@Kaeru: the original poster left off a few important words. The quote from the help file is from the entry for Base64Decoder().Kaeru Gaman wrote:think!Melissa wrote:No, read harder.Seymour Clufley wrote:Surely 25% larger?Melissa wrote:Fail, it should be 25% smaller.It's recommended to get a slightly larger buffer, like 30% smaller to avoid overflows.
... how can you avoid an overflow by using a smaller buffer?
... if only a smaller buffer was needed, how can using a bigger buffer produce an overflow?
surely, it should read "like 30% larger"
If one is using Base64Encoder() the buffer is larger, but if one is using Base64Decoder() the buffer is smaller It would appear that the entry in the manual for Base64Decoder() duplicated portions of the entry for Base64Encoder() but failed to adjust (i.e. replace 'larger' with 'smaller') it to match.
Since the original post four days ago, most of that has been corrected in the help file (though it still says 'overflows' which doesn't make since when it recommends a smaller buffer). The poster still thinks that the precision of the percentage difference of the buffers needs to be adjusted from 30-33% to 25%.
I am in agreement that the language should be adjusted. I'm not concerned about the percentage.