Comment on the TW

[Joe Armstrong]

I'd like to help improve the quality of the information

at https://tiddlywiki.com/

Conversations/questions/comments about tiddlers

are taking place at the wrong place.

Discussion about  tiddlers should take place *at the tiddler* 

as a comment appended to the tiddler 

and not in this mailing list.

I feel like shouting this from the rooftops!

Would this work?

1) Comments are submitted by email by submitting

through a standard form. 

This should have a load of fields

TiddlerName, permission to publish etc.

(this is for, copyright, permission, GDPR rights etc.)

2) Comments are reviewed by a moderator once a day (by volunteers

- this is to avoid spam etc)

3) A "daily version" of the TW is made

(could be higher frequency if necessary)

4) Comments are cross-posted to this list

This is so people can stumble over the discussion when they search.

(It has been previously observed that TW's themselves are badly indexed - the solution is to cross post to (say) this list and make sure this list is well indexed :-)

One of the bad things about the web is that

comments about X - should take place at X and not Y where X and Y 

are different locations.

This mechanism could eventually be built into the node server to improve the feedback mechanism to the author of the TW.

Cheers

/Joe

[TonyM]

Joe,

I agree with this idea of comments in close proximity to the actual issue. I am personaly trying to build a mechanisium for this but others may be better qualified. This is my approach;

Allow tiddlywiki to save in the browser session like noteself allows. Allow the fully updated wiki to be saved to disk but also allow comments, notes edits and new tiddlers to be exported to a json file or idealy emailed as an attachment from the browser, or submitted to file queue. Then provide an import tool which allows a review and apply changes during the import process.

Such a system would assist in rapid feedback and enhancement. Allowing lessons learned and observations made by people to be captured in real time.

Tony

[Jed Carty]

I made a wiki that used hashover to let people leave comments but I don't think it is available anymore. It used hashover to save the comments using a php server and then in the wiki a widget would query the server for new comments and make tiddlers out of them and display them as a threaded conversation. That approach was a bit unwieldy, but it is possible.

I am hoping to use Bob to make a much more streamlined commenting system, it has enough control over permissions that it could allow comments without a login but still prevent any other changes to the wiki. 

As a note, any commenting system should almost certainly only save the rendered html of the comments to prevent malicious wikitext or javascript from being embedded in the wiki.

[Jed Carty]

As a related idea, I have been looking into running a Zulip server. It is like an open source Discord but not run using a business model based around spying on people using it and you don't have to pay for features. One of the nice features is you can make very flexible bots using python or node, so you could have conversations on it and have the bot send interesting parts to the list here or include them in a wiki.

[Joe Armstrong]

There is an easier way of saying this:

    I'd like to send mail to a tiddler

Tiddlers don't have email addresses so you can't send emails to them.

It down be *very nice* if all tiddlers had a "mailto" link in the dropdown region of the tiddler.

Clicking on this should allow you to compose and send an email to the tiddler.

Well not really - it sends it to an address of some person or program that will do something with

the mail. What happens when you get such a mail is not part of the TW - it's an out-of-band problem.

Seems to me like entire TW's should have an associated email address.

There is a slight problem here, who gets the mail if a tiddler is modified.

 I think the original author of the tiddler should get the mail

 If the tiddler has been modified the person who modified the tiddler gets the mail

 and might possibly forward it to the original author.

Processing such mails would open a range of possibilities.

/Joe

[SylvainComte]

HelloThere

I don't know if it's relevant here but if the projects is still moving forward to gitlab, we may use the send issue by email functionality. The dropdown menu could create the email on the fly using retain-original-tiddler-path so that the object of the email contains the title of the tiddler.

cheers,

Sylvain
@sycom

[Jed Carty]

I think that extending any of the existent node back-ends (the build-in server, Bob or tiddlyserver) to send an email when a comment is made wouldn't be a terribly difficult task. In my rather limited experience the problem is that most hosting providers block port 25 (or whichever port is used to send email). I think that it would be simple enough to have some convention for putting the tiddler name as the email subject or something similar and probably some date or time information would be enough to uniquely identify a tiddler and allow the recipient to set up email filters.

[Jed Carty]

Sorry, I wasn't thinking there. If it was a mailto link than the back-end wouldn't need to be involved at all. So it would just be a matter of having a macro that creates the email template for comments and then whatever filtering past can be done in the email client.

[Mark S.]

If you have an email client set up. Which a lot of people don't these days thanks to Gmail, work requirements, mobile machines, etc.

For a back-end server, couldn't comments actually be saved to the server? Not an exploitable public place, of course. But a log somewhere that developers could pick up daily and screen.

The overall problem is, that one-way communications tend to generate low-quality feedback.

-- Mark

[@TiddlyTweeter]

Interesting discussion. But I'm having difficulty with it. 

Speaking from what I know ... authoring from single TW via email, or posting to Twitter etc is fairly straightforward. In these cases I think of my use as an originating author. 

The fact I could expose such mechanisms to visitors would make them originating authors too (even if the thing would not save on line the Tweet or email would still go out).

Maybe I'm missing something. But where is the interaction?

What Jed is developing with Bob looks like it could eventually accommodate well structured multi-party discussion. But I'd imagine it would be discussion of the themes presented on the specific server instance. Would a TW instance actually be able to do better than GG?

The missing piece here for me is I don't yet get, Joe, what gap you want to bridge?

As one of the most vocal critics of the Google Groups swamp I'd guess we might in a TW make a meta index to it? The issue is that such groups have a lot of needed, yet ultimately ephemeral, content. Who will do the index?

Just thoughts

Josiah

[Mark S.]

On Monday, January 14, 2019 at 8:33:32 AM UTC-8, @TiddlyTweeter wrote:

Maybe I'm missing something. But where is the interaction?

Yes. Exactly. One-way communication is going to get a lot of random, non-useful responses.

-- Mark