On Thu, 28 Feb 2002 16:53:59 -0000, DL Neil wrote: >BTW: this apparent accreditation (below) is incorrect/an accident of editing? I >didn't say this. It was Doug (he >says pointing an accusative finger). Doug whilst I'm (appearing to) round on you, >contrary to your comment about >lack of evidence, Benjamin Pflugman and I responded to Richard's first msg and BOTH >reproduced the original >phenomenon. > >It is true that I tested on a Windows box, but I can't say what Benjamin's is, to >know if it has anything to do >with M$ particularly - but don't let the mere facts get in the way of a salacious >story (M$ don't!) >
I missed Benjamin's and your replies. I apologize for suggesting there was apathy being exercised. ;) I spent the last couple of hours replacing the mainboard in another box I have here and doing some testing. It now has the same board/cpu as this Win98SE machine. It is running Linux-Mandrake 8.0 with MySQL 3.23.38 which came with the LM distro. I then created a table and populated it with fifteen, or so, rows of data. The SELECT statement that started this thread : SELECT number,round(number,digits) FROM testme was then tested as well as a few variations. I also retested on this Win98 machine. +--------------------------------+--------------------------------+ | Win98SE / AMD cpu | Mandrake 8.0 / AMD cpu | +--------------------------------+--------------------------------+ | MySQL 3.23.43 | MySQL 3.23.38 | +--------------------------------+--------------------------------+ | ROUND(cola,colb) works | ROUND(cola,colb) works | | Rounding is DOWN from 0.500 | Rounding is UP from 0.500 | | F.P. format is inherited by | F.P. format is inherited by | | calculated column | calculated column | +--------------------------------+--------------------------------+ I don't believe I made it clear that ROUND(cola,colb) always worked for me on the Windows machine, too. >Thank you to Roger for a concise summary of the two issues: the somewhat familiar >C-rounding >library-inconsistency problem had (to my mind) started to overshadow the original >issue with ROUND(). Perhaps. It was very well stated. Though, my results above would lead one to conclude that there is only a problem with C- rounding. I'm not suggesting that should be a final assessment; merely commenting on the one set of data and the possible difficulty of nailing down the cause. > >I agree, feeling that it is either a bug in MySQL or a 'bug' in the documentaitno - >as I said at the time and >was restated/reworded today. All of the manual examples give a constant as the number >of significant figures for >this and similar functions, but the command templates seem to say that a column value >or variable could be used. > >I had forgotten that Benjamin had announced his firm grasp on the initiative - I >guess my less teutonic (but >nevertheless tidy, if small) mind felt that I should ensure Richard scooped that >particular pool - some people >think such things deserve some 'credit', because it is certainly not everyday anyone >finds a genuine bug in >MySQL! > >More of interest, I'd like to hear from one of the MySQL AB people, to find out if >there is a misunderstanding >of how things should really be, or if it is in fact a bug? Moved and seconded. Doug (Bot bottleneck botulin: query) --------------------------------------------------------------------- Before posting, please check: http://www.mysql.com/manual.php (the manual) http://lists.mysql.com/ (the list archive) To request this thread, e-mail <[EMAIL PROTECTED]> To unsubscribe, e-mail <[EMAIL PROTECTED]> Trouble unsubscribing? Try: http://lists.mysql.com/php/unsubscribe.php