Charlie
2003-04-11 04:35:58 |
Long Comments
Is there a limitation on the size of comments posted toward problems? I know that I have had to break apart a comment and continue it on another comment if it exceeded about 25 lines, and I've seen others do that also. Yet today I see a very long comment (Software Solution) posted against the puzzle Grid Pathways. Is it a function of the browser? I'm using IE 6.0. There's no difficulty posting long solutions in the submitted puzzle area--the limitation applies only to comments under posted puzzles. |
levik
2003-04-11 05:34:34 |
Re: Long Comments
I'm honestly not sure why this would be. There's nothing in the software that makes it behave that way. One hting I can think of is thta perhaps this limit is only effective for preview (i.e. previews of longer comments don't work, but posting would be OK). Can somebody email me with a copy of a comment that they were unable to post to the site? I will then look into it. |
DJ
2003-04-11 05:40:06 |
Re: Long Comments
I have experienced the same problem with a few lengthy problems (for example, I have been unable to post my full solution for the yet-unsolved 'Locked Safe' problem). I'm actually on a computer at school right now, but if it will be helpful, I'll try to post here the entire comment (it's saved in a text file on my home computer). There have been a few other comments that didn't work, I'm assuming because of length, but I cannot remember any others offhand.
=) |
Charlie
2003-04-11 08:29:08 |
Re: Long Comments
The next time it happens, I'll have to check. Perhaps it had been a confluence of two things: inability of the preview to handle it, and taking so long to compose the long post that my login timed out. Or perhaps composing it offline in Word and copying and pasting introduced bad codes, etc. Next time I see a need for a long comment, I'll compose in Notepad so as not to get wrong codes, and make sure I'm logged on, and ignore an inability to preview. I'll see what happens then. |
DJ
2003-04-11 09:46:42 |
Re: Long Comments
I am having the same problem, Charlie, having done everything you just suggested. I always write only in notepad, I'm already logged in, and I skipped the preview - to no avail. One such comment (and I have modified it several times but it is still too long, if that is the problem) follows (after *** to the end of this message):
***
I had 'proved' an 8-digit solution before, but my oversight was that the reverse of x is not necessarily x. So, I will redo the problem, with the invented notation of X being the reverse of x (if x=123, X=321). As before, I will attempt to prove that a number relates to itself, which by property E means that it must be a solution and open the lock. The shortest solution still presumably has the form of xx2 or x2x2, where x is a series of 1s, 5s, and 9s that enumerates the steps needed to find that very arrangement.
First, for xx2, lets work backwards:
xx2 - 2XX - XX - xx - x
or
xx2 - 2XX - XX - X - x.
A consecutive reversal (5) and repetition (9) are interchangeable. This yields the following proof:
2x2 ~ x
92x2 ~ xx
592x2 ~ XX
1592x2 ~ 2XX
51592x2 ~ xx2
:. x must equal 51592, and this shows that 51592515922 relates to itself and therefore is a combination that opens the safe. Also (using the 5-9 commutability), 51952519522 opens the safe by similar proof.
Now, work backwards again to find x2x2:
x2x2 - 2X2X - 2X - X - x
or
x2x2 - x2 - 2X - X - x.
Thus we have the proofs:
2x2 ~ x
52x2 ~ X
152x2 ~ 2X
9152x2 ~ 2X2X
59152x2 ~ x2x2
:. x = 5915, and 5915259152 relates to itself and must open the safe. Also:
2x2 ~ x
52x2 ~ X
152x2 ~ 2X
5152x2 ~ x2
95152x2 ~ x2x2
:. x=9515, and 9515295152 also relates to itself and must open the safe.
Such solutions were come upon previously by other people. The next step is what eluded me for some time; that finding a solution one way may make it possible to find others another way.
An inspection of the latter two valid combinations reveals that both are repeats of some number. A simple proof, from the definitions, shows that:
2y2 ~ y
92y2 ~ yy
So, for any repeated number yy, 92y2 is related to it. And, if yy opens the safe (see where I'm going now?), then 92y2 must also open the safe, to conform to property E. Therefore, the solutions 9515295152 and 5915259152 each yield another solution, and we have that:
2951522 ~ 95152
92951522 ~ 9515295152
and
2591522 ~ 59152
92591522 ~ 5915259152.
Therefore, we have found two 8-digit solutions, 92591522 and 92951522.
!! |
DJ
2003-04-11 09:47:33 |
Re: Long Comments
So, that's pretty lengthy, and I can think of no reason why that should prevent it from being submitted; nor can I think of any other explanation. Thoughts? |
TomM
2003-04-11 12:00:38 |
Re: Long Comments
Whatever the cause, the long comment problem has been with us a long time. Most of us have learned to adapt and break up our long responses into two or more comments. Not an ideal solution, but not terribly onerous either. |
levik
2003-04-14 05:39:01 |
DOH! (long comments)
DOH!!!!
Just realized the cause of the long comments issues! The form submitting comments was set to be a method "GET" form (those who know a little bit about how web scripts work will now let out a groan and think: "Levik is an idiot").
For everyone else's sake, I will say that "Get" forms are posted in the URL bar of your browser - this means that the long comment you just typed in is being crammed in to that tiny field JUST big enough to fit "http://www.flooble.com/perplexus/". Naturally, most browsers cut this field off at SOME length, resulting in the problems people have been having.
I just changed it to using the POST method, so hopefully this will resole everything. |
Charlie
2003-04-15 05:55:33 |
Re: Long Comments
Works great! The squeaking shoe gets the new sole.
I think that the following is something new: posting the comment after time out results in a blank comment being posted to the puzzle, rather than merely forgetting about the post altogether. But that's better than not being able to post long comments. |