There are no genealogical records on the internet. There are garbled extracts, inaccurate transcripts, more-or-less complete copies with (if you’re lucky) an accompanying image. But there are no actual records. For anyone doing research online, this is a very basic point, but one that can be difficult to grasp. The original records are all offline, “reality-based” as Donald Rumsfeld used to say, in archives and manuscript repositories and presbyteries and books.

The reason to keep this in mind is the seductive ease of access that the internet allows. This can mask great gulfs in apparently continuous runs of records, or transcription errors compounding record-keeping errors, not to mention large gaps in the search interfaces themselves.
This is most apparent to anyone searching the (otherwise wonderful) Latter-Day Saints’ site, familysearch.org: records are lumped together and described with the broadest of brushes, making it difficult to be sure what exactly you’re looking at. Have I really searched all Irish marriages up to 1898? (No) Are these three separate records of three different events or three different transcripts of the same record?

And the same problems crop up on every record site. Parish registers that are listed as transcribed and searchable are somehow missing from search results, while General Register Office records not listed somehow appear in search results (rootsireland.ie). “Mc Dermott” and “McDermott” are treated as completely unrelated surnames (irishgenealogy.ie). Record images exist online, but for some reason remain untranscribed (census.nationalarchives.ie).
The only solution goes back to the very first principle of all research: know absolutely, precisely, exactly what records you’ve searched, their dates, page numbers, locations, shelf references, gaps, original purposes … If you don’t, and you find nothing, you will certainly end up searching the same records all over again some time. If you do find something, and don’t have details of the record source, it becomes impossible to interpret. And you will certainly end up searching the same records all over again some time.
I know. I have been both of those researchers.
The rule is very simple: If you don’t know what you’ve searched, you don’t know what you’ve found.