Hi Mal,
Reading your post, your number one priority is, to "protect the content". And as a "side effect" you'd like to recreate "older" versions.
I think your usecase asks for a TW5 dev server and an optional VCS system. It seems you are ok to create some (probably tricky) "bash scripts" and you are in control of a server.
- TW5 comes with a "development server" that can be used in local networks.
- So it would be possible to run this one on the server and directly edit content with TW5. This option will require a network connection.
- modified tiddlers would be stored as single files.
- Version controlling could be done with the system described below.
- The file watching mechanism, that you use can be reused to detect the single tiddler changes.
- It can trigger a build process, that creates single fiel TWs and copy them to the DP sync folder.
-- Version controlling --
My "brainstorming" depends a little bit on some info, that you didn't provide.
- how many _different_ users do modify / edit the "core" content ?
- which OS does the server use?
- which OSes do the "editors" use, if they are different to the server ?
- Is some type of access management needed?
So as I know git a bit, I use it as an example. Othere VCSes may work too.
- On your server you could install git.
- This server is the "master of content".
- If the "nodejs dev server" (from above) is active, it can be used to modify the content tiddlers.
- If no dev server is runnung you can modify the tiddlers with any text editor
- Your file watch script will do the rest
If you have more than one Editor
- Editors, can have there own "local backup" with a TW5 installation, that is able to build a single file TW using a custom edition [1]
- TW5 editions basically are "directories" that contain some configuration info and many content tiddlers.
- The TW5 repo contains some scripts, (that can be optimized) to work with custom editions and build "single file TWs" [2]
- The build script could build a file TW and store it in the DP sync folder.
So what would be the advantage of such a configuration.
- Every "Editor" has a local backup on there computer. So if the server breaks, everything could be rebuilt using those local backups.
- If your IT department is responsible for the server, I hope they do regular backups of the whole stuff, so everything you need should be backuped allready.
- You can still use dropbox for syncing your single file TWs. So DP is an additional backup
- Every Editor can _offline_ modify the source and it will be possible to do "real merges".
- A TW core update could be done with a pull from the original TW5 repo.
I use TWC, and I just found http://mptw.tiddlyspot.com/#LessBackupsPlugin. I wonder if something like this, if converted to TW5 would work at all for you?
Thanks un advance
I'm really interested in this kind of implementation but I don't understand how users can edit offline their own copy.
I would like to carry my TW file in my pendrive for editing and consulting it while working and merge changes after.
I would be glad if you explain that a bit further. I have a NAS server and I think I can implement this scenario on it.Thanks un advance
Editors, can have there own "local backup" with a TW5 installation, that is able to build a single file TW using a custom edition [1]
- TW5 editions basically are "directories" that contain some configuration info and many content tiddlers.
- The TW5 repo contains some scripts, (that can be optimized) to work with custom editions and build "single file TWs" [2]
- The build script could build a file TW and store it in the DP sync folder.
So what would be the advantage of such a configuration.
- Every "Editor" has a local backup on there computer. So if the server breaks, everything could be rebuilt using those local backups.
If every editor has a local backup, why I can't have the same? Could you write a simplified scheme of this implementation?
And where can I find that server version you're talking about?
And where can I find that server version you're talking about?
Thank you for your detailed explanation.
So your idea is, mainly based in a git server/git clients and the node JS implementation of tiddlywiky. I suppose the node part is just for creating individual files for an easier git control.
I don't like using different browsers, that's why I carry node webkit and tiddlyDesktop for a consistent experience. Is node webkit able to run a node implementation of tiddlywiky?
--
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+...@googlegroups.com.
To post to this group, send email to tiddl...@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
For more options, visit https://groups.google.com/groups/opt_out.
Jeremy, that approach sounds good, but I would prefer an automatic version. Is there any way to detect new Tiddlers within a single file implementation? Maybe I could try send new Tiddlers to my own server and merge then.
Sync back to the web will be difficult because limitations or security reasons? As far as I know, node webkit has the same privileges as a native application.
Jeremy, that approach sounds good, but I would prefer an automatic version.
Is there any way to detect new Tiddlers within a single file implementation? Maybe I could try send new Tiddlers to my own server and merge then.
--
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+...@googlegroups.com.
To post to this group, send email to tiddl...@googlegroups.com.
Visit this group at http://groups.google.com/group/tiddlywiki.
For more options, visit https://groups.google.com/groups/opt_out.
The discussion has touched on both node-webkit and ordinary browsers. The restrictions apply to browsers, code in node-webkit can indeed freely talk back to a server.Having said that, TiddlyDesktop currently runs TiddlyWiki files in a sandbox with much the same restrictions as a browser. This is done for the same reason that browsers do it: it would be undesirable for a buggy or malicious TiddlyWiki document to be let loose on your computer. Anyhow, we can arrange for privileged operations to be carried out via the same bridge that TiddlyFox uses.
We'd all prefer something more convenient, the point is that browsers explicitly disallow the scenarios that would be most useful.Is there any way to detect new Tiddlers within a single file implementation? Maybe I could try send new Tiddlers to my own server and merge then.
The issue is that browsers have shut down the ways that an HTML file loaded from a file:// URI can talk back to a server. That's why we have to workaround by initiating the sync from within the server, or use node-webkit.Best wishesJeremy
That's the point. Do you have any pre-implementation or workaround using node webkit? I will be glad to try it out.
A small offtopic : is there any easy way I can try the new feature from 5.0.7 for syntax highlighting in the current one? I'm not very familiar with Git.