Detail warning.
@Michel re: Draft Help Page Help:Reliable Sources (wikitree.com)
There is one critical improvement to this draft page which should be considered: at no point in the draft are the key terms "source" or "reliable" defined.
The word "Source" is linked to Help:Sources (wikitree.com), where one finds: "A source is the identification of where you obtained information." This G2G discussion indicates a much more nuanced approach and a fundmentally different definition. Rather than "the identification of where you obtained" something, most posts here suggest that the "source" is actually the person originally supplying the original data (regardless of where a genealogist may have found it later). There is a second usage which is almost as prevalent, in which the term "source" refers to the first (written or other) record made based on the data provided by the (source) person. Then there are numerous other "sources" in which that initial record was preserved, copied, collected, compiled, transcribed, translated and/or interpreted by one or more individuals to create some meaningful information or narrative. The definition of "source" found on the (linked) page seems to focus entirely on what happens next: some genealogist "finds" information and "identifies" the location, so that another genealogist might also find it. That definition would probably fit the concept of "citation" better than "source".
As for "reliable": A common understanding of this essential term is simply taken for granted in the draft Help: page. We learn that "reliable sources" are required and ask "how do you know if a source is reliable?" There are guidelines for "judging reliability" and even a sweeping statement that "primary sources are the most reliable" because they are recorded "at or near the time of the event by or with someone who has first-hand knowledge". But at no point is the most basic question even adressed: What do we mean by "reliable"?
One "dictionary definition" of reliable is: "consistently good in quality or performance; able to be trusted" That definition works well enough for our purposes - if others agree, there should AT LEAST be some mention of this in the draft page.
Once "reliable" has been defined, it can be very usefully applied in a discussion of which people, records, and other things referred to as "sources" are (recommended, approved, accepted - take your pick) by the WikiTree leadership to support the entry of information on pre-1700 profiles.
a) we can't rule out that the PERSON who originally supply the data made a mistake - either intentionally or unintentionally. But parents and relatives reporting birth and death dates, as well as city clerks, priests and immigration officials may generally be considered "consistently good in quality or performance; able to be trusted".
b) we similarly can't rule out that some time-honored genealogical works included errors of commission or omission - for whatever political and/or pecuniary reasons one cares to imagine. But certain works - especially those which show evidence of "scholarship", can also be considered "consistently good in quality or performance; able to be trusted".
c) At the end of the day, putting ANY resource, repository or volume on a list of "reliable sources" is a decision (in our case, by a dedicated group of experienced volunteers who have accepted responsibility to help the WikiTree community by becoming "project leaders") to recommend, approve or accept the items on that list as "consistently good in quality or performance; able to be trusted".
IMHO, a carefully constructed, well-defined statement of what WikiTree means by "source" and by "reliable" would go a long way toward ending the endless discussion about "reliable sources". We could still use the term as a shorthand, but it would be clear that:
1) the lists include "sources" which, in the opinion of the relevant project leader, are recommended/approved/accepted as "consistently good in quality or performance; able to be trusted" - and therefore "reliable". MUCH more important is the converse: anything NOT on the list has simply proven to fail this test! What we are REALLY trying to avoid is for new, enthusiastic WikiTreers to more-or-less "blindly" copy data which can only be found in places which are constently poor in quality and performance and therefore not able to be trusted.
2) Each bit of data ("My name is "Friedrich Müller"), each transcription (port official writes "Miller"), transcription (handwriting is unclear, so typed list reads "Mills"), and interpretation (biographer claim Fred Mills arrived, based on typed list) is only as reliable as the individuals involved. Each genealogist must decide whether or not it is "able to be trusted" - especially if conflicting indications exist. None of that diminishes in ANY way the invaluable help which WikiTree project leaders provide by creating and maintaining lists of "reliable sources" - or the important quality-assurance factor involved in ensuring that anyone working on pre-1700 profiles is aware of and willing to abide by the corresponding safeguards.