| |
Polls |
| |
(14:26 27/9/2001)
johnstlr (08:41 28/9/2001)
rich (08:58 28/9/2001)
Phlamethrower (14:51 28/9/2001)
geoff_youngs (17:00 28/9/2001)
Phlamethrower (16:47 29/9/2001)
rich (09:25 1/10/2001)
|
| |
|
Phlamethrower |
Message #5505, posted at 14:26, 27/9/2001 |
| Unregistered user
|
Here's a quick one: As a gauge of how popular TIB is, how about putting the number of votes received up with the poll? |
| |
 [ Log in to reply ] |
| |
|
johnstlr |
Message #5506, posted at 08:41, 28/9/2001, in reply to message #5505 |
| Unregistered user
|
Surely this would only tell you how popular a particular poll is. I don't always vote because I don't always have an answer. Take the current poll. I would probably buy both Tek and ID if they appeared but I've pretty confident ID won't appear on RISC OS. Ok so I can vote for Tek only but this does kind of miss the point a bit. |
| |
[ Log in to reply ] |
| |
|
rich |
Message #5507, posted at 08:58, 28/9/2001, in reply to message #5506 |
| Unregistered user
|
Part of the point of the current poll is to gauge if bringing out ID would be a good idea or not (albeit in an unofficial, not-related-to-Artex way). So assume they both will be brought out and vote accordingly! |
| |
[ Log in to reply ] |
| |
|
Phlamethrower |
Message #5508, posted at 14:51, 28/9/2001, in reply to message #5507 |
| Unregistered user
|
Actually the reason I'm asking for this is that I'm interested in how many people are promising to buy TEK/ID, rather than just the percentages. I'm sure Artex would be interested in the numbers as well. |
| |
[ Log in to reply ] |
| |
|
geoff_youngs |
Message #5509, posted at 17:00, 28/9/2001, in reply to message #5508 |
| Unregistered user
|
Actually the reason I'm asking for this is that I'm interested in how many people are promising to buy TEK/ID, rather than just the percentages. I'm sure Artex would be interested in the numbers as well. At a extremely crude lower edge estimate, I would suggest a minimum of 99 (at latest count). 10 : a%=0
20 : REPEAT
30 : a%+=1
40 : UNTIL FNcheck(10) AND FNcheck(20) AND FNcheck(11) AND FNcheck(66)
50 : PRINTa%
60 : END
70 : DEF FNcheck(v%)
80 : LOCALb%,c%
90 : b%=v%*a%/100
100 : c%=b%/a%*100
110 : =(c%=b%)
Which is probably a highly innacurate and flawed approach, making far too many assumptions about the similarities of rounding routines in BBC BASIC runnning on a RiscPC and PHP on Linux, has the disadvantage that it can only try to estimate the minimum number of voters, doesn't work if more than 100 people have voted, etc, etc, etc. |
| |
[ Log in to reply ] |
| |
|
Phlamethrower |
Message #5510, posted at 16:47, 29/9/2001, in reply to message #5509 |
| Unregistered user
|
Um, are all those FNcheck's meant to add up to 100? 10+20+11+66=107... Anyway, I doubt that method would catch on. |
| |
[ Log in to reply ] |
| |
|
rich |
Message #5511, posted at 09:25, 1/10/2001, in reply to message #5508 |
| Unregistered user
|
Actually the reason I'm asking for this is that I'm interested in how many people are promising to buy TEK/ID, rather than just the percentages. I'm sure Artex would be interested in the numbers as well. Just to state that voting isn't really a promise, before someone starts wondering if we're logging voter details and giving them to Artex to chase up orders! But we will give Artex access to raw numbers if they want them, because hey, we're nice like that.The problem is that the poll system was designed to be just a gauge of public opinion, and I've never been keen on putting the raw numbers out there. I fear that it would be more about the numbers, not about the result, as I learned when working at ArgoNet. <snip long, rambling story about a problem with people demanding access to ArgoNet logs they weren't entitled to> So basically it'd be a fairly simple job just to print the raw numbers next to the poll graph, but I've never been keen. I'm open to persuasion though. |
| |
[ Log in to reply ] |
| |
|
