• (disco)
    "I just wanted to report a on bugreport.apple.com," Simon E. writes
    I think TDWTF's QA forgot a <!--word--> there.
  • (disco)

    Am I missing something? I'm pretty sure the last one is Windows XP...

  • (disco) in reply to Joris

    I think that's the joke :stuck_out_tongue:

    …unless @mark_bowytz got the caption wrong, that is…

  • (disco)

    Some people, when confronted with a bug, think “I know, I'll report it to bugreport.apple.com.” Now they have two bugs.

  • (disco) in reply to RaceProUK

    Yeah, I think it needs just a little more sarcasm to make the joke clear. Something like

    I'm glad to know they'd never do something as stupid as run this on XP or Vista...

  • (disco) in reply to Joris

    I have seen and used some strange windows version called

     windows XP POS ready
    never really understood what it was. It did not have ALL the standard API but enough of it to be a real pain in the ***.

  • (disco)

    For the email one there are a bunch of sites that don't accept hyphens either. Or any special characters in the domain half for that matter... I actually have an alias version of my email domain for just those occasions. It is the same minus the hyphen.

  • (disco) in reply to Dlareg
    Dlareg:
    never really understood what it was.

    “Point Of Sale” is the marketdroid-approved expansion of “POS”. Everyone else has their own ideas.

  • (disco)

    "Wow! I can type the exact same speed I do now if I just get some training!" wrote Abner Q.

    Well what do you expect for free.

  • (disco) in reply to RaceProUK
    RaceProUK:
    …unless @mark_bowytz got the caption wrong, that is…

    Dammit.

    "but at least I know they're not using Vista" FTFY.

    JBert:
    I think TDWTF's QA forgot a there.

    "I just wanted to report a bug on bugreport.apple.com," FTFY!

  • (disco) in reply to mark_bowytz

    XD

    Descriptive is body; try to be a little more invalid

  • (disco)

    James C. wrote, "When invited to sign up for the Microsoft Partner Research Panel, I was presented with a question that I couldn't quite answer."

    Maybe the question was "what language does this nonsense resemble?"

  • (disco) in reply to powerlord

    @powerlord Maybe you're expected to inspect the source and fix it, thus qualifying you to join the panel.

  • (disco) in reply to powerlord
    powerlord:
    Maybe the question was "what language does this nonsense resemble?"

    I was amused to note that part of it looks a bit like SQL....as in SQL injection opportunity. Lots of sites seem to have SQL injection opportunities, but most of them don't advertise it by posting, "To join, bring down our website."

  • (disco) in reply to powerlord

    lojban?

  • (disco) in reply to shinji257
    shinji257:
    For the email one there are a bunch of sites that don't accept hyphens either. Or any special characters in the domain half for that matter... I actually have an alias version of my email domain for just those occasions. It is the same minus the hyphen.

    HP laptop installation/registration software as of a few years ago wouldn't accept + in an email address.

  • (disco)

    Regarding the u-verse one, long ago, I was a moderator on an MSN forum, and got free premium service forever (this is when they were a portal, and actually had such a thing.)

    They apparently didn't have any way to signal "no expiration date" so my actual expiration date in the system was in the year 2830.

  • (disco) in reply to FrostCat

    U-Verse isn't the only one. Currently I have a lifetime plan with VPN Unlimited and they signal that by giving you 100 years on your account. Their program does pick it up as a lifetime plan but their website shows the 100 year bit.

    Some other sites have done that and usually it is set to some crazy high expire date.

  • (disco) in reply to shinji257
    shinji257:
    Some other sites have done that and usually it is set to some crazy high expire date.

    Like the year 2830 (or so; I don't remember the exact year)? I got that account in 1996.

  • (disco) in reply to FrostCat

    The + sign bit is really common since it is rarely expected. Currently the only legit email addresses are ones where they part after is used for filtering purposes. I don't think I've seen one that wasn't used that way. However if you do accept email addresses like that you can't assume that to be the case at all since there is still that 0.0000001% chance that someone legit wanted a + sign in their email address for some odd reason.

  • (disco) in reply to shinji257
    shinji257:
    The + sign bit is really common since it is rarely expected.

    Nonetheless, it's actually legal. Presumably, someone could write a mail server that used it as an actual character, not a filter indicator, which is, of course, how I was actually using it.

    Once again, you should do next to no "validation" of email addresses--as has been amply argued about here over the years--because you're almost guaranteed to get it wrong.

  • (disco)
    FrostCat:
    They apparently didn't have any way to signal "no expiration date" so my actual expiration date in the system was in the year 2830.
    shinji257:
    lifetime plan with VPN Unlimited and they signal that by giving you 100 years on your account

    Seriously? Where do these developers come from? Just do something like:

    • Add a Lifetime or No_Expire flag to the DB. Was that so hard?
    • If you're concerned about DB space, make the Expiration_Date field nullable. Anything with a NULL in that field doesn't expire.

    Wow, that took a ton of work to figure out. I better take a nap now.

  • (disco) in reply to abarker
    abarker:
    Seriously? Where do these developers come from? Just do something like:

    In my case, Microsoft.

    Go figure.

  • (disco) in reply to FrostCat
    FrostCat:
    Once again, you should do next to no "validation" of email addresses--as has been amply argued about here over the years--because you're almost guaranteed to get it wrong.

    Something like:

    • Make sure it has an @ symbol somewhere in the middle. AIUI, that should be safe.
    • Send a verification email to the user. This is definitely a safe step if you need the email.

    Anything besides that is bound to be wrong in some case.

  • (disco) in reply to abarker
    abarker:
    Make sure it has an @ symbol somewhere in the middle. AIUI, that should be safe.

    Nope! (Although I don't know if anyone still uses bang-paths. And I vaguely remember that BITNET didn't use @ symbols but I might be wrong there.)

    The only universally-meaningful validation of an email address is to actually attempt to send an email to it.

  • (disco) in reply to FrostCat
    FrostCat:
    Nope! (Although I don't know if anyone still uses bang-paths. And I vaguely remember that BITNET didn't use @ symbols but I might be wrong there.)

    Huh. I'll have to check the official spec again. That's RFC2822, right?

    Or is it 2821?

  • (disco) in reply to abarker

    822, like God intended.

    Get off my lawn!

  • (disco) in reply to FrostCat
    FrostCat:
    822, like God intended.

    AHA! You are WRONG good sir!

    From the Abstract of RFC 2822 (emphasis mine):

    This standard supersedes the one specified in Request For Comments (RFC) 822, "Standard for the Format of ARPA Internet Text Messages", updating it to reflect current practice and incorporating incremental changes that were specified in other RFCs.

    In other words, RFC 822 is no longer the standard. :musical_note: Suck it. :fa_music:

    Now, regarding the current address standards, section 3.4 indicates that an addr-spec should be used for addressing. Section 3.4.1 defines an addr-spec as (emphasis mine):

    An addr-spec is a specific Internet identifier that contains a locally interpreted string followed by the at-sign character ("@", ASCII value 64) followed by an Internet domain.

  • (disco) in reply to abarker
    abarker:
    AHA! You are WRONG good sir!

    Did you miss my Get off my lawn! ?

  • (disco) in reply to abarker
    abarker:
    Make sure it has an @ symbol somewhere in the middle. AIUI, that should be safe.

    Practically, yes. Technically, no.

    FrostCat:
    Although I don't know if anyone still uses bang-paths.
    If they do, take them out and :gun:-bang them.

    There is technically the possibility of a local user for whom the username alone is a valid email address.

  • (disco) in reply to HardwareGeek
    HardwareGeek:
    Practically, yes. Technically, no.

    Someone else who needs to read RFC 2822. :P

  • (disco) in reply to HardwareGeek
    HardwareGeek:
    There is technically the possibility of a local user for whom the username alone is a valid email address.

    Generally not a concern when designing web software these days.

  • (disco) in reply to abarker
    abarker:
    Generally not a concern when designing web software these days.

    True. I started writing "If you're designing a form for local users, you should blah blah," but decided Belgium it, and deleted it.

  • (disco) in reply to HardwareGeek
    HardwareGeek:
    There is technically the possibility of a local user for whom the username alone is a valid email address.

    You've also gotta remember unusual stuff like parens, which probably most don't think of.

  • (disco) in reply to abarker
    abarker:
    Generally not a concern when designing web software these days.

    What if you work for the company developing the software?

  • (disco) in reply to FrostCat
    FrostCat:
    What if you work for the company developing the software?
    That's what I started to write about. At my most recent employer, AFAICR, everything either asked for your username (mostly), or if it asked for your email, explicitly asked you to include the @employer.com part.
  • (disco) in reply to abarker
    abarker:
    Make sure it has an @ symbol somewhere in the middle. AIUI, that should be safe.
    I should really get myself a separate e-mail-address just for testing X.400 support ...
  • (disco) in reply to aliceif

    Good luck finding a providor anymore. Support is fading for X.400.

  • (disco) in reply to abarker
    abarker:
    Support is fading for X.400.

    It's still alive? The only part of that entire stack that seems to have really survived robustly is X.509 and the bits that support it.

Leave a comment on “Tell QA They Missed One”

Log In or post as a guest

Replying to comment #:

« Return to Article