Number formatting fail



  •  

    I saw this and thought of you guys.

     

    The url the link was pointing to also had the scientific notation, so it didn't actually work.



  • If it had worked, that would have been TRWTF!



  • @snoofle said:

    If it had worked, that would have been TRWTF!

    If it had worked (it just depends on whether the server side language understands the number syntax - since it produced it, it probably will), it would probably have loaded 48588670779700, which stands a 1% chance of being correct. The useful point is that it's fairly simple to try all numbers from 48588670779700 to 48588670779799, and see which one you get access to, in order to work out the correct number. Just be thankful it wasn't E+999 ...



  • @TarquinWJ said:

    If it had worked (it just depends on whether the server side language understands the number syntax - since it produced it, it probably will), it would probably have loaded 48588670779700, which stands a 1% chance of being correct. The useful point is that it's fairly simple to try all numbers from 48588670779700 to 48588670779799, and see which one you get access to, in order to work out the correct number. Just be thankful it wasn't E+999 ...

     

    Don't forget rounding: it might be from 48588670779650 upward



  • I'm not sure about that: Remember, integer to double conversion is a strange beast. You had an integer, then converted into floating point, which was then rendered in decimal-with-exponent form.

    I'd probably start around ...9620 odd, and not give up until I have reached 9700.

     (I often get job sheets with important numbers that have got this treatment: usually with only 5 or 6 significant figures, which are all boilerplate anyway. Thankfully, it is rarely a trouble to get the correct ones from the helpdesk.)



  •  It's because all the morons out there insist on taking data that came originally in some perfectly reasonable transfer format, like comma-separated values (CSV), and load it up in M$ Excel and save it back out from there... they might as well print it out, feed it to a goat, and then submit the goat's crap.



  • @dtobias said:

    It's because all the morons out there insist on taking data that came originally in some perfectly reasonable transfer format, like comma-separated values (CSV), and load it up in M$ Excel and save it back out from there... they might as well print it out, feed it to a goat, and then submit the goat's crap.

    I've bolded the stupid parts of your post for you.



  • @morbiuswilters said:

    @dtobias said:

    It's because all the morons out there insist on taking data that came originally in some perfectly reasonable transfer format, like comma-separated values (CSV), and load it up in M$ Excel and save it back out from there... they might as well print it out, feed it to a goat, and then submit the goat's crap.

    I've bolded the stupid parts of your post for you.

     

    CSV isn't a reasonable file transfer format?



  • @dtech said:

    CSV isn't a reasonable file transfer format?
     

    No, it's a reasonable database format



  • I've been surprised recently how good CSV can be for a database. 50,000 names in an hierarchical database for an online organisation chart gives virtually instantaneous response times when the $10,000 piece of software that HR bought takes 30 seconds to initialise its crappy view. (Not to mention the consulting fees, which must have exceeded the cost of the software in order to choose a set of colours that makes it look even crappier and compromises the usability.)


  • :belt_onion:

     @Zemm said:

    @dtech said:

    CSV isn't a reasonable file transfer format?
     

    No, it's a reasonable database format

    For file transfers use XML. You can save meta information like datatypes and localization! Wouldn't that make for an excellent database format as well?


  • @bjolling said:

     @Zemm said:

    @dtech said:

    CSV isn't a reasonable file transfer format?
     

    No, it's a reasonable database format

    For file transfers use XML. You can save meta information like datatypes and localization! Wouldn't that make for an excellent database format as well?
     

    For L10N you just make a copy of the CSV database. Everything is a string, why would you need anything else?



  •  @bjolling said:

     @Zemm said:

    @dtech said:

    CSV isn't a reasonable file transfer format?
     

    No, it's a reasonable database format

    For file transfers use XML. You can save meta information like datatypes and localization! Wouldn't that make for an excellent database format as well?

    My meter is broken today.  That is sarcasm, yes?


  • :belt_onion:

    @sylvic said:

     @bjolling said:

     @Zemm said:

    @dtech said:

    CSV isn't a reasonable file transfer format?
     

    No, it's a reasonable database format

    For file transfers use XML. You can save meta information like datatypes and localization! Wouldn't that make for an excellent database format as well?

    My meter is broken today.  That is sarcasm, yes?

    I don't know! I wasn't expecting the Spanish Inquisition.


  • @bjolling said:

    I don't know! I wasn't expecting the Spanish Inquisition.

    Nobody expects the Spa...

    @bjolling said:

    Our chief weapon is suprise. Fear and surprise!

    ...oh, bugger!


Log in to reply