• Prime Mover (unregistered)

    Anthony? Can you help?

  • Anthony (unregistered)

    Contact me. You know my hourly rate.

  • (nodebb)

    Without the context, nobody can say if it is wrong or right. There's a reason some languages have unreachable() pseudoinstruction (or something similar to it), and the presented code seems like C's implementation of that concept. I could bet some compiler can take this as a guide for optimization.

    Seeing assert(false,...) in code - the first thought that springs into my mind is a guard for non-exhaustive switch in a language that doesn't check/enforce it. A guard unreachable in the present code, but a safety net for further editions. One who never forgot break in switch let first throw his spacebar.

  • Anonymous') OR 1=1; DROP TABLE wtf; -- (unregistered)

    It's legacy code, so there's no meaningful commit history that far back

    If I had a nickel for every time a code base got migrated from one VCS to another without importing the full commit history, I'd have a lot of nickels. All too often the git blame terminates with some crazy commit that adds a thousand files and everything gits attributed to that lucky importer.

    Or maybe this company imported from Visual Source Safe, or worse.

  • (nodebb)

    Whose words these are I think I know They're ANTHONY's.

  • my name is missing (unregistered)

    This was clearly in code written for Cleopatra.

  • dpm (unregistered)

    If I were the programmer who wrote that assert(), I wouldn't use my real name. Which explains why no one remembers an employee by that name.

  • Just another Embedded Designer (unregistered)

    Ah the new inverse Spartacus

    I ain't taking the blame or owning it EVER

  • "Get" the error? "Get" it??? (unregistered)

    Clearly an advertising gimmick for Anthony's pizza. Actually, this makes me wonder if it would be a type of justice if such a thing got into web developers' code... (or Oculus these days, I hear).

  • DFYX (unregistered) in reply to Anonymous') OR 1=1; DROP TABLE wtf; --

    We imported a ton of stuff from VSS and actually managed to get a decent git history out of it. Takes a bit of fine tuning as VSS does its versioning per file instead of per repository but there are some good scripts out there.

  • (nodebb) in reply to Pawel Kraszewski

    Without the context, nobody can say if it is wrong or right.

    It's wrong. There is no context where leaving out the actual error is correct.

  • Edd (unregistered)

    If something beside remained that explained the error this could have been a useful suggestion for the user at the time, to be updated as the program and organization changes.

  • MJ (unregistered)

    A co-worker, long ago, put such a message into a Unix SCSI disk driver, back in the early days of SCSI...for a certain error condition that 'should never happen' he left a console message "If you see this message, <last name> screwed up." Sure enough, a year or so later, some CE from Germany emailed him with the question "Vas ist los '<last name>' screwed up' ? His last name was uncommon enough that they were able to locate him in the company-wide directory, even in those pre-Web days, and contact him directly.

  • Carl Witthoft (google)

    It's obviously Anthony Martignetti (RIP) , who ran home fast every Wednesday because it was Prince Spaghetti Day. If it hadn't been Wednesday, he'd have hung around long enough to explain the fault in the comment.

  • Appalled (unregistered)

    "Yeah, we're used to it now. The infamous Anthony error means a database record got mangled. We never found the source of the mangling. We just find the record from the screen shot and go fix it". Seen situations like this many times before over the last 40 years.

  • (nodebb)

    The new York central park charges upwards of $10,000 to etch your name into a bench for a proud of a few years. Anthony etched his name into history forever and he didn't pay rather got paid for it. He won

    Addendum 2021-06-28 12:17: *period

  • jeremy (unregistered)

    Have they checked the local bar? Anthony might offer the solution there

  • (nodebb)

    In the 1960's, the source listing for the CDC 3600 Algol compiler contained the following: Hardware Stack Underflow, Please send me the dump. The comment had a name and home address.

  • Donald (unregistered)

    This kind of makes sense. Most pieces of code will fail in the first week if they fail at all.

  • (nodebb)

    I like to think that the title is a reference to Sam Jackson's line from the underrated "Long Kiss Goodnight", though it's an easy joke for anyone to come up with.

  • Worf (unregistered)

    I've done it - if you have an enum and you've switch()ed on every possibility, then adding a default is perfectly reasonable. It actually happened to a coworker of mine - i wrote a state machine, enumerated the states and transitions, and in the transition code, added a default that should never be reached.

    Somehow though, my coworker managed to hit that unreachable code. Never figured out why since it's unreachable. I enumerated all the state transitions explicitly - not even making any assumptions on the numbering of the enums.

    What I did do was make the code do an early return when it hit the unreachable condition which crashed the system, making it an easy fault to find.

  • xtal256 (unregistered)

    "It's legacy code, so there's no meaningful commit history that far back"

    You know, I think there's a point where we have to stop blaming legacy code for things. I've seen articles here where "legacy" code is only a few years old, so I would expect it to use source control and have commit history. On the flip side, I work with code where some of it is 20-25 years old and it still has commit history.

  • Donald Klopper (google) in reply to xtal256

    Yeah this is not a WTF at all.

    The code outlived the risk. This rather points to poor management, a lack of handover and improper ownership. It's like when markets run rampant bullish or bearish and you now blame the origigal architects of an investment solution for poor returns, or celebrate them for the gains years down the line.

    Addendum 2021-06-29 03:17: Yeah this is not a WTF at all.

    The code outlived the risk. This rather points to poor management, a lack of handover and improper ownership. It's like when markets run rampant bullish or bearish and you now blame the original architects of an investment solution for poor returns, or celebrate them for the gains years down the line.

  • (nodebb)

    Although it's possible that there used to be a guy named Anthongy, I bet that today it's a role. Probably a rotating one, with each maintenance developer being The ANTHONY for one week.

  • kapue (unregistered)

    If the code works, it was written by me. If not, it was written by Anthony.

  • Trupik (unregistered)

    When false evaluates to true, you surely have problems that only Anthony can solve, whoever he is. Hopefully not a priest...

  • David Mårtensson (unregistered) in reply to Anonymous') OR 1=1; DROP TABLE wtf; --

    The project I spent most of my time working on started without version control (or devops or really any deploy strategy at all), once we got Visual source safe, that was a big improvement, then we moved to Team foundation and then to Git.

    My only solace was that any legacy code that was from an older system and lacked history was usually mine :P

    And since we moved from vbscript and jscript through visual basic and vb.net and then c# on net.framework and then netcore I am not sure really old history would be understandable or relevant.

    In total I worked just over 17 years on the system before moving on.

  • Paul M (unregistered)

    "Anthony" was the business analyst who swore blue that "this never happens".

Leave a comment on “Out of You and Umption”

Log In or post as a guest

Replying to comment #:

Log Out

« Return to Article