User:Wantnot

= My Editing Philosophy =

I try very hard to have complete accuracy in my edits. Any mistake I make is potentially amplified by many thousands of readers getting the wrong information. I strive for accuracy.

Before submitting, I will always preview my changes, no matter how small, and check every link I added or changed.

(But I'm sure I mess up sometimes. Sorry for when I do.)

My Music Editing Philosophy
[I can't believe I made a "My Editing Philosophy" section on my user page without a subsection for music - and I kept it that way for years...]

I actually do have interests, and even strengths, outside of music and actually exercise them on Wikipedia sometimes. But my biggest strength, and the place where I have the most to contribute, is music.

I will almost never edit album or track details, e.g., album or track name or composer or producer credits, etc., unless I have visual confirmation of an error or missing details.
 * "Errors or missing details" means a difference in a Wikipedia article and actual artwork (i.e., visual confirmation) of the release.
 * "Visual confirmation" means actually seeing the details in the album artwork. This is typically from LP back covers and/or gatefold details, LP or CD label details, or CD booklet details.
 * Sites with artwork which are usually, but not exclusively, useful are: discogs.com, 45cat.com, 45worlds.com (the 78 and 45 section is excellent and the LP details are getting pretty good now, too), rootsvinylguide.com (can be excellent) and popsike.com (which is hit and miss but has proven to be invaluable). rateyourmusic.com is okay but for the front cover only.  Of course there are countless less-well-known sites, often fan sites, with visual details.  Some are simply amazing.
 * Of the resources specifically listed above, discogs.com is, at the time of writing, singled out in Reliable sources/Perennial sources as "user-generated, and is therefore generally unreliable" though all of these resources would surely meet the same criteria that made discogs "generally unreliable".
 * I absolutely agree that the user-entered text on the site can't be trusted. Sometimes the details are painstakingly transcribed from the album artwork or similarly culled from reliable resources (with comments detailing this).  In other instances, unfortunately, it's much less trustworthy.
 * But I steadfastly disagree that discogs.com, and the others as discussed above, should be considered "generally unreliable", per se, and, essentially, not to be used as references. Visual confirmation from the artwork is almost always significant, if not definitive, for artwork from the country where the album was first released.  Of course there are typos and errors, most often in the composer, frequently in track durations, and sometimes even in the track name.  Composer differences are often due to litigation and should be discussed in the article text. Differences in durations I don't touch as they're so hard to conclusively confirm even if you think you have original media.  But if the track title in the article is different than that seen on the relevant album artwork, without explanation to the contrary, I believe the text used in the album artwork should be used.
 * Unfortunately a lot of hard work, maybe from someone as assiduous as myself, might have to be discarded due to uncertainty of the diligence of the author.
 * And what do I do when the relevant album artwork has contradictory details such as different track titles on the LP back cover and media label? After a bout of crying, I rarely edit the article in this case unless it seems overwhelmingly obvious which is correct.
 * And what do I do when artwork details differ between an original vinyl LP release and a CD reissue differ?
 * Research! And if that doesn't result in a definitive answer, bail!  Updating questionable data with poorly-verified information is not a win.
 * And when worse comes to worst, there are sites like eBay. When an article lacks details that require using such a site, as said sources could disappear moments after I access the information, I will absolutely include the precise text used along with the precise location from which it was retrieved, i.e., an LP back cover or and LP media label.  I will also include the catalog number of the release from which the details were obtained.

I munge all this up and, if I believe I have sufficient details from this information, I will update the article. I will add citations where reliable. Knowing they may disappear at the blink of an eye, such as for eBay listings, I try to, and hope I have, include links to the visual sources and include catalog numbers and the exact text I am using for whatever I'm editing (e.g., track title, album title, ...) where relevant and possible, so that I, or future Wikipedians, have some hope of finding alternate sources if necessary.

The more ephemeral the source, e.g. eBay, the more I try to leave breadcrumbs for future use.
 * That said, I'm sure I was less strict in earlier days. Sorry...  I'm getting better, I swear.

I avoid using text-only sources unless there is no alternative - and only if the text-only sources are reliable
As so tempting as it can be when visual sources can't be found, I do not use text-only sources such as allmusic.com (a very frequent source for Wikipedia as well as a frequent source of errors on Wikipedia (sorry allmusic, that's just the way it is)), last.fm, or freedb.org. Similarly, I rarely accept Wikipedia itself as a source unless the sourcing is sufficient.

I don't want to besmirch the hard-working Wikipedians that kick butt to the smallest detail, and there are many like that. But, again, I don't want to update Wikipedia unless I'm sure of my change. So, "It's not you, it's me" :).

That said, I do sometimes make updates based on text-only details but if and only if they are significantly more authoritative than what is currently on the site. An example can be seen in the Correct Composer Credits section of the The Spencer Davis Group Gimme Some Lovin' talk page (for which I actually didn't update the main article other than a citation request).

...Especially allmusic.com
Again, allmusic.com, a very common citation on Wikipedia, simply can't be considered a reliable source across the board, in my opinion, "just cuz", which is unfortunately frequent, noting that allmusic does not appear in External links/Perennial websites at this time. While there are surely allmusic pages that are bang-on accurate in every way imaginable, there is no justification for this to be assumed in general. However, allmusic has been used to show that my visual resources might be wrong simply because allmusic has differing, unreferenced, text-only, data.

IMO, it's time to treat allmusic like IMDb, an unreliable source in general as opposed to "assumed accurate until proven otherwise", a position which it appears to hold at this time.

So how do I determine the most appropriate release country?

 * This can be tricky, too (but it's usually not too bad). The Beatles are unequivocally a British band.  Similarly, Steely Dan is unequivocally a US band.  But bands like U2 or Thin Lizzy are Irish though most releases after they got popular will be from labels in England.  Somewhat similarly, for bands like Rush, I will try to source the Canadian release for details like the record label (Moon Records for the very early years, then Anthem, and then Mercury), but I'll accept a US release for details like composer and producer if and only if a Canadian visual source is not available.
 * But, of course, it's not that simple. Donovan, for example, despite being from the UK, could not release such monumental records as Sunshine Superman in the UK until much after the US release due to contractual disputes.  So, in this case, the US release is more appropriate than the UK release.
 * That said, releases from outside the home country can sometimes be helpful. The Netherlands releases are often remarkably reliable for details other than the record label for UK releases.
 * But it's still better to be safe than sorry most of the time.
 * If a release from an appropriate source country is not available, I will usually give up.

Whatever the appropriate release country is, is the earliest release the most authoritative?
Not necessarily.

Sometimes the initial releases have simple typos or technical errors (typically typos in track titles or composer or producer errors) or credits are amended after initial release due to legal action. I check the details on the initial CD release from the appropriate countries as well as any significant later releases such as deluxe editions, etc.


 * Actually, I check every single release that could conceivably be considered a reliable source. I check artwork from LPs, cassettes, CDs, etc. and reconcile any differences before submitting (or I don't submit).


 * NOTE: Cassette artwork should not be ignored! For some releases, cassette artwork might actually be the best visual source for details.  There have been quite a few times in my Wikipedia or personal research where there may have been numerous scans of the front and back cover and media, etc., from many countries and media types.  But, for both LPs and CDs, some details, such as composer or producer, only appeared on the LP gatefold or CD booklet for which no LP or CD scans could be found.  But a relevant cassette scan sometimes saved the day!
 * That said, cassettes (and 8-tracks) can not be relied on, in general, for track length, number, or ordering, as cassettes and 8-tracks may have a different track ordering and the tracks may be longer or shorter than on the LP. Also, tracks are often omitted or added on tape releases, or the tracks may differ.


 * An example of an extra track being added to a cassette release is "Something Cool" initially only appearing on the cassette release of Rickie Lee Jones' Girl at Her Volcano. An example of a track differing on tape releases is the two "Pigs on the wing" parts being combined on the 8-track release of Animals with the two parts linked by a guitar solo not included on the LP (which is not performed by David Gilmour, actually).

So what do I do when the above details conflict?
Tough question. And there's not a single, simple answer. First, I try to have the "optimal" source for information. This usually means:


 * the appropriate release format, i.e., LP or CD (or 45 or 78 or...), for the time frame
 * more importantly, the home country for the artist, or best appropriate substitute (more above)

Trying to maximize the above conditions, I absolutely need visual confirmation of the music details in all but the rarest of circumstances.
 * If I can't see an appropriate release with details that contradicts what is currently on Wikipedia, I will almost certainly not update the article.

So how do I determine the appropriate release format when the formats overlap (such as LP vs CD in the mid-1980s or 78 vs 45 or LP circa 1949-1955)?

 * That's another tough question.
 * For LP vs CD in the 80s, it's pretty rare that the LPs and CDs have contradictory details. And when they do, falling back on my general philosophy, if I can't make an edit with confidence, I usually won't make the edit.
 * For 78s vs 45s or LPs, it's trickier. Unlike the LP to CD switch, which pretty much all labels encouraged, switching away from 78s was tougher.  Not only did the industry have to deal with consumers with existing equipment, e.g., record players vs CD players, or, in this case 78 rpm record players vs newer-format record players - whatever they may have been.  In the 1980s, all labels were gung ho on CDs as they were cheaper to mass produce and had fewer returns (like scratched or warped LPs).  But in the late 40s-mid 50s, it wasn't that simple.
 * Think VHS vs Betamax. Columbia Records started producing 33 1/3 rpm LPs in 1948.  RCA Victor introduced the 45 rpm format in 1949.  In general, the industry went with LPs for, well, long-playing releases and 45s for singles.
 * But there are still wrinkles. In this era, a 45, or even 78, might be the predominate release for some song despite the availability of an LP with the same song, for example.  It depended on the label, e.g. RCA Victor preferring their 45 rpm format vs Columbia preferring LPs, but also on the release itself and what equipment those consumers had.  (People didn't upgrade hardware nearly as often then as they do now.)
 * This can be pretty tricky.

Speaking of 78s, how do you handle a single 78 released on multiple labels?
In the early 78 days, it was fairly common for an artist to have a track released on multiple labels, and possibly with different artist names. An example is the song "Ory's Creole Trombone" which was released in August 1922 on "Nordskog 3009" with the artist as "Spikes Seven Pods of Pepper Orchestra". It was also released the same month on "Sunshine 3003" by the artist "Ory's Sunshine Orchestra", which I believe is the same artist.

Often, there is an original release followed by copies of that release made on dime store labels. An example of this is Robert Johnson who released 78s on the Vocalion label which were then copied and released on dime store labels such as Conqueror, Oriole, Perfect, and Romeo.


 * If I can determine the original label on which a 78 was released, I will use that. If I can't do this, I will not update the article without details from a reliable source which I would cite and/or discuss on the talk page, as appropriate.  At the time of writing, this situation has not arisen.

Well, now that you've got 'appropriate release details', what do you do?
Another tough question. Getting it right is "still tricky".

This is getting pretty deep.
 * Is the band name "Eagles" or "The Eagles"?
 * Or "Pink Floyd" or "The Pink Floyd" (or either, depending on the era, actually).
 * Is the track name "Goutelas Suite" or "The Goutelas Suite"?
 * Should the composer be "Pearl King" or "Earl King"?
 * Or should the composer be "Autry DeWalt" or "Jr. Walker" or "Junior Walker")?

These are tough questions (à la In Good Company, where Dan Foreman, played here by User:Wantnot, destroys his/her credibility and career in front of the whole company): ""Teddy K. (CEO)> Dan Foreman. Sports America. You ask some excellent questions. Excellent, excellent questions. I'm glad you asked them.""

But Wait, There's More
So, you've got artwork with the desired media type (LP/CD/45/78/...) from the most appropriate release country and year, etc. We're good to go, right?

Of course not.

The funny thing with finding what should be the most authoritative release is that, sometimes, due to the release being so new (typically), there can sometimes be labeling mistakes that no one noticed until after release or even years later when reissued on CD.

Sometimes the from the  has more accurate details (or more commonly, a later release from the appropriate release country).

Sigh...

However, Not All Music Edits Can Be Visually Confirmed
All that said, many of my music edits don't come from album artwork. There are many cases where this could happen but the most common is release year, though the following comments apply to other details.

In my opinion, the release year is by far the most important detail for any release (LP, single, ...). Other than serious album or track detail errors, few errors are more important than release year errors.

Unfortunately, such details can't be confirmed with album artwork. Yes, (C) and/or (P) on the artwork can be used as a basis but is not authoritative. It's not uncommon that an album/single/... is (C) and/or (P) in some year but not actually released until the following year, or even several years later, for whatever reason.


 * (C)/(P) on artwork basically means that that the release was not made before that year. It does not mean that the release was from that year.  Of course, when a song is re-recorded, the (C)/(P) will change.  In that case, you have to consider which release is being discussed (i.e., original or re-recording) when editing an article.

Since visual details aren't relevant for such details, I have to use other sources. This significantly limits my sources, as it should. For North American releases, I often use the Billboard release charts for this. See the 'Release Year for "Rock and Roll Music" LP' section of the The Frost talk page for an example of how I deal with this. I can't think of cases where I accepted anything other than Billboard as a source. If I did, it must have been a great resource. A possible example might be Anthony Phillips' awesome website, though I doubt I solely relied on even such an amazing site for anything other than confirming details before publishing to Wikipedia.


 * Whatever the resource, Billboard or otherwise, I will almost always discuss any changes I made on the relevant talk page so that other Wikipedians can confirm or refute my edit as necessary. And I cite/reference as much as I can (though I wasn't great at that in my early days :.


 * And, BTW, if you aren't familiar with Anthony Phillips' music, some of it is pretty awesome. I'd recommend Wise After the Event as a place to start.

I treat other such edits, i.e., not just the release year, just as seriously. As always, if I'm not completely sure of an edit, I won't make that edit.

= TLDR =

I try pretty darn hard to be as accurate as I can possibly be for any edit I make (music or otherwise). It's, uhm, tricky. (I might have said that a few times already.)

But, when I mess up, which I surely have done, please know that it was not due to a lack of effort or sincere desire to improve the accuracy on Wikipedia.

Please, let me know my of any errors I have made, and/or please fix them as necessary.

But, despite whatever errors I may make, please don't just change things because your CD artwork differs in some way from what is here. Please ensure that the differences you see are generally applicable. "My CD is different" may not be enough to warrant discarding or reverting a previous edit (made by me or anyone else), and in my experience, is sometimes simply incorrect.

Together, we can make Wikipedia a better place. :)