo fucking clever and sooooo fucking important can't even edit Unix code properly! What a bunch of fucking dicks!

    I blame the varsities - or perhaps the people designing the curriculum at those varsities. The attitude the graduates have is handed them from their teachers. One of the problems is that teachers are often (though not exclusively) Academics. One of the problems with grads is that they believe the textbook. Now, I'm not bagging the text book per se', but in context the text book is at best "how things work in an ideal world". This has often been revised by numerous academics each of who have some great theories on all sorts of stuff. Bottom line, the text book is usually wrong - especially if your scenario is anything but EXACTLY what is in the text book (yes, that's right "You are writing a system to track university staff and student, and you must use Java and not worry about how well it would scale up" doesn't necessarily apply to "You a working in a team whose project relates to designing and creating an actual functioning system for a college to manage administrative tasks of the university, including keeping a record of all students, the courses they are enrolled in, and how high a salary each lecturer gets.").

    Anyway, the point is thus: Most graduates don't understand that they graduate (thereticallY) equipped with the skills they might need to be able to investigate and solve problems, but rather think that they graduate equipped with all of the knowledge they need for an entire career in their chosen discipline. This is not specific to Mathematical/Engineering/Sciences - this is all graduates. It seems like it affects IT peeps more, because chances are you work in IT and so you're running into the IT ones more regularly.....

  • Norman Diamond (unregistered) in reply to JimM
    JimM:
    eric76:
    I'm still puzzled why the developers would be logging into root.

    "were given root access" != logging into root

    I'm not even sure how you would log "into" root... Oo

    me> su - password: password

    Or as I had to do in a now-obsolete version of Knoppix: $ sudo su -

  • Norman Diamond (unregistered) in reply to john
    john:
    PS. You have to "down" the interface first before "up" with a new address, unless you use a sub-interface
    Who were you talking to? You weren't talking to me. I don't have to "down" an interface first before "up" with a new address. But I'm just a developer.
  • Jeff Dege (unregistered)

    If the network admins were competent, they would have established a section of IP addresses, within the DHCP configuration of the development net, specifically for locally-assigned static IP addresses.

    There are reasons why a developer might want - or need - to assign one or more static IP addresses, and why those needs might be far too ephemeral for it to be reasonable to push those assignments through some external IT organization.

    The right answer is to give the each team a sandbox to play in, so that their playing doesn't interfere with the larger organization. And to let them fight out any conflicts that might arise within the sandbox among themselves.

    If the developer used an IP address outside the sandbox, I'd blame the developer. If there was no sandbox, I'd blame the network admin.

  • (cs) in reply to Matt Westwood
    Matt Westwood:
    Far be it from me to be rude about a person I've never met ...

    Matt? You feeling OK?

  • (cs) in reply to Matt Westwood
    Matt Westwood:
    Jeremy:
    QJo:
    Jeremy:
    QJo:
    The real WTF is naturally to expect veteran experienced developers to be instantly familiar with all aspects of the IT world the very moment it becomes available.

    "Master developers" with "years of experience" may not even know that they don't know what they're doing with a particular aspect of a technology -- but have enough of a pragmatic approach to get something working (which is why they have reputations for being "masters").

    It's a matter of learning opportunities. Having found that line in the .bashrc file, it is then incumbent upon Alex to communicate to the master developers why that is the wrong thing to do, and what is the right thing to do.

    So tell me (I know enough about Linux to be able to hack together bash scripts if I need to, and I have a working knowledge of what the various directories are for, and I know how to use man), why is it wrong to configure a network redirect in the .bashrc file?

    No, the real WTF is: "Alex died a little that day". Would it not be a wholly more appropriate response to come down off his know-it-all high horse and understand that not everybody has had the opportunity of an expensive education in computer science in an institution of learning? Some of us pick up our knowledge through on-the-job experience.

    You think those of us with the highfalutin CS degrees earned them by learning about ip configurations and bash scripting?

    You'd be surprised. I watched over the (figurative) shoulder of a participant in an IT course once, and one of the exercises was "Write a Unix shell script to do an admin-ish sort of function. Then make a menu to run several useful admin-ish sort of functions. Then do the same thing in DOS."

    Well, it was never my implication that those things were NEVER taught, and certainly never learned in passing, but if Mr. Highhorse there thinks a CS degree is "learn how to configure a server 201" and so on then he's wrong, at least if my university is any guide.

    Far be it from me to be rude about a person I've never met, but Jeremy is a stupid fucking prick.

    The way I read this is that QJo is pointing out that, even if you've been in the business some time, you can't know everything. On the other hand, people who are new to the business, who have just finished a course of learning (who gives a fucking shithead bugger if that fucking course is called "computer science" or "IT" or "plugging in boxes") fact is these little shithead brats come flouncing in thinking they know every fucking thing. And maybe they do know every fucking thing that was presented to them neatly on a little fucking (metaphorical) plate, and so they've got a job where they can exercise this stuff they've learned.

    And, haw! haw! haw! Those developers who think they're sooooooo fucking clever and sooooo fucking important can't even edit Unix code properly! What a bunch of fucking dicks!

    And Jeremy's point is that anyone who thinks a CS degree is in about being a Unix sysadmin is utterly fucking mistaken. Worse, he's not just mistaken, his ignorance is the sort of bullshit that keeps WTF developers with shit degrees from half-ass schools floating around, and depresses the pay scale for better developers who actually learned the things they need to learn, and not just how to type the words into the black box.

    Any school that offers a computer science degree that is basically a thinly veiled technical support certification course ought to closed down for fraud. Call it Network Administration. Call it Computing Fundamentals. Call it Infomatics. Call it whether the fuck you want to, just don't call it something that confuses the employment landscape. Because it really IS important what you spent 4+ years learning.

  • The Truth, The Hole Truth, And Nothing Butt (unregistered)

    So he get's a request, probably takes an age to do it and when he does he does it wrong. Any wonder that people resort to using other routes to get the work done?

  • tehslayor (unregistered)

    Posts here are usually fun to read and you ever learn interesting stuff sometimes. But for the love of god, please use pronouns. It's like reading a child story.

  • Wrexham (unregistered) in reply to Sam
    Sam:
    JimM:
    eric76:
    I'm still puzzled why the developers would be logging into root.
    "were given root access" != logging into root

    I'm not even sure how you would log "into" root... Oo

    Well I can think of a number of ways, here's 5 off the top of my head:

    1. At the login prompt (assuming you allow physical access)
    2. Over SSH (if you are insane and allow this)
    3. Using the 'su - ' (assuming you can execute su)
    4. Using 'sudo -s' (assuming you have sudoers rights)
    5. Using 'sudo su -' (assuming you have sudoers rights)
    Norman Diamond:
    JimM:
    I'm not even sure *how* you would log "into" root... Oo
    me> su - password: password #

    Or as I had to do in a now-obsolete version of Knoppix: $ sudo su - #

    I think that JimM was quibbling about eric's use of "logging into root". I assume he would prefer "logging in to root" instead. He probably didn't mean to suggest that it's impossible to login to root.

  • boogerboy (unregistered) in reply to Wrexham
    Wrexham:
    Sam:
    JimM:
    eric76:
    I'm still puzzled why the developers would be logging into root.
    "were given root access" != logging into root

    I'm not even sure how you would log "into" root... Oo

    Well I can think of a number of ways, here's 5 off the top of my head:

    1. At the login prompt (assuming you allow physical access)
    2. Over SSH (if you are insane and allow this)
    3. Using the 'su - ' (assuming you can execute su)
    4. Using 'sudo -s' (assuming you have sudoers rights)
    5. Using 'sudo su -' (assuming you have sudoers rights)
    Norman Diamond:
    JimM:
    I'm not even sure *how* you would log "into" root... Oo
    me> su - password: password #

    Or as I had to do in a now-obsolete version of Knoppix: $ sudo su - #

    I think that JimM was quibbling about eric's use of "logging into root". I assume he would prefer "logging in to root" instead. He probably didn't mean to suggest that it's impossible to login to root.
    or perhaps logging in AS root.....

  • robbak (unregistered) in reply to anonymous
    anonymous:
    eric76:
    I'm still puzzled why the developers would be logging into root.

    Well, they may need to install some extra packages during development. Granting them root access can sometimes be reasonable. (However I would normally granting them apt-get/yum with sudo only.)

    I can work with that. I'll create a dummy package with a configure script that runs bash, and then returns false. Either that or a package that installs a copy of bash suid.

  • Bill C. (unregistered) in reply to Wrexham
    Wrexham:
    Sam:
    JimM:
    eric76:
    I'm still puzzled why the developers would be logging into root.
    "were given root access" != logging into root

    I'm not even sure how you would log "into" root... Oo

    Well I can think of a number of ways, here's 5 off the top of my head:

    1. At the login prompt (assuming you allow physical access)
    2. Over SSH (if you are insane and allow this)
    3. Using the 'su - ' (assuming you can execute su)
    4. Using 'sudo -s' (assuming you have sudoers rights)
    5. Using 'sudo su -' (assuming you have sudoers rights)
    Norman Diamond:
    JimM:
    I'm not even sure *how* you would log "into" root... Oo
    me> su - password: password #

    Or as I had to do in a now-obsolete version of Knoppix: $ sudo su - #

    I think that JimM was quibbling about eric's use of "logging into root". I assume he would prefer "logging in to root" instead. He probably didn't mean to suggest that it's impossible to login to root.
    Logging in to root was standard back in my day, signing a log book in the lobby of the girls' dorm.

  • (cs) in reply to chubertdev
    chubertdev:
    Also, TRWTF is "recieved"
    As not a native English writer, I'm somewhat surprised at how often this mistake is made even by native writers. It's not like learning to spell this word is that big an acheivement.
  • (cs) in reply to Zecc
    Zecc:
    chubertdev:
    Also, TRWTF is "recieved"
    As not a native English writer, I'm somewhat surprised at how often this mistake is made even by native writers. It's not like learning to spell this word is that big an acheivement.
    or dyslexia coupled with not using a spell check enabled editor
  • nasch (unregistered) in reply to Snooder
    Snooder:
    Because it really IS important *what* you spent 4+ years learning.

    I heard about some research that found zero correlation between the quality of a software developer and whether s/he had a college degree. I don't remember which public radio show it was on, though.

  • agbeladem (unregistered)

    I too heard somewhere that people who comment on posts on internet are all idiots. I don't know how relevant it is however, and I just can't seem to remember where exactly I heard that.

  • nmclean (unregistered) in reply to lauren
    lauren:
    nmclean:
    Everything you said could be said about just about any post on this website. It normally requires specialized knowledge or experience to understand why something is a WTF. Why is this different from writing inefficient algorithms and unmanageable spaghetti code? Because it isn't your area of expertise?

    I think (and this is a guess) that the difference in this WTF and most others on this site is that the "WTF" in question was an error made by someone who was working outside their field of expertise. Most WTF are WTF-y because it's someone (usually a "guru") with supposedly tons of experience coding making basic coding errors. Whereas this is developers who aren't necessarily experts in network admin making a networking error (which is probably why the company hired some networking guys?) Alternatively, it could have been a WTF if "Alex" had then confronted the dev about the mistake only to have the dev give him some ridiculous justification for why he wanted it that way.

    Basically, it's funny when people who should know better make mistakes - but when people who have no reason to know better make mistakes would it not be better to help them learn instead of sighing and chalking it up to ineptitude?

    Yes you should help them learn; my point is that is true for every post on this site. Every WTF boils down to a lack of experience, regardless of whether the perpetrator claims to be a "guru". I don't believe it's the level of supposed qualification that determines whether it's funny. It's simply funny when the reader knows better.

    The problem is the hypocrisy of becoming offended when you don't get it. Somehow I doubt that QJo, a developer, would be writing the same defense if this post was about a script containing laughably poor coding practices and logic errors, written by a Linux admin.

  • Tux "Tuxedo" Penguin (unregistered) in reply to JimM
    JimM:
    I'm not even sure *how* you would log "into" root... Oo

    First you log the (directory) tree, then dig up the root, dummy! :P Also no big wtf here, have seen worse (one disgruntled sysadmin has set up start up scripts so it would constantly kill core processes leading to kernel panic. I had to connect server's HDD to another computer and chroot into it to fix the issue).

    Captcha: incassum. Incassum you didn't know My name means that I am employed as sysadmin in small company using certain variant of Unix that has many subvariants.

  • (cs) in reply to snoofle
    snoofle:
    Zecc:
    chubertdev:
    Also, TRWTF is "recieved"
    As not a native English writer, I'm somewhat surprised at how often this mistake is made even by native writers. It's not like learning to spell this word is that big an acheivement.
    or dyslexia coupled with not using a spell check enabled editor

    I'm surprised at how few people do this. It's a minor thing that completely changes how people percieve you. And I really need it with my lysdexia.

    hehe, "percieve"

  • (cs) in reply to nasch
    nasch:
    Snooder:
    Because it really IS important *what* you spent 4+ years learning.

    I heard about some research that found zero correlation between the quality of a software developer and whether s/he had a college degree. I don't remember which public radio show it was on, though.

    I'd believe it. I worked with a man who had two Master's degrees in whatever seemed relevant at the time, and couldn't code his way out of a wet paper bag.

  • Deanis (unregistered) in reply to chubertdev

    "I before E except after C"

  • foxyshadis (unregistered) in reply to elroy
    elroy:
    I used to do just that at work. I would use the super fast ethernet for most things, but I had to use a slow, public wifi connection to access my email through thunderbird. For some reason, that wasn't allowed on the wired connection even though we could access almost anything else online.
    Email viruses, phishing notices, and the like are still a very popular infection vector. I don't blame IT for wanting to mostly cut it off; at least internal email is guaranteed to be filtered at least once, usually twice. Remember all of the people who fell for the huge "UPS Package Tracking" scam that was going around last year? Personal email was one of the main ways it got into corporate networks.

    Obviously, with skill you can work around it, the hope is that the technical skills to do that translate into the technical skills to avoid infection.

  • foxyshadis (unregistered)

    Also regarding the central argument of this thread, everyone seems to be arguing from the stance that devs can't be IT and vice versa, when it's actually quite easy to be highly versed in both, especially if you regularly go back and forth and bother to do a little research from time to time. Most people just don't want to be, and if you don't want to be, you shouldn't really be playing with others' specialties too much.

    (Every enterprise application with a fragile, manual install procedure that breaks if anything's updated, customized, or non-English? That's devs playing IT and failing. Every "app" that's a collection of shell scripts that you have to place in specific locations and edit in a dozen copy-pasted places because they don't have command-line parsing, let alone GUIs? That's IT playing dev and failing.)

  • (cs) in reply to chubertdev
    chubertdev:
    snoofle:
    Zecc:
    chubertdev:
    Also, TRWTF is "recieved"
    As not a native English writer, I'm somewhat surprised at how often this mistake is made even by native writers. It's not like learning to spell this word is that big an acheivement.
    or dyslexia coupled with not using a spell check enabled editor

    I'm surprised at how few people do this. It's a minor thing that completely changes how people percieve you. And I really need it with my lysdexia.

    hehe, "percieve"

    I usually run it through a spell checker, but I wrote this one on my phone using the "note" applet; never got around to grabbing the submitted text and checking it later - mea culpa

  • (cs) in reply to Deanis
    Deanis:
    "I before E except after C"
    That is not always scientifically sufficient for our species, but then you computer-scientists should know that...
  • (cs) in reply to snoofle
    snoofle:
    I usually run it through a spell checker, but I wrote this one on my phone using the "note" applet; never got around to grabbing the submitted text and checking it later - mea culpa

    That explains a lot. When I'm bored (aka, the afternoon), I read older articles, and there's been a steady flow of Muphry's Law throughout them.

    Addendum (2014-03-11 16:17): (for all authors, not just you)

  • (cs)

    Maybe the WTF is same as at our place - windoze devs got charged with DNS such that machine "foo" was now known as "foo.AD.somebollox.wibble.dept.crap" and any attempt by the unix team to have a DNS alias "foo" entered was met with blank looks.

  • Alex (unregistered)

    Hi! I'm Alex. I'm surprised and flattered this got accepted as a Daily WTF since it happened quite a while ago; I was much greener and optimistic, and assumed every developer knew everything under the sun. I've since moved on to another company; I still consider myself very new in IT, but I'm slightly less all-assuming about others' knowledge in my field if their expertise lies elsewhere.

    My two cents:

    1. I changed their virtual NIC on that VM to another virtual network adapter, but that IP address range (/24, by the way, not a /16) was also using static IP addresses. I wasn't an admin, either, just a technician aspiring to adminship. (Still not quite there yet.) Every group has their own VM pool and IP address space, and I didn't realize I had placed their VM in the wrong pool/NIC because I had followed the direction of a senior, I was too new to know the intricacies of the company's setup, and it had worked at the time and for several months before all this happened.

    2. QJo: Hah, tell me about picking up skills on the job. I'm a liberal arts major; I'd say 95% of what I know now is knowledge I learned on the job. I love learning this stuff!

  • (cs) in reply to QJo

    So tell me (I know enough about Linux to be able to hack together bash scripts if I need to, and I have a working knowledge of what the various directories are for, and I know how to use man), why is it wrong to configure a network redirect in the .bashrc file?

    There's two of them. You have your user-local .bashrc and in most implementations there's a global .bashrc. For instance, if I wanted (redhat) all users to inherit the command "lc" I'd put it into /etc/bashrc whereas if only you want "lc" you'd put into your $HOME/.bashrc. It is complicated somewhat by .bash_profile. Without explanation you want to put your stuff into .bash_profile for a login shell (redhat). Once you can understand shellscript, looking at them will be obvious.

    It is in no way good to (in all instances I can think of) to designate network configurations in your bash login file. At the login prompt 'ifconfig' should reveal your network settings. If it doesn't (show some "eth") then you need to post into a group for your particular linux OS as to why.

    Ordinarily linux will (better than windoze) discover ethernet cards. If you're having to bring it up manually it's likely a 'modprobe' issue (ie getting the kernel to load it on boot).

    TIA Guy

  • QJo (unregistered) in reply to Deanis
    Deanis:
    "I before E except after C"
    ... said Sheila seizing Keith.
  • QJo (unregistered) in reply to Snooder
    Snooder:
    Matt Westwood:
    Jeremy:
    QJo:
    Jeremy:
    QJo:
    The real WTF is naturally to expect veteran experienced developers to be instantly familiar with all aspects of the IT world the very moment it becomes available.

    "Master developers" with "years of experience" may not even know that they don't know what they're doing with a particular aspect of a technology -- but have enough of a pragmatic approach to get something working (which is why they have reputations for being "masters").

    It's a matter of learning opportunities. Having found that line in the .bashrc file, it is then incumbent upon Alex to communicate to the master developers why that is the wrong thing to do, and what is the right thing to do.

    So tell me (I know enough about Linux to be able to hack together bash scripts if I need to, and I have a working knowledge of what the various directories are for, and I know how to use man), why is it wrong to configure a network redirect in the .bashrc file?

    No, the real WTF is: "Alex died a little that day". Would it not be a wholly more appropriate response to come down off his know-it-all high horse and understand that not everybody has had the opportunity of an expensive education in computer science in an institution of learning? Some of us pick up our knowledge through on-the-job experience.

    You think those of us with the highfalutin CS degrees earned them by learning about ip configurations and bash scripting?

    You'd be surprised. I watched over the (figurative) shoulder of a participant in an IT course once, and one of the exercises was "Write a Unix shell script to do an admin-ish sort of function. Then make a menu to run several useful admin-ish sort of functions. Then do the same thing in DOS."

    Well, it was never my implication that those things were NEVER taught, and certainly never learned in passing, but if Mr. Highhorse there thinks a CS degree is "learn how to configure a server 201" and so on then he's wrong, at least if my university is any guide.

    Far be it from me to be rude about a person I've never met, but Jeremy is a stupid fucking prick.

    The way I read this is that QJo is pointing out that, even if you've been in the business some time, you can't know everything. On the other hand, people who are new to the business, who have just finished a course of learning (who gives a fucking shithead bugger if that fucking course is called "computer science" or "IT" or "plugging in boxes") fact is these little shithead brats come flouncing in thinking they know every fucking thing. And maybe they do know every fucking thing that was presented to them neatly on a little fucking (metaphorical) plate, and so they've got a job where they can exercise this stuff they've learned.

    And, haw! haw! haw! Those developers who think they're sooooooo fucking clever and sooooo fucking important can't even edit Unix code properly! What a bunch of fucking dicks!

    And Jeremy's point is that anyone who thinks a CS degree is in about being a Unix sysadmin is utterly fucking mistaken. Worse, he's not just mistaken, his ignorance is the sort of bullshit that keeps WTF developers with shit degrees from half-ass schools floating around, and depresses the pay scale for better developers who actually learned the things they need to learn, and not just how to type the words into the black box.

    Any school that offers a computer science degree that is basically a thinly veiled technical support certification course ought to closed down for fraud. Call it Network Administration. Call it Computing Fundamentals. Call it Infomatics. Call it whether the fuck you want to, just don't call it something that confuses the employment landscape. Because it really IS important what you spent 4+ years learning.

    Yes, I may well be mistaken. Basically, I haven't got a clue what they teach on this subject in institutions of learning nowadays. My lack of knowledge in this area is completely irrelevant to the discussion -- whether you call it "computer science" or "IT helpdesk 101" has no bearing on the observation that brand-new "graduates" from such academies tend to have an incomplete overview of the various functions within an organisation -- and as such have not quite got their heads round the operational remit and general skill-sets of "developers".

    The very fact that said "developers" have not learned what the IT helpdesk staff did in school is something which causes irritation to said helpdesk staff.

  • anonymous (unregistered) in reply to Alex
    Alex:
    Hah, tell me about picking up skills on the job. I'm a liberal arts major; I'd say 95% of what I know now is knowledge I learned on the job. I love learning this stuff!
    Lucky you. If you were a CS major, probably 100% of what you know now would be knowledge you learned on the job. Liberal arts majors tend to already know how to read and write English...
  • Fake' Nagesh (unregistered) in reply to anonymous
    anonymous:
    Alex:
    Hah, tell me about picking up skills on the job. I'm a liberal arts major; I'd say 95% of what I know now is knowledge I learned on the job. I love learning this stuff!
    Lucky you. If you were a CS major, probably 100% of what you know now would be knowledge you learned on the job. Liberal arts majors tend to already know how to read and write English...

    Plain Rubbish! I know nothing about my job or my CS Skills, but I can still copy-pasta.

  • anonymous (unregistered) in reply to Fake' Nagesh
    Fake' Nagesh:
    anonymous:
    Alex:
    Hah, tell me about picking up skills on the job. I'm a liberal arts major; I'd say 95% of what I know now is knowledge I learned on the job. I love learning this stuff!
    Lucky you. If you were a CS major, probably 100% of what you know now would be knowledge you learned on the job. Liberal arts majors tend to already know how to read and write English...

    Plain Rubbish! I know nothing about my job or my CS Skills, but I can still copy-pasta.

    You've a fair point. If you know nothing, it can hardly be said that you've learned anything.

  • 0815 (unregistered) in reply to yolocaust
    yolocaust:
    ¯\(°_o)/¯ I DUNNO LOL:
    For bonus points, configure DHCP to return the same IP address for a laptop's wired and wireless interfaces.
    A very very bad idea what if they connect both wired and wireless at the same time? illogical but possible

    I have a wired dock, but have to run around a lot in the building - it's easy to forget to switch off the wireless when I work from my desk.

  • EatenByAGrue (unregistered)

    What I would have expected to happen is this: Alex goes to the developers and complains, and several senior developers start snickering.

    It's kind of like how on naval vessels a new recruit is asked to go searching for something that either doesn't exist or isn't what the young buck would think it was.

  • Kasper (unregistered) in reply to anonymous
    anonymous:
    yolocaust:
    ¯\(°_o)/¯ I DUNNO LOL:
    For bonus points, configure DHCP to return the same IP address for a laptop's wired and wireless interfaces.
    A very very bad idea what if they connect both wired and wireless at the same time? illogical but possible
    Then it would probably receive everything on both interfaces.
    Nope, IP does not work that way.

    Let's assume the laptop is running an OS which does support assigning the same IP to multiple interfaces (I know for a fact, that Linux supports that). And let's assume the DHCP server and client will accept such a setup (I have never tried that part).

    What will happen next is, that any computer needing to send packets to that IP will ask across the network, who has that IP address. Assuming wired and wireless is on the same segment, the laptop will request that ARP (or ND if you are little more up to date) packet on both interfaces. (If the two interfaces were not on the same segment to begin with, then assigning the same IP to them using DHCP would be an even worse idea).

    At this point the behavior of the kernel on the laptop makes all the difference. How exactly does it respond to those two copies of the ARP request? A likely result is, that when the request arrives on the wired network, the laptop responds over the wired network with the MAC address of the wired network interface, and when request arrives on the wireless interface, the laptop response over the wireless network with the MAC address of the wireless network interface.

    Most of the time, the reply with the wired MAC address will arrive first.

    If the laptop does answer with both addresses, it is up to the other computer, which of the two it will be using. It will have an IP packet queued (that's why it send the ARP request), and that IP packet will be sent as soon as the first ARP reply arrives. So that queued IP packet will be sent over the wired interface.

    Then another ARP reply arrives. On receipt of this ARP reply the ARP cache entry may be overwritten, or the old may be kept. That will probably depend on the implementation, and if the implementer did think about which choice was best, it was most likely based on an assumption that the two ARP replies originated from different computers, which may lead to a very different decision, on what is best.

    The end result will be that each IP packet is send to one interface or the other, but not to both (unless it is broadcast or multicast).

    But I say it is a bad idea to configure it this way because it is hard to predict which interface the packets will arrive on. And you definitely don't want all the packets arriving on the wireless interface, when wired is available. But that is a possible outcome.

  • Dan (unregistered) in reply to WL

    It works in Slackware

  • WillRay (unregistered) in reply to hobbes

    The programmers will be happy to respect networking and security's separate spheres of knowledge, the day that networking and especially security decide that it's reasonable to respect the programmer's expertise and authority in the programming and development domain.

Leave a comment on “It's My Network, I'll Configure it if I Want To”

Log In or post as a guest

Replying to comment #:

« Return to Article