there are times when i will make a quick fix to some application, upload it, and everything will work. but it's almost certain that the times when i do not make a temporary copy of the application, and instead replace the live version of the application without testing it will not be one of these times. at these times, more often, the "quick fix" i made turns out to be a "quick break" which turns into a "long fix" during which the application that was previously mostly working with one minor bug is not working at all.
not really. dave winer (and i'm not going to link to dave winer on principal) says "Mark Pilgrim reports that the optimizations that UserLand and other aggregator developers deployed in the last few weeks have saved him substantially on bandwidth costs." the thing is, mark hasn't written on his weblog in the last few weeks, so this bandwidth statistic may be skewed. anyway, that got me looking through mark's statistics pages and the page of incoming searches has a surprising number of searching for "hey everybody i'm looking at gay porn" or something very similar. following this search on google shows that mark does indeed come up on the first results page for a post he wrote on referrer spamming. it seems he fell into the same trap i did by discussing an odd text in his logs thereby causing it to turn up more in his logs. and now i'm doing my part to increase this trend by talking about it more.
i could go on and make analogies to quantum particles that change when you observe them, but instead i'll say that i want to know why mark pilgrim hasn't written for two weeks because i miss his writing.
one of my news feeds contains the following description of an article: "Now, a computer in every othe rroom isn't just for itinerant geeks (like me)." i can't link you to anything that will show you that sentence, because the actual article doesn't have the same typo, and the typo is the whole reason i think this is interesting. the typo, you may have noticed was "othe rroom", which should have been "other room". when i read this, i first noticed the missing "r" on "othe" and mentally corrected the mistake. then i noticed the extra "r" on "rroom" and mentally corrected that mistake (which, upon further reflection proves to be just another part of the same mistake). then when i got to "itinerant", i assumed the word was meant to be "internet" and had been mistyped, and i mentally edited it. when i got to the end of the sentence, i wasn't sure what it meant, so i re-read it and realized that "itinerant" is actually a word and not a typo of "internet". the interesting (i think) point here is that within the few seconds during which i read this one sentence, i adjusted my evaluation of the typers tendancy toward mistakes and accordingly adjusted my own tendancy to assume mistakes have been made. brains are amazing.
i've been working on a project for over a week now and i just tonight made it do something remotely useful. it can give you a good number (28) of grammatically correct sentences that contain only the words 'a','the','fish','cat','ate',and 'quickly'. it's not particularly useful, but it's a nice start. it's still quite slow, and it breaks easily, but maybe you can "smell what i'm cookin'".
when using php...you should not recursively call a function from within a foreach loop on a global array variable and expect the foreach loop to continue after the recursive function call returns a value. this is because foreach loops depend on a pointer within the array itself, and this pointer will be reset by the first nested instance of the function to complete the loop, so that when the call returns, the original loop sees the pointer as completed and stops looping. to avoid this, you should use foreach(array_values($array) as $items), (or something similar) which will iterate over a pointer within a local (to each instance of the funciton loop) copy of the array values.