Two proposed variations on the normal Wiki signature to enable easier editing and review of refactored ThreadMode
- ?- WayneCool means the preceding Wiki contribution of WayneCool's has been edited by a third party and the amendments have not (yet) been approved by WayneCool
- !- WayneCool means the preceding Wiki contribution was edited from an original by WayneCool but no longer entirely reflects the views of WayneCool
The allowable state transitions would be:
- -- WayneCool => (nothing)
- (nothing) => -- WayneCool
- -- WayneCool => ?- WayneCool
- ?- WayneCool => (nothing)
- should only be carried out by WayneCool unless the text is being further refactored
- ?- WayneCool => !- WayneCool
- should only be carried out by WayneCool on reviewing his text edited by someone else and deciding that although he is not now happy to be seen as the "sole author" of the text he is willing to be known as the originator of it or a major contributor to it
- ?- WayneCool => -- WayneCool
- should only be carried out by WayneCool on reviewing his text edited by someone else and deciding that he remains happy with being seen as the "sole author" of the text
- -- WayneCool => !- WayneCool
- should only be carried out by WayneCool on changing his mind
I've used the third transition on a very mildly edited section by StevenNewton
. -- RichardDrake
If you can edit my stuff to make it say what I meant better than I did then go ahead and do it with out any further action. Otherwise: quote it in your own work, paraphrase it or even delete it if that's appropriate, but please don't change what it means and leave it attributed to me. -- PhilGoodwin
Agreed in many circumstances. Two situations where I don't think the first sentence is adequate or realistic enough:
- I believe that you will think that I've improved your "stuff to make it say what you meant better than you did", but I'm not 100% sure. 90% I've saved you significant time and hassle. 10% you have to put a !- on it, remove your signature or do something more time consuming. As a GoodWikiCitizen I believe that the probabilities are such that I should definitely go for it but I must inform you of my 10% uncertainty.
- An extract to a new page, a refactoring of ThreadMode, a local MindWipe or just a simple edit to a referred-to contribution (the ChristianIntellectual situation as of last week) leads to a new context, where some words in a signed comment need to be changed to preserve their original meaning. (That can only be as far as humanly possible, which is limited, but still potentially helpful to the original authors.) I've accepted the use of [this kind of thing] for inserted text in some such situations but I have also been looking for more readability.
In either case the ?- would indicate to the original author and
to third parties that something has been changed and that the signature is no longer authoritative in the "normal" sense. (And how authoritative is that? Answers please in the MeaningOfDoubleDash
.) The original author doesn't strictly need to know which words were added, deleted or changed, he or she just needs to ask "Do I mind my signature on those words
from now on?".
Whether the refactorer should also email the signatory every time he does a -- to ?- transition is a moot point. -- RichardDrake
Phil's right. This is a fun idea, but not a good one. Ordinary editing skills and good sense are much better. -- WayneConrad
There's nothing here which downplays "ordinary editing skills and good sense", far from it. One needs good sense to recognise the need to check amendments with the original authors on occasions. I'm looking for the very lightest weight way of doing that. -- rd
This seems as if it would only complicate the use of Wiki, almost like a mini--programming language
that would require people to learn it before they can understand or use it. Further, it isn't apparent why any of this delineation is even desirable. Are more and more rules embedded in cryptic ascii characters really the direction Wiki should go? Is this really a solution that highlights the strengths of Wiki? It seems the ideas on this page exist only to support ego in some way. Is that the reason all of this is being proposed? - anonymous
This is intended to be the lightest weight way to ask the question of the original author and let other know that it needs asking. No more rules are needed and the !- may not be. Note that this touches the one piece of "shorthand" Ward originated on Wiki. (Take it either way. Out and out heretic or faithful, creative disciple. I don't mind.) It does not imply massive ego on the author's part, just common courtesy on the refactorer's part. Plus a recognition that some Wiki contributors
are very sensitive to "silent" edits to signed text, something that has been brought home to me over the last six months and that I increasingly sympathise with. -- rd
I prefer, when making a significant change (something other than fixing a typo) to someone's signed contribution, to leave both the new and the old on the author's name page with an invitation to correct or undo my change. This is direct and requires no DecoderRing
I'd never say that you shouldn't do this Wayne. If you have the time and the patience, great. I'd only say that this will take longer than the above style and this might inhibit people from attempting more complex refactorings.
If I'm changing a page to document mode, then I usually don't notify the author, because in those cases I strip the signatures, usually putting all of them onto a single "Contributors" line. I've done this a few times with good success (that is, lack of stones thrown in my direction). -- WayneConrad
Well done indeed for those lack of stones! There's a preliminary issue here though. Are there complex ThreadMode situations, sets of pages that would benefit from skilled refactoring but where attaining DocumentMode is either practically impossible or even unhelpful? I agree with AlistairCockburn (who made the point again trenchantly in person last night) that the latter possibility is well worth considering (see ImproveSignalAndReadability). My view is that there are lots of such messy, intermediate pages, with some good stuff and lots of temporary, redundant or waffly stuff. And not many highly skilled refactorers. I'm not going to lose sleep about these "tools" not becoming a de facto standard. I won't use them myself for a little while, if ever, while they're debugged/destroyed on the runway here. But I believe that they might be lightweight enough to work in giving semi-skilled refactorers courage to "have a go" at improving ThreadMode without making it perfect, which I think is what Wiki needs. -- rd
it will sometimes be better to concentrate on avoiding this in the future than going back and trying to refactor something that may be IrrevocableThreadMode
. -- anonymous
Are you saying something like RefactorFasterDeleteMore? It got us into a lot of trouble last time that seemed to be the consensus. I still feel that something like this, without the "in your face" deletion of whole pages, will be necessary to avoid IrrevocableThreadMode happening so much.
I think the skill is in choosing which pages to refactor. I've learned that some long-dormant pages are easily refactored into document mode without anyone getting too upset. Other long-dormant pages, I look at and can't figure out what to do with them. I leave those for smarter refactorers. If the page isn't stone cold, I fear making much of a change to it. -- WayneConrad
I respect your efforts a lot Wayne but the metric here concerns me quite a bit:
"without anyone getting too upset". AlistairCockburn and other later readers might take the view of such a refactoring that he describes in ImproveSignalAndReadability. But in the nature of things they won't notice till later, they probably wouldn't know who to be upset with or where, they most likely won't be that upset, just resigned to more loss of signal. I believe with Alistair that Wiki's emphasis of improvement only coming from moving directly towards DocumentMode has become a pretty faulty and impractical one. It's fresh in my mind from a stimulating meal together with him last night. -- RichardDrake
Yeah, editing old pages could easily be a way to dodge responsibility. So far, it hasn't worked out like that (I don't think
). Usually people don't seem to care - nobody undoes the changes, and usually nobody adds their own changes. That means that the page still has my name stuck to it so people know who to go to. People on wiki have always been very willing to speak out when they think I've goofed, so when nobody says anything, I take it as a good sign. The only dodge taking place is that I keep forgetting to turn cookies on when I'm Wiki-ing from home, so my edits end up without my name on them - not the greatest way to take responsibility. -- WayneConrad
Wayne, thanks for sticking with this. Ok, you sign, you use your UserName (mostly), you're a great guy and they'd probably see it was you that last touched the page. (Unless by putting it back into RecentChanges you trigger completely new ThreadMode from people who've just discovered or remembered it. We've all seen that happen too!) A stronger point is that one seldom has the original ThreadMode to compare with and even if one does it's a major effort to evaluate it against the "cleaner" version. Who with any kind of charity or sense of gratitude or even aesthetics wants to add a "I preferred the previous version" or the even more pathetic "I would have done the refactoring better" at the bottom of a "clean" page after considerable effort.
The issue of "where" critical evaluation goes for me is more pressing than the crude "who's to blame?". I now feel that in creating WikiReductionists (the page) I inadvertently met a big pent-up need for the "where do I complain that Wiki isn't as great as it used to be?". It and the nastier spin off pages that referenced it (or us) became the places for Wiki to release its negative emotion at that time. The negative emotion wasn't being harnessed to produce positive refactoring effort and we all knew we were creating something more like a pig-sty than it ever used to be. It wasn't a good feeling and it was cool to be able to blame a small group for some easy to understand, grave errors (real or imagined) that took our minds off our joint responsibilities.
Rather than such a destabilising flight from reality I want a lot more imperfect refactoring around here, so that people can release their sense of frustration in a positive way. I admire consistent refactorers like you who almost always get it right but I want more, far more, who aren't that good. I want them to know that they can make mistakes and that it's Wiki's great strength that it easily copes with this over time. I have a dream ... oh no, wrong speech. I do think that something like "?-" could really help but I may have veered off somewhere back there. -- rd
Let me give an example that presented itself yesterday. On saying au revoir
(not goodbye) to AlastairRae?
on his home page I wanted to give a short history, with references, of what various folk on Wiki had got up to on Christian or related themes over the last year or more. One page I remembered well was ReligiousWar
. Or so I thought.
I'll make the following observations (section details will soon be out of date, so hurry through this):
- I was involved at the time and was the first person to respond to the original creator of the page (called Nigel, now showing in section 3). My response was anonymous and now heads section 6 (thanks guys!).
- The fact that my initial contribution no longer follows Nigel's means that it has been rendered pretty inane, even more inane I would argue than Nigel might say it was originally.
- The prevailing ethics of Wiki have been followed. None of my contributions, including my first signed contribution (in section 6) have been changed as far as I can see (although "WhiteHat" is arguing in the second section for the removal of all of them, with a lot of others, for the greater good). As I've been saying elsewhere I don't think that this approach to Wiki ethics is good enough. By the change of context my meaning has been changed. Without my permission. Meaning is more than words, it also has to do with context and cotext and referents.
- I didn't notice the latest changes to the page, I guess because I was away from Wiki in August. Someone clearly tried to start a refactor and, as in other situations, the page "got stuck" in an intermediate state (for 44 days as of today). Am I to blame for not spotting the changes until now? TomStambaugh argues powerfully "No" in WhoStuffedWordsInMyMouth and I would thoroughly agree with that.
- The refactoring efforts so far have not in my view met the goal ImproveSignalAndReadability.
- The issue of signal is subtle. Although I say it myself, I think that my first anonymous response was a good example of a polite reply to not such a polite beginning of a page on Wiki. Such an example in itself is signal, in an important sense, signal that affects the future quality of our dialog. This has already been lost from the page.
- It isn't at all easy for me to understand who has been responsible for the refactoring in this case.
- If I wasn't writing about this here, arising from my dialog with WayneConrad, should I really pitch in on ReligiousWar itself, with any kind of complaint? Let me say for sure that I wouldn't. I want to encourage more refactoring and I don't think that such complaints help. But I am very interested in how we can make refactoring as fair as possible to people who have already invested time and effort in a page. I don't think that we have "cracked it" yet. Hence the new idea on this page and my legendary openness to other new ideas.
's experience with ReligiousWar
does point to a need for more careful editing of pages. We only disagree on how. Anyone who is aware enough and cares enough to use the FooDash
convention is someone who would do a good job editing without it. If someone changes the meaning of a thread then they're either new to this or having a moment of carelessness. I don't think that FooDash
will help in that case.
Good editing comes from being thoughtful and careful. Someone having these traits doesn't need conventions. Someone not having these traits won't be helped by conventions. -- WayneConrad
"If someone changes the meaning of a thread then they're either new to this or having a moment of carelessness"
I love the confidence here Wayne but I don't share it. I think editing ThreadMode
without changing meaning is very hard. In fact, apart from the most trivial changes, "impossible" is the word that comes to mind. What I think you're talking about are changes of meaning that are big enough to bother any of the participants. Even if I take this as your meaning (correctly? huh?) I don't share your confidence. AbsenceOfEvidenceIsNotEvidenceOfAbsence
of contributor discontent. Some people leave silently.
I agree with you that tools are less important than skills, habits and motives, but not
that they are irrelevant. I agree both with GkChestertonOnWiki
and with Ward enjoying his WikiWikiKudos
. -- RichardDrake
See the earlier PlainEnglish
discussion. Also, the WikiNewspaperAnalogy
is probably suitable for review.
As an alternative to the FooDash
proposal, perhaps people should feel more free to make changes in "editorial brackets". Examples (from what used to be text above):
- Changing a word or phrase: "Are you saying that our [inspiring example] in"
- Deleting text: "from the beginning. [...] It's worth thinking about when"
- Noting other authors: "-- rd [and others]"
- Noting possible disagreement " -- rd [and others who hope he won't mind the changes]"
- Noting disagreement with the current version: " -- rd [who now disagrees with these edited ideas]"
The editorial bracket convention is common and requires little explanation. ("Text in square brackets like [sample] is commonly used to indicate editorial comments or corrections by people other than the original author. Some authors also use brackets for parenthetical remarks within their own text.") This convention is even widely accepted for direct quotes in the established print media. (Direct quotes are (arguably) the closest equivalent to many Wiki signatures.)
Are there any cases where a new signature convention works better than editorial brackets? -- CliffordAdams
On using editorial brackets within the text before a signature
Using no square brackets and a FooDash
would almost always make the text before the signature easier for third parties to read and usually easier to produce. And marginally easier for the signatory to say "yep, I still believe that" or "nope, that's not me any more", rather than having to remove the brackets in the former case (not that people often get round to doing that).
Editorial brackets are better for showing which parts have been changed, that's also clear. I'm not saying "never use editorial brackets in text", I'm saying that readability for future generations is important, being able to trust the integrity of signatures is very important (both methods address that) and here's another option that could be somewhere between mildly interesting to quite a major breakthrough.
On using editorial brackets with PlainEnglish after signatures
I like all the examples you have given and would be glad to see a Wiki culture that used them much more. It would take a little longer to do than FooDash
of course. If FooDash
doesn't take off and save the planet, this would be a great way forward. Even if FooDash
was used on some pages during refactoring (with a reference to this page at the top say) I'd be very glad to see increased use of editorial brackets for this kind of thing. Even with
it would be helpful at times. (Couldn't resist just now in WhoStuffedWordsInMyMouth
, for example.)
But next time you propose such a sensible idea, Mr Adams, please say clearly that I need to read everything you've written before I reply! -- RichardDrake
Horses for courses. I never cease to be amazed how intensive it can be to try to coordinate and get agreement among n Wiki editors, where n is greater than 2. Such attempts as happen on the face of Wiki have sometimes really wrecked readability for future viewers. In this area, as in WikiBallotBox, I would like a really small, asynchronous convention, easy to learn and recognise, that everyone shares. But I can't decide culture around here. And I like the PlainEnglish alternative from Cliff. -- rd
See also MeaningOfDoubleDash