Every edit must stand on its own feet. Before making multiple edits, think twice about the order in which you make them. Do not make an article worse with one edit and justify it on the grounds that with others coming soon it will be better. It is weak to argue that it only causes limited disruption in the interim, because every edit is an edit. You cannot group edits together, categorize them or in any way make several edits into one. Articles are visible both to readers and to editors at all times, and members of either community who see the article degenerate may be confused about your intentions.
Reculer pour mieux sauter. Retreat to make a better jump.— Napoleon I
It contains the advice or opinions of one or more Wikipedia contributors. This page is not an encyclopedia article, nor is it one of Wikipedia's policies or guidelines, as it has not been thoroughly vetted by the community. Some essays represent widespread norms; others only represent minority viewpoints.
This page in a nutshell: Small changes are good, but each change must improve the article and preserve its integrity.
Notice how Edith, who goes by her nickname "Edit", prefers to stand on her own feet.
This does not mean of course that every edit towards that goal must be independent of every other, only that it must move towards that goal and not away from it.
A foolish consistency is the hobgoblin of small minds. — Ralph Waldo Emerson
For example, augmenting or formatting references can take some time and be done over several sessions. While doing so there may be a combination of styles in the references list, but on each edit it becomes slightly more consistent and so slightly better. If, in the alternate, the <ref> tags were renamed to something else in the meantime so that the reference list was consistent — consistently empty — that would be a backward step since the primary purpose of the reference list is to enable readers to find references.
Readers
Toto, I've a feeling we're not in Kansas anymore. — Dorothy Gale
Remember that at any time a reader can view the article, and remember that the vast majority of Wikipedia users are readers, not editors. If a reader sees a degenerate article it does not give a good impression of Wikipedia. Ideally, every edit makes an incremental improvement, large or small. Sometimes this means having to plan carefully how to make a large edit in small stages, so that at each incremental stage of improvement the article still stands on its own feet.
Another editor may wish to work on the article between your first and subsequent edits and revert or undo your changes or add their own, which may disrupt your ultimate goal if they are unaware of your intention. You may then have to incorporate their edits into your own changes later. For this reason, it is best to ensure that each edit can be justified in its own terms, that is, that it stands on its own feet. While there are several means to help stave off unnecessary attention by other editors, the article is still always live.
Many editors keep a list of articles on their watchlist and will look at an article and start editing it as soon as you make changes, even if it has not been changed for months or years. This can be surprising, even frustrating, since often it seems little attention is paid to the article's talk page, yet as soon as you make some edits lots of editors start to take notice. This is not opportunism or bad faith, but because the article has been stable for a time other editors with an interest in it will look to see what changes have been made and why. They may even have forgotten that they had intended to make more changes themselves, but because their edits stood on their own feet it was not that important because the article was still an improvement on what went before it, and the change reminds them and stimulates them to add their own changes.
This behaviour by other good-faith editors exacerbates an editing style where individual edits do not stand on their own feet: a backwards step is likely to be undone or reverted. On reflection, it is not surprising that if a stable article suddenly gets changed, especially if it is by an editor who has not had any involvement in the article before, other editors will scrutinize the edit carefully. Stability does not mean that the article is not cared for.
Using "Edit in progress" template tags
A "Beware of Dog" sign hung on the gate where the pit bulls escaped, indicating the owners knew the dogs were dangerous. — various websites.[Note 1]
The templates {{underconstruction}} and {{inuse}} can be used to show that an article is incomplete and rapidly changing. They warn readers and editors that an editor is actively working on an article, so not to expect it to be complete and perhaps to be more cautious before making changes that may conflict with another editor's. Their legitimate aim is to avoid duplicated or conflicting work between editors and to warn users that the quality of the work is less than some end goal that will be achieved reasonably soon. But the quality of the work should always be no worse than the version before it. Using these templates is no excuse for making edits that do not stand on their own feet.
If every edit stands on its own feet, the {{inuse}} tag is, strictly, unnecessary. However, it still serves the purpose of giving pause to other editors who may otherwise, from a legitimate starting point, take the article in a different direction from that which you had intended. Do not cry wolf by overusing it.
{{underconstruction}} is different in that it is intended to show that the article is being actively edited but not in the very short term (e.g. the user is going to sleep or to the library, which often of course is the same thing). It does not imply even informal protection of the article, simply that it is recognised it is not finished. Consider using more specific tags such as {{unreferenced}} or {{cleanup}}.
Make small edits
Add little to little and you have a big pile. — Ovid
By "small edit" we mean less the actual size of the edit, but that it addresses one particular concern about the article, however many characters the change that is. So, adding wikilinks is a small edit; augmenting references is a small edit; changing an erroneous spelling is a small edit; doing all three at once is a large edit. Of course sometimes in passing one might spot a typo or something that can be fixed without comment, but think carefully before trying to mix two types of changes in one edit: is it really a typo or just spelt in a different variety of English?
Making small edits that stand on their own feet also allows other editors to use differencing tools to see exactly the nature of each change, and good edit summaries can also help enormously to explain the reason for the change. In the edit summary, don't say what, say why. Necessarily the two overlap, but there are tools to say what, to give a short reason for the change is very useful, even if it is just "typo": it may save other editors the need to look at the change at all, if they understand its gist.
Limitations in the differencing tools may make it hard to spot small changes when bound together into one large change. It may be difficult to spot small slips in white space, sentence reordering, and so forth. This may incline other editors to revert the large change if it seems drastically different from the previous version but they cannot directly compare them; for example if the differencing tool does not recognise the reordering of paragraphs, and marks them as deletions and insertions (as Wikipedia's built-in tool does).
Large changes get hard to justify under Be Bold!. While making many small changes should not be done simply to sneak past the attention of other editors, making small incremental changes gives a clear indication in the history of the path being followed and probably suggests the goal. Of course this does not obviate the need for discussion on talk pages and so forth, but if as often happens no discussion takes place, it can save effort to make small changes where editors may be watching the page but do not participate on its talk page. They may start to contribute after the first small change rather than having to argue against a large change that is essentially a fait accompli.
With smaller edits that stand on their own feet, other editors can just revert that one edit. Even if they have to do it manually (if other related edits have happened since), at least it allows the individual edits to be able to be referred to in their own edit summary, or in discussion on the talk page.
Most editors on Wikipedia act in good faith and so will happily keep a good edit or even improve it further. Making small edits that stand on their own feet helps towards that goal, because at least if there is disagreement over a change, all the other changes can be kept with (implicit) consensus.
Bots
For 'tis the sport to have the engineer
Hoist with his own petard. — Hamlet III iv
Bots make huge numbers of edits in short periods of time: any error, no matter how small, will quickly create problems, and it can take a long while to unravel them either automatically or manually, once they are discovered by frustrated editors who would rather be improving Wikipedia than undoing a bot's bad behaviour. The actions of a bot must be carefully thought out and tested before being "set free" to run in the mainspace.
Just because a bot has passed through Wikipedia:Bots/Requests for approval does not make it fit for a particular purpose: administrators approve it on the basis that it works correctly, not on whether its aim for a particular case is necessary or valuable (which, unless the bot is intended only to be run once, cannot be predicted at the time of approval). It can help to discuss your intent and purpose on the talk pages or on an appropriate project talk page before running it.
Most Wikipedia editors have good faith and run bots to improve Wikipedia, but it can be surprising to other editors if a bot suddenly starts changing many articles when it is not clear why, and this is especially true if bots make multiple edits which, taken individually, do not stand on their own feet. At the worst this ends in an edit war between the bot and human editors: the bot changes something, the human editor changes it back, the bot changes it again, and so on. There is usually good faith both by the human editor and the bot's author, but the bot's aim is not achieved because the two have different goals.
