A successful search often takes two or three tries, sometimes more. Especially so when checking to see a topic has already been posted. Should I just go ahead and post?
That’s it.
Peace,
mangeorge
I would. Waiting five minutes to try again, while it saves the server, isn’t worth the effort. If it’s a duplicate posting, someone will report it and it will be merged or redirected. 
I agree with mangeorge. Five minutes seems excessive.
What’s the purpose of the delay anyway? I understand the 60 second rule between posts (How come only one post every 60 seconds?), but those reasons don’t apply to searches.
Searches, depending on the implementation of SQL logic and other factors, can be VERY system intensive, probably more than adding a single post. It was the thought of the mods at one time that too many simultaneous searches was the cause of not only slowdowns but crashes. I ha’ me doots, but nobody’s proved 'em wrong yet.
I can see why a search would be extremely resource intensive if it had to plow through the content of every single post, but my understanding is that all words over tree letters are indexed, so I wouldn’t expect that to be so bad. Of course, this is outside my expertise, so it may well be the case.
Anyway five minutes??!! I’m afraid that will make people want to blow off the search entirely and just post duplicate threads.
I don’t know about crashes, but I think it’s clear that it didn’t do anything for the slowdowns.
its fine by me.
Irritating as hell - particularly when people snark me for a repost - but I get it.
I don’t know much about the internals of the SQL version in use, but I do know quite a bit about internals in database management programs in general and others specifically.
There is a considerable difference in how search procedures are executed and a lot of the resulting speed or lack of it is due to the high-level coding. B-trees are marvelous things, but poor procedure calls can negate much of their efficiency.
To put it simply, a well-structured query will access only a handful of data records to find (or not find) the desired record(s). A poorly-structured one may have to search ALL records or a very large sub-group. It is difficult for the end user to know how his query is handled when he hits “search”. To get to the bottom of this, someone would have to have extensive knowledge of SQL, PHP, and vB to see how they interact and how they might be improved. I hope someone like that pops up sometime in my lifetime.