On Sunday, July 17, 2016 at 1:13:24 AM UTC+2, Dragon Cotterill wrote:
>
> [...] For example how do we handle the importing of Tiddlers with the same 
> Title?
> I feel that it is very disingenuous to force users to create shared 
> Tiddlers with different names. At the moment we are doing this by adding 
> the creators name to the end of the Tiddler's Title. This I feel is a big 
> mistake. The naming of the Tiddlers should NOT be enforced like this. The 
> creator should be free to name them however they want. But of course this 
> throws a spanner in the works when names clash.
>

While I do also think universally unique IDs are a good idea, I think there 
are some misunderstandings (hopefully not on my side, but it may be!)

First, just to be clear; the user isn't forced to do anything manually - 
the titeling is created in the fetching-of-the-tiddler process. Second, and 
again just to prevent misunderstandings; it is not "the creators name" that 
is added but the unique identifier used in TWederation, namely the URL for 
the wiki. Third, there is also an addition of a time stamp as part of the 
title. So the tiddle becomes named something like

original title - URL - timestamp

(I might have the order wrong)

Now, it is in deed simple to *present* this name as only the original 
title. One way is the caption field as Eric mentions and another way might 
be my aliases <http://aliases.tiddlyspot.com/> concept (it works but there 
are some minor tidbits I must straighten out.)

As I said though, unique hash key for each tiddler does seem sensible to 
me, but I have very limited experience in databases so I can't quite argue 
on the matter. In the pub discussion we had, at least Mario and I, there 
was however another point in this:

To use a tiddler ID as a way to track its origin. A kind of DNA as Mario 
put it. So that in (eg) an identifying character sequence, a few of the 
characters remain from the origin but other characters concern the current 
version of that tiddler. Mario will have to explain the details but it is a 
fascinating idea and it would perhaps enable us to create overviews of how 
tiddlers (e.g plugins) evolve. I think it would also be a way to "branch" 
and be able to automatically credit original authors better.


Tiddlers change. At present if a federated tiddler is changed, that change 
> is not propagated (or have misunderstood the workings of TWederation? 
> Mat?). If a parent tiddler is deleted, what becomes of the orphans? Yes, 
> there is still much to discuss on this.
>

You're right; no "propagation" since the whole idea centers around 
*fetching*, i.e copying a remote tiddler into your own TW but renaming the 
copy in the process. So there is no "propagation" unless you fetch *again*.

If a parent is deleted in your TW... well, then you have loose orphans just 
like if you have a tiddler tagged with some other tiddler title (or 
be-fielded with some other tiddler title) and delete that "parent". It is 
up to you what you do with the orphans, they're just normal tiddlers. We 
could easily create a batch-delete macro to delete a complete "discussion" 
(side note: "discussions" are just one use for TWederation. The general 
case is, I guess, simply "related tiddlers").

@Jed, @Jeremy - A serious problem if the ID is based on the title is; if I 
"fetch again"... but the tiddler has been renamed since last time... then 
what? There is no way to tell the renamed tiddler has anything to do with 
the old title. Or am I misunderstanding something?


<:-)

-- 
You received this message because you are subscribed to the Google Groups 
"TiddlyWiki" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to tiddlywiki+unsubscr...@googlegroups.com.
To post to this group, send email to tiddlywiki@googlegroups.com.
Visit this group at https://groups.google.com/group/tiddlywiki.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/tiddlywiki/6eab4caa-dec9-49b2-8264-f0e0fd572570%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to