Not dead, just sleeping...
Greetings, any Everythingites misguided or unlucky enough to stumble onto my home node.
Since I have your attention for a few moments, I'd like to say a few words, and possibly correct a few misapprehensions you might have.
Upon looking at the above stats, you might think something like "whoa, crappy user who added a few w/u and then pissed off. What a loser!"
While I don't dispute the loser part :), I would like to tell you of why such an impression does not fit the facts.
When I first encountered Everything, it excited me greatly. What a wonderful idea! What a glorious project! And so I wrote all sorts of stuff, some of it even competent, and in no time was almost at Level Four of 'achievement'.
However, at this point I became angered by some of the inequities and hypocrisies of the Everything implementation (and some of the users), and removed my writeups as a gesture of protest.
That action might have been a little extreme; I can only plead that my brain biochemistry, and thus my emotions and reasoning, are somewhat misadjusted as compared to what is considered 'healthy' for humans.
Anyway, after some thought (and sulking) I added some of the better of my writings back onto the system. As a result, I have more than enough XP to coast into Level Three - if I wrote a few more writeups, I could easily do so.
Deciding that the merits of the Everything system outweighed the detriments, I wrote some additional things, some of which received an angry response from the established userbase. One of my nodes got nuked, unfairly I thought, and I wrote a node protesting this action - which also got nuked.
Censorship is one of my pet hates, as is the abuse of power. After these actions, I decided that it would be more than my time was worth to contribute further content to the Everything system, and resolved to spend no more time adding to a flawed system.
Since that time, I have devoted much thought to the design and implementation of Everything2, and have produced a long list of areas which (at least in my own, ever so humble, opinion) could use correction. I believe that the Everything implementation is poorly optimised, and that certain of the major features are simply wrong. These conclusions strengthen my earlier decision.
I wrote a summary of my conclusions: "Everything2 is a wonderful idea, held in thrall by inept conceptual design and poor optimization, while being further handicapped by certain of the userbase." While I'd like to further contribute content in the way that Everything users do, I cannot in good conscience do so under the existing Everything2 system.
Perhaps one day I will contribute to a system similar to Everything, but correcting the deficiencies (as I see them) present in Everything2. If such a day ever arrives, I would like to invite you to join me there.
To summarize: No, I'm not dead. No, I haven't left, I just am not writing much anymore. No, I didn't (mostly) write crap. And no, it isn't (just) sour grapes.
I'd like you to think seriously about Everything2's design, and also to encourage you to publish what you write elsewhere as well as here, as well as keeping a copy of what you write on disk. And be warned - using any system with poor design can frustrate and anger you.
Thank you for taking the time to read this. You may or may not agree with what you have read - that is your choice. All I ask is that you give what I have written a fair hearing in the court of your opinion first.
Peace and love to all.
z_evil1Note to editors: please LEAVE THIS HERE!!!! It is here and not in Suggestions for e2 for a REASON. Thank you.
This document contains a list of changes and corrections that I feel e2 would benefit from. Peruse at your leisure, if you desire, and remember, this is just my opinion. I'm not demanding these occur - however, I would be very happy to see those responsible for the current state of the e2 system taking a good hard look at the contents herein.
Before I start, let's just step back a bit and look at why we are doing the whole Every thing? Here's my understanding of it:
- e2 creators: They make and administer e2 to make the world a better place, and also to gain status. ("I _made_ e2, therefore my dick must be huge.")
- e2 users: They contribute their time and energy in adding content and rating pre-existing writings in order to make the world a better place, and also to gain status. ("I spend many hours a week writing grteat stuff for e2, therefore my dick must be huge.") They might also want to troll, spam, graffiti, vandalise, demonstrate how studly they are (not) by h4x0ring or proprieting multiple users in order to mutually vote each other up, or just have one thing in their pathetic, stunted lives that they have some vicarious control over - we must act, if possible, to render the latter abuses of the system as impractical as possible, if e2 is to be of use to the world.
Some users will write better content than others, and some will write content that other users will rate as better. (Note the distinction!) To make e2 of more value to the world, we must reward better content - this might be by recognition and subsequent status enhancement ("People reckon that what I write for e2 is pretty cool, therefore my dick must be huge"), or it might be by granting these users more priveleges, provided they can use them responsibly.
Users will get better at writing for e2 over time, and they will be inexperienced at most e2 tasks to start. So, we should limit the damage they can do to start off with, and allow them more latitude as they prove that they can use their priveleges responsibly.
The users of e2, being drawn from the Internet as a whole, will tend to be a miscellaneous lot. Mechanisms to protect minorities from abuse by the majority would therefore be appreciated, and would likely lead to a healthier database overall.
Whatever one writes, it is bound to offend someone. If you are writing for a despised minority (despised for whatever reason - remember school, Linux geeks?) then you will be offending most people. Mechanisms to prevent opression of unpopular content, and in particular arbitrary removal by editors serving as self-appointed arbiters of taste, will lead to a more diverse and healthy database that is of more value to the world as a whole.
Users of e2 will inevitably want to compare themselves against others, and so a robust and fair method of rating what they have written (true node-fu) would be appreciated for that purpose, as well as helping to guide authors in deciding what is most appreciated by other users, as well as what is considered valuable. An unfair system, vulnerable to abuse like the current system, will breed resentment, alienate users and decrease the quality of the database as a whole.
An e2-type system should address all of these areas by appropriate policy and algorithms, therefore, in order to support all of these aims, and to properly support the aims of the system, namely (as best as I can tell) to rule, to create a massive and awesome Web resource, and to give talented writers an outlet for their genius.
In my understanding, the current e2 system only addresses some of these areas, and some of the areas it does address leave a lot to be desired. The e2 idea - a communally-edited, collaborative, free-form database open to anybody on the Web - is a brilliant one, but the way the current e2 system supports this idea with algorithms and policies leaves a lot to be desired. However, until now it has been the main system of it's type, and even if it is flawed, 'half a loaf is better than none', and so people have contributed vast amounts of effort to enriching e2. I have contemplated somehow starting a competitor to e2 - an 'e2 done right' - in order to address the numerous issues I have with the current e2 design, but why waste all the effort put into e2? If e2 can be fixed, it should be. A fixed e2 would also avoid all sorts of other problems - laziness on my part would delay implementation considerably, users would be split between the two systems, holy wars and arguments would spark, and all-new content would have to be entered into the new system. I may yet stir my butt into taking just this course of action if e2 continues to suck, since I think the e2 concept is too brilliant to be hamstrung by the lackluster e2 implementation. Using e2 in it's current form is like swimming with your legs and one arm tied behind your back, or using a computer with a Microsoft operating system.
In the interests of a better world, I have developed RSI authoring this extended diatribe and wishlist. I submit it for your information, and hope you gain some benefit from reading it. Also, I would humbly request that you not be too harsh on me for my writing style - I am nowhere near as brilliant as I wish I was at the moment, and this state, sadly, reflects in my work.
Note to bones and co.: please, please, please, PLEASE read this and give some serious thought to the issues it raises! And don't be too pissed off by all the criticism. You have done some brilliant work in e2, and I could make this document even longer by pointing out areas of e2 which rule and need no improvement. However, IMOSHO there are many areas that DO need attention, and this document covers as many of them as I can think of. e2 (and the world) would be better as a result of their improvement.
Now, the ways that I reckon e2 should be changed (improved?):
Retire the poorly thought-out XP and level advancement systems
Voting and XP changes:
- Make node-fu an official rating, composed ONLY of reputation gained from votes of other users
- No vote bonus, vote exhaustion bonus, or node creation bonus - that's right, none whatsoever!
Why do this? It eliminates the temptation to node-split or NFN in ALL it's forms. Users are not tempted to vote dump. And users are not rewarded for voting - voting is it's own motivation and reward, having an XP reward only encourages thoughtless and compulsive voting. Compare the number of new writeups per day with the number of votes available to distribute...
- Vote banking - votes do not expire.
- Reduce number of votes available per day.
- Make Cools bankable, and reduce the number available greatly - some days, nearly all submitted nodes are cooled - is this what we want? Surely only the best of the new nodes, no matter what the average quality, should be Cooled?
- Review the Cool vote bonus (for recipient - no bonus for giver, of course) - I would suggest a drastic downward revision of it's power!
There are many, many possible schemes that could be implemented. I would lean toward a reputation-only level advancement scheme - i.e. the more rep you get, the more power you have. Why on earth should the number of w/u be significant in level? Isn't one carefully written and well-researched w/u worth fifty puerile and assinine personal opinion nodes? I would also suggest a node-fu modifier for higher levels - you have to have a high node-fu to attain the topmost levels. This would help limit abuse by users proprieting multiple, cross-voting accounts, 'bootstrapping' themselves into the upper levels.
Such a scheme would be simple to implement and understand. I would suggest an increment in node-fu of at least 1 per level - new users require 2 to advance, 3 to next level, and so on, up to 15 for Psuedo-like levels.
Current levels and XPs:
Those who gained by an old system will always be threatened by a different, new system. As I understand it, level and XP seems to be at least partially dynamic - so a change to the calculating mechanism will result in a drastic shake-up of levels. However - if the new system rating is more 'fair' at rating a user's contributions, then of course they should shift around - users who previously languished in the shadows despite valuable contributions will gain their deserved place in the sun, while those who mainly wrote vast numbers of nodes filled with crap and obsessively vote-dumped will get relegated to the small time, as is their lot. Perhaps it's time to trim the dead wood?
Anyhow - if the opinions of those who stand to lose under a new, fairer system are valued, then there is nothing to stop their old XP and level rating being posted alongside their new node-fu and rep ratings. Perhaps, although this would be grossly unfair, they could have their old level and XP converted to equivalent node-fu and rep as of the changeover? Of course, if they persist in their old habits then they will slowly slide back down to humbler levels...
Level advancement and retreat
If your node-fu and rep are sufficient, then you may gain a level, with the associated benefits of more votes to award, the power of Cool, associated status, and perhaps even other treats. However, other users may gift both positive and negative rep - you might skate over the level threshold multiple times, if you post controversial content at that point. So, I propose at most one level change per day, and that if you slip slightly under the level requirement, that you retain your high level. (Level rating -5% sound fair?)
Change other limits in the system
Writeup quota per day
At the moment, anyone may write as many nodes as they like. This lends itself to abuse by script-assisted persons or bots, or persons who like to split up nodes which really should be in one piece. IMHO, this leads to a decline in the quality of each node. If a user was only permitted a certain number of writeups per day, this might encourage them to spend more time researching and polishing their nodes before submission, as well as (under the current system) reducing the bonus for node creation awarded per unit time, all other things being equal. As a user increases in level, they may author more writeups per unit time - since they have proved that what they write isn't mindless drivel or crap, they are thus not polluting the database.
What to do about bots?
Bots take pre-existing content from elsewhere, re-package and re-format it, and then regurgitate it at high speed into the database. If a bot was forced to climb the levels, same as a person, then this would act to screen the quality of the content they provide. (Remember, if node-fu and rep are the only determiners for level advancement, then users advance in level by writing material considered to be more and more valuable to the database.) An alternative system might be to provide 'bot accounts', where an editor or admin authorises a login to ejaculate content at high speed into the vitals of the database, subject to that editor/admin's approval of the quality of the content that the bot is supposed to be emitting. However, this idea is prone to numerous abuses, not least that it relies only upon one person's opinion and bypasses the peer review process, and that the bot might provide a few pieces of high-quality literature to evaluate, then simply emit shite the rest of the time. As a result of these considerations, I would urge that bots be subject to the same limits as other, human users.
Rewards for level
At the moment, the rewards for posessing a higher level are:
- More votes - you have more say in rating the content of the database
- A Cool! or two - point out nodes of hiher-than-usual quality
- Editor powers - if you are high level enough! More drastic powers of rating than normal, and the power to censor. Is that such a good thing?
- The power of invisibility in chat - yay.
- And let us not forget the enhancement in status afforded by being able to say 'look, I am high level, therefore my dick must be huge' (you know...)
Under the proposed changes, the 'rewards' would become:
- More votes -
One vote per day for level 1, two for level 2, four for level 3, and so on
- More writeups allowed:
One w/u per day for novices, two for initiates, and so forth - perhaps not a binary progression, as for votes...
- Cools awarded
I propose that there should be a fixed fraction of Cools per writeup availabe - say one Cool per twenty writeups. This way, the best of what is coming in is pointed out as what you should strive toward - which would ideally lead to continuous improvement in the quality of writeups coming in, somewhat like the principle of the donkey and carrot on a stick. (One disadvantage of this scheme is that, as the quality of nodes climbs to stratospheric levels, novices become discouraged from joining. I would argue that the advantage of brilliant nodes would outweigh this disadvantage.) With this system, Cools are awarded (for gifting) to users in ratio of their node-fu (remember - node-fu is the default measure of overall user quality), or perhaps to users in the top five percent of node-fu in ratio of their node-fu. One disadvantage of this system is that low-level users may whore to the tastes of the highest level users in order to receive Cools, or that high level users might only Cool each other - possible fix: you may only Cool those of lower level or node-fu than yourself.
With this system, if two thousand writeups come in one day, then one hundred Cools are available. However, if only five writeups appear, then there is only a 25% chance that anyone will receive a Cool. Note - It would probably be best to award Cools randomly within the node-fu dictated ratio - high level users should not get limited Cools preferentially.
- As now, really high level users get the power to edit node titles, edit node contents for nodes created by now-dead users, split up huge nodes, and remove nodes AT AUTHOR'S REQUEST ONLY - although the latter feature would be best automated. They might also get powers to edit 'system' nodes, important to the functioning of the system (such as the Everything FAQ), and to write on the front page or in possible news and discussion pages.
Changes to features of everything2:
Display of node reputation
I think I speak for most Everythingites when I say I would like to see a writeup's reputation BEFORE I vote on it. I might want to use my vote powers to modify the rep towards what I think it should be, rather than voting up a node in case it doesn't have nough rep and finding that it is actually 400+ rep - I have just effectively wasted a vote, in that case. Always display node reputation.
Editors having the power to unilaterally remove w/us that THEY don't like is a Bad Idea. It bypasses the peer review process, infringes on the rights of those who might want to view such w/us anyway, and gives Editors powers over database content that they should probably not have. They have votes, just like everyone else, anyway! I propose that ALL nodes, no matter how low the rep (except when we are getting ridiculous, like w/us that infringe on the Terms and Conditions of Use of the system) shall remain available for persons to view, if they so desire. Who says that the majority is always right? The majority has been, and I would claim still is, usually wrong on most matters. Besides, some nodes might be of interest to only a small subset of the database's users. Who says that you have to write for everybody? And no, you don't have to read these objectionable writeups if you don't want to and the next feature is implemented...
Learn some lessons from Slashdot
Slashdot comments on a particular comment are rated by priveleged users. However, everyone may choose to see the lowest rating that they wish to see. This lets them screen out content that the majority of others deem to be unacceptable or uninteresting. While simple and limited, this system works fairly well. A more sophisticated system for Everything might allow filtering by rep, and an alternative system where a form of meta-moderation is employed ...
Say you are a user who has interests substantially different from the main mass of Everything users. (Sound familiar, computer/Linux geeks?) In this case, you don't want to 'see' the ratings of people whose judgement you don't trust or disagree with. As a result, you can rate the relevance of a particular user's votings and opinions, and filter their voting contribution out altogether or partially if you don't like it. You can rate them for yourself, or submit your rating on their ratings for the community as a whole to use as they see fit. Perhaps they agree with this user, in which case they discard your rating - of course, such is their right. Perhaps there are bigoted users who you especially despise for their ignorance and intolerance? A negative recommendation from these people might be a positive recommendation for you! Such a person has a negative weighting in your moderation settings, instead of zero, one, or somewhere in between. Such a system, while no doubt a complete bitch to set up, would allow minorities to use the database without being oppressed by the opinions of the (possibly ignorant and wrong) majority.
A default setting for this system might be all or none of people's opinion. That way, you could start with trusted user's opinions, or filter out those who you find particularly objectionable.
Who says that a writeup is set in stone once it appears in ENN? Perhaps the author maintains or updates it. Or perhaps he goes nuts and screws it up entirely. In such cases, you'd want to review your earlier vote, wouldn't you? I propose that you should have the option of altering your vote on a given w/u whenever you like... subject of course to the number of votes you posess! (Altering a vote does not use votes, if you have already cast a vote for that topic). Also, changed nodes should be signalled, just as new nodes or writeups are. (A limited number of node changes per day should be implemented in such a case, to prevent users spamming the ENN nodelet, and to stop users from removing content out from under people's feet when voting, thus making respectable persons appear to be voting for trash.... Perhaps the same number of node edits per day as node creations? If you run out of node change allowance one day, the system might post news of the modification the next day for you...)
Right First Time
Well, I don't know about the rest of you, but I definitely carry around inside my head, in addition to the complete contents of Webster 1913, a complete image of exactly how my current node is to look, as well as a database of nodes that it is suitable to hard-link to. I can soft-link all relevant nodes with the greatest of ease, and in short, my nodes are all wonderfully grammared, polished paragons of perfection by the time they hit ENN. That's right, every single one of 'em. For the rest of you mortal types, however, the almost complete lack of node editing facilities surely contributes to the fact that everything you have written so far goes chomp chomp chomp on the nutsack of Satan. Here are a few ideas for things that could help you improve the quality of what you write, before _anybody else_ gets to see it!
Node Edit Area
As the system currently stands: Either you are expected to do 95% of the work of creating your node offline (you know... _premeditated node creation_! a woefully rare thing), doing things like writing it, pretty-formatting it, running it through ispell and your favourite HTML formatting script (oh yes! that's right... some of us either don't have access to UNIX/Linux or aren't wizards. How remiss of me) and so on, and then making about a billion requests of the e2 engine to check the validity of intended hard and soft links. (You _do_ take the time to make each and every one of those time-consuming requests, don't you?) Then you merge the results of this research into your carefully-tuned gem of inspiration (don't screw up any of that formatting, now), cut-and-paste the whole mess into your browser window (not too much at one time, X goes into a thrash loop with giant cut-and-pastes, and Netscape only allows small pastes to be added to a text box at a time), submit it, and pray you can fix the inevitable massive buggerups introduced somewhere in the process before your new work hits ENN and a million bored users swarm all over it, looking for the tiniest excuse to vote it down to Hell and send you copious abusive messages about how much your latest work sucks...
Surely there is a better way?
Here is my proposal:
Instead of a textbox and a submit button, let's just have a button at the bottom, saying 'Add new writeup' that takes you to (say) the same page with a bunch of editing tools provided at the bottom - for instance, a textbox containing the raw text of your new writeup, boxes for mass searches of the database for other nodes that might be suitable for hard- or soft-linking, a spell-checking option (pipes through something like ispell with Everythingized dictionary on server side), search for relevant nodes button (checks which words or phrases of your text are pre-existing node titles... good for total linkage, or just to show you what's possible), 'preview' button, and of course, once you are happy with the way it all appears, the all important 'Submit' button, with the option to not appear in ENN! of course, having written a node that doesn't suck for once with all these neato new tools, why would you want to hide it from others' eyes? And for those REALLY big node projects, or for when you run out of node change allowance for the day, a 'Save' option for a partially complete node might prove to be useful... (limited quantity of unfinished nodes at any one time - I suggest five)
The current person-place-idea-thing categorization is about as useful as the Usenet top-level designations or the new Web top-level domains, and even less useful than the existing Web TLDs. Perhaps a few extra categories? (animal, mineral, vegetable...) Or perhaps, we could add a tree structure to the database, with your writeup or node slotted in under the relevant category. (Root->Concept->Computer-related->Adult->pornography, or Root->Thing->Mineral->Furniture->easy chair...) Symlinks in such a system might have a certain utility, as the same node would do well to appear in multiple places at once... It would also be a LOT easier to find other nodes relevant to the one you are writing, for hard- or soft-links, also.
Simple change... softlinking is OFF by default. This way, you don't have to go to another node via your home node if you don't want to tie the two together. ("Hey... who softlinked "masturbation" to my "gospel lyrics" w/u, and why?" "Sorry, but I was reminded of it by something in ENN...")
No more drive-by votings
Perhaps we are getting into controversial territory here, but I think it would make for a better system if your vote towards a writeup, positive or negative, is accompanied by who cast it, when, and an optional area for a comment (private messaged to the author only, if the author addresses the area then the comment becomes obsolete very quickly, thus it should not become public). This way, an author knows what people don't like about what they wrote, and they get to see which people are doing the voting up or down, so that they may if necessary adjust their moderation appropriately, indulge in revenge >:), request more details on the voter opinion through the Catbox, or simply be reassured that all those worrying negative votes are coming from "the usual gang of idiots" and not people whose opinions you value. This would also tie in nicely with the proposed moderation system, mentioned above.
Should node change anonymity be allowed?
Sure, you're allowed to add a node secretly, by not allowing it to appear under ENN.. this hides crappy nodes from their rightful downvoters, however. Surely you should be made to stand by the content of your w/us? Also, changes suggested elsewhere in this writing would largely eliminate the need for traditional, legitimate uses of secret node posting...
And all the other crap...
Here's a list of other possible changes and modifications that might make e2 nicer to use.
- Colour coding of users' login names - See at a glance who is dead, who is logged in, and who has been away for how long.
- (Colour?) coding of comments or writeups - See at a glance whether a writeup is factual, supplemental, opinion, personal expression, complaint, praise, extending, correcting, differentiating, speculating, protesting...
- Selectable CSS - By switching which CSS you use, you can gain the benefits of themeability without the server having to reprocess every request, just for you. Plus, you could perhaps tune and tweak your own personal CSS, for if the existing ones don't quite cut it - perhaps you hve some nice fonts on your system?