When you say “not really posted” – do you mean that it’s not visible to other users, or do you mean that it’s not inalterable?
Not posted at all in fact. It’s a delayed post. But it apprears like posted to the user. When then time is elapsed the Pod process the post and its “really” posted.
It is inspired of the gmail cancel feature. It feels like posted so you act like it is posted but if needed and not too late you still can cancel to avoid the mess.
OK – to me that sounds like a preview, though. Maybe that is the actually workable solution… but I do think that what people want is the ability to actually post, and it be visible to everyone, but then to be able to correct spelling errors and so on. I feel like this feature is available on comment sections on many media websites.
Sounds like preview but is not.
Actually: I post something. I forget to make it plublic. I’m stuck. Must redo all the edit since I can not see the markdow. I probably double post because of the lag.
With the proposal: I post something. I forget to make it plublic. I still can edit and just change it to public.
- The proposal is not heavy since it does’nt deal with the federation.
- The concept has been tested and approved by gmail users. If it feels like posted then its posted.
- It can happen way more faster than changing the protocol.
- It improves the user experience as a better solution is being searched. And the better solution will just extend the edit time.
- It may be sufficent: if the complains drop drastically, then no need to deal with how to do the complicated solution.
Ideally, the edit period would be after it is visible to others (in your own pod?) but before it gets replicated anywhere – to other pods, to an email notification, etc.
But since that sounds a little complicated, Antoine’s proposal gets the job done with a nifty psychological trick – technically it’s just a preview, but the user doesn’t really know that, and the clock is ticking. This creates a feeling of urgency and concern that you just don’t get from a preview that can sit there as long as you like. You imagine everyone else looking at your post, hoping you didn’t say something stupid, scan through it again… and realize you did.
Is the federation problem actually that significant? I can delete comments, and the deletion seems to propagate out to the other servers fairly quickly. And if it doesn’t, that’s an issue with federation, not the delete function. Nobody says that we shouldn’t be able to delete comments because federation might be inconsistent. Why is editing any different?
Editing is different than deleting because it opens the possibility to two different versions of the same thing. Deleting just brings up the question of whether the thing exists or not, which isn’t as liable to cause confusion and problems.
@brentbartlett
The proposal is a step forward the goal of a post that can be edited anytime. But rather making a bing step we start with a babystep that allow to having editing in a short delay with a time restriction rather than nothing until we figure out how to deal with it. As @rob12 stated, it’s not that simple. Let have something in the mean time. It will help to get traction. Or at least reduce frustration of actual users.
The golden solution implies:
- having revisions
- when to allow revision
- how to store revisions
- how to display revisions
- how to propagate revisions
- notifications when revised
- how comments fit in all this mess
- a lot more work to implement the possibly wrong solution
@rob12 Mike McGirvin had pointed out earlier that Friendica allows unlimited edits, and there haven’t been any problems with it. I can think of about a dozen other systems (social networks, message boards) which allow unlimited edits and I haven’t heard of any problems with them, either. They just prominently display a notice that the message was edited (often with a timestamp), so that if somebody did abuse the system like that, it would be pretty obvious. So, if there haven’t been any problems with these other systems, why would there be any with Diaspora?
To me, it makes more sense to implement editing and then see what (if any) problems crop up, and then go from there, rather than assuming a worst-case scenario up front.
1 Like
I forgot about the timestamps. That does help ward off obvious abuse. So now I feel sheepish about my example. But I’m in luck: Loomio also allows me to edit my comment at any time. So I did, and now your response to me makes less sense.
Except… Loomio also saves the full revision history. Which is good cause now we can go back and see my original comment to make sense of your response. But which is bad for the reasons you pointed out at the beginning of this thread. So if we don’t save the history then you’re left looking nonsensical… unless you also edit your response, and the next person his response… and it starts to look more like a wiki than a discussion thread.
I dunno, maybe it wouldn’t be the end of the world. Depends on the kind of users you’ve got. A bunch of software geeks can handle it no problem, other demographics might run into trouble. Or they may not. It would be interesting to see I guess.
IMHO the preview enough. Keep it as simple as possible - for the users and the infrastructure.
I think this discussion concerns far more than the simple question of editing posts. It concerns the entire federation.
diaspora* is the first attempt to to build a decentralized social network. It causes a lot of problems that don’t even exist in a client-server architecture. These problems are really important ones. We can’t deal with them just by providing workaround. Otherwise, we should just go back to a client-server social network. This workaround won’t improve the federation protocol and we all know it needs to be improved.
This is a very complex question that concerns more than only diaspora* itself. It’s an algorithmical and mathematical problem. Finding just bandages to fix these kind of problems will just throw us through the wall. The federation must be improved to deal with it’s structure, or solving further problems will be harder and harder each time until diaspora* is just stuck.
@rob12 loomio allows to edit the comments but not the proposal. I’ve seen it when I tried to change the default pending time and remove de optional SendNow.
@faldrian “The user can use preview to get this straight, this pending post thing is nothing more than a preview with a auto-submit feature.” You will always make some mistake.
“Too stressfull” So not being able to edit at all is zen ?
@macieklozinski @stefofficiel “It’s too complicated” ? Anything dealing with the federation is more complicated. Revision + federation is worst.
Pending post is:
- add a pending timestamp
- having a process that handle pending posts
- some UI
Federation + revision is:
- having revisions
- when to allow revision
- how to store revisions
- how to display revisions
- how to propagate revisions
- notifications when revised
- how comments fit in all this mess
- more UI
@alexb The proper edit is the next step if only possible. Lets enhance the user experience now with something we know is feasable in a shorter time. The user will see incremental improvement. If we want the proper edit, the user is stuck with nothing until it’s done.
@augier “postponing the problem” It is about solving the user problem with a bicycle so it can avoid walking. Then we can see how to provide a car. But in the mean time the user is not stuck with is feets.
“I think this discussion concerns far more than the simple question of editing posts. It concerns the entire federation.” This discustion from my point of view is about the user experience. You can improve it with a big step that will take a lot of time and is unsafe and if it’s not feasable ot if it fails you’re left with nothing. Or you can make a small step that improve slightly the UX and then have more time for a better solution, if it exists, since the user in less angry about no being able to edit at all.
@all
It’s not about the proposal itself it’s about improve the UX in 3 month with this proposal or anything else rather than maybe in 3 years with the “proper” edit.
Again: https://pbs.twimg.com/media/Btm2-6oIIAItR9w.jpg:large
That will be my last comment on this proposal since I can not see any other way to argument on it.
“postponing the problem” It is about solving the user problem with a bicycle so it can avoid walking. Then we can see how to provide a car. But in the mean time the user is not stuck with is feets.
Nope, postponing is about proposing the user to take him on your back and hoping that you won’t feel any backache one day…
Solving the problem is improve the federation to let it do what it should do. So yes, it’s a bit tougher, but it’s the right solution. Timestamp is a bad one.
We should not patch diaspora* to provide functionnalities while the good solution would be to make things right. Because one day we will have to deal with a spaghetti code, a lasagna of code and nobody will have the willing to maintain it anymore…
It does no good to make little bits here and, other way … It’s better to focus efforts, do something solid and the proposed update instead of doing small patch from time to time.
@augier spaghetti code comes from doing things wrong not doing the wrong thing. Doing the right thing with wrong coding will end in spaghetti anyway.
@stefofficiel my bad. I still disagree. But it doesnt’ matter the proposal is closed now.
What I usually do (might be the dumb thing) is to delete the post (usually typos or I forgot to put something or something equally stupid) and then re-post after cleaning or adding it up. I do wish there was an edit post but do understand that it would need some sort of patching of messages and could quickly be convulated for pod admins.
I have no idea when I do delete posts, are the posts eventually removed from other federated network (within the diaspora fold at least) or do they remain forever ?