32-bit and 64-bit

[doaji]

I index over 100k pages, with my unique words going over 1million, V5 tells me i need a 64bit version, which i believe is only available in v6.
i have a 32-bit system.
can i run the 64-bit on my 32-bit system?
is the limitations increased for 32-bit (above 1 million) in v6?

[Ray]

i have a 32-bit system.
can i run the 64-bit on my 32-bit system?

No. You need a 64-bit system to run 64-bit software.

is the limitations increased for 32-bit (above 1 million) in v6?

You will still most likely need the 64-bit version if you are attempting to index over a million unique words. As noted by the message given to you by Zoom, this depends on the actual content indexed.

But ultimately, you are not able to use any more than 3GB of memory on a 32-bit system. That's not a limit in Zoom, that's a limit in Windows. And when you're indexing that many files, it is very likely you will need more than 3GB of memory in your system.

Do you know for sure that you have over 1 million unique words? Or are you just making a wild guess? Note that a unique word is not the same as the total number of words. The English dictionary only has 50,000 unique words. A re-occurring word does not count as an unique word.

If Zoom is apparently indexing many words, it's worth checking why. For example, you may have configured Zoom to index binary files as text files. Then it will index alot of binary junk as words.

What file types are you indexing? How big are each of these files? Do you have many massive PDF files containing product codes, or phone numbers? (which count as words)

[doaji]

i have ran zoom, and it maxed out on the unique words, and yea i do have pdf files that contains phone numbers, and numbers. is there anyway you think i can configure zoom to overlook this kind of "words"?

[Ray]

You can either:
(a) Skip the PDF files in question by adding them to the skip list
(b) Limit the number of words indexed from each file ("Configure"->"Limits"->"Limit words per file"). This means you only index, say, the first 1000 words from each file. This will prevent a PDF file with a huge number of numbers from swamping the dictionary as only the first portion of it will be indexed.