I'm grateful to everybody for taking part in this discussion. It's very helpful to bring things to the surface, and I'm tremendously appreciative that there is interest in my activities.
The original question was about splitting this group, and the rationale included some comments about the relationship between TWc and TW5 that are worth responding to in some detail before turning to the original question.
The first thing to note is that TiddlyWiki classic has been slowly dying since BT acquired Osmosoft in 2007. Here's the Google Trends graph of search volumes, usually a pretty good metric of interest:
That's the time frame that Osmosoft invested significant effort in the TiddlyWiki universe. It has enjoyed the attention of a devoted but dwindling core of end users, but it hasn't been enough to stem the decline.
Against that background of dwindling interest, one might ask why I'm investing my own time and energy of working on a new version at all. Fundamentally, the reason is that I think that the success of TW classic thoroughly validates the idea of a single file wiki but that the execution was too flawed to sustain the attention it deserves - particularly the attention of developers.
I have personal experience of this issue while I was running Osmosoft inside BT: there was little enthusiasm from many of the developers to work on the TiddlyWiki core code. Coming from a conventional development background, TWc is a very strange beast. I wrote most of it while I was learning JavaScript and HTML, and at a time when TiddlyWiki was blazing a trail, pushing browsers beyond their commonly understood constraints. There was no jQuery and so a lot of my effort on TWc was just to get the code working consistently across a range of browsers.
At Osmosoft we successfully developed TiddlyWeb, a clean, well engineered, production quality serverside for TiddlyWiki. But we struggled to get the TiddlyWiki side doing everything that we wanted. We were working against assumptions made in the core code, and in the flock of popular plugins. Over the years I think Osmosoft did a pretty good job on the suite, but even today there are some troubling limitations - for example, TiddlyWiki under TiddlySpace can't dynamically reflect the changes made by other users to the same space.
Other technical limitations of TiddlyWiki have dogged it from the start: for example, there is still not a good experience for search engines to index TiddlyWiki. Not everybody cares about such capabilities, but those that do have little choice but to abandon TiddlyWiki.
So, for a while, TiddlyWiki has been an unfriendly environment for developers, but with lots of limitations that really need developer effort to overcome.
Turning now to end users, with open source software a good rule of thumb is that the more users a piece of software has, the more useful it becomes. Accordingly, I believe that the best way to assure TiddlyWiki's future is to try to make it much more popular than before. I bandy around the figure of making it 100 times as popular.
I believe that we can make it that popular by making it easy to learn and easy to use, and ensuring that the underpinnings are flexible enough to cope with whatever the future throws at us. As I said before, the basic proposition of TiddlyWiki has always seemed to go over quite well with prospective users. This time around I think we can do a much better job of explaining it to new people.
If we achieve this increase in popularity, then the community of incoming TW5 users will rapidly outnumber our core community of TWc users. Therefore, I believe that we should optimise the experience of discovering TiddlyWiki for new TW5 users, and we need to do that now.
In my opinion, TiddlyWiki5 and TiddlyWiki Classic are different versions of exactly the same product, with the same goals and the same basic design. There are incompatibilities in the way that they are customised and extended, but the elevator pitch is precisely the same: a JavaScript wiki that works from a single HTML file. TW5
The changes in TW5 are resolutely intended to benefit end users. For example, the new plugin architecture makes it possible to drag-and-drop plugins between wikis, not just cut and paste them. And now a plugin can contain a bundle of related tiddlers (which is handy for developers) while still being a single unit for end users.
As I've shown above, underpinning those improvements with a good developer experience is important for end users because without a stream of developers interested in working on TiddlyWiki and it's plugins we wouldn't have the rich world of extensions and adaptations that makes TiddlyWiki so useful in so many different niches.
There a few misunderstandings in the thread that suggest that I need to do a better job of explaining TW5. For instance, the idea that the presence of the node.js edition of TW5 might somehow compromise/diminish/complicate the single file edition. The truth is that TiddlyWiki has since 2006/7 had a complete set of command line tools to enable the developers to work on the core. They were kind of klunky weird things written in Ruby, with a high learning curve. All that's happened in TW5 is that that job of building a single file wiki from it's constituent parts is undertaken by TiddlyWiki itself. This makes the tiddlyverse easier to use and easier to understand.
Finally, turning to the original question about splitting the Google groups, I have a couple of observations:
* Experience suggests that people find it hard to figure out which Tiddly* group they should address with a particular post. There are already too many groups
* I don't think we can rename the existing group without messing up links to it
* Based on past behaviour, Google is probably going to kill Google Groups as soon as they decently can
So, my response to the original suggestion is:
1) We should add a welcome message to the TW groups asking people to flag their subject lines TWc or TW5 (I've actually already done this)
2) We plan a migration away from Google Groups:
a) TiddlyWikiDev could migrate to StackOverflow; several other open source projects use it in that way
b) The main TiddlyWiki group could migrate to a homegrown TiddlyWeb+TiddlyWiki5 host
I haven't had time to address all the points raised in the thread, do please fire away with any questions or clarifications.
Best wishes
Jeremy