Inventors are continually told, "Keep an Inventors' Log Book" — which is then explained as a bound, page-numbered book in which the inventor is to date and document all new ideas and all actions he takes in the development of those ideas — a literal diary of his inventive activity — and in ink, without blank areas, without erasures (corrections are done with line outs). Periodically, new pages are to be read by a knowledgeable, disinterested acquaintance and each page dated and signed as "read and understood" by that person (the witness).
The purpose of this book is to evidence the dates and actions taken by the inventor in the creation of intellectual property if the validity of that property is ever challenged, e.g., was he really the "first to invent", did he "diligently" pursue development, etc., etc. The structure of the book (bound, page-numbered), its entry restrictions (dates, no blanks, no erasures) and the witnessing eliminate many possible recordkeeping challenges — provided that the witness is still available and willing to testify on your behalf.
Yes, you could have left blank areas and filled them in later — but is your witness going to swear — under oath — that there were no blank spaces on the pages he signed. Yes, you could have rewritten a whole new log book — but is your witness going to go along with that (i.e., re-sign and pre-date) and then swear under oath that he didn't. Of course, your witness will be challenged — does he have a monetary interest, a familial interest, long-time friend...?
The Log Book and its associated procedures are a good method for recording potential intellectual property — if done diligently and completely. Unfortunately, in my experience, it seldom is — even in the corporate world. The typical Log Book records a burst of activity — when the independent inventor is into recording, or right after Legal comes down on the corporate engineers (for poor recordkeeping) — and then nothing — and then another burst — which leave the Books wide open to "diligence" challenges.
A good loose-leaf recordkeeping system (e.g., Acco-bound Manila folder), maintained by "habit", will beat the Log Book, maintained sporadically, all day long. If you want to keep a Log Book and you're willing/able to do it "right" — do so.
But if you're not willing/able — over the long term, many years — get into the habit of maintaining a chrono file. Put everything into it — your original scratchings of a new idea, your original notes from telecons and meetings, correspondence in and out, receipts, etc., etc.
Yes, someone challenging your records can argue that you could have faked some of them. But that you "could have" and "did" are two different things — that come down to "believability". If your records show consistency and continuity — you've kept all that's relevant by "habit" — you have "believability" on your side — and you're likely to prevail.
And that applies not just to intellectual property matters — but to all your business dealings. In any legal dispute — other factors being anywhere close to equal — the party with the better records is most likely to prevail.
Since good recordkeeping is pitifully rare — even among big companies — this gives an enormous edge to those who make a "habit" of keeping good records.