selecting non-contiguous text

[nicka]

You're asking to add new features that speak to your specific needs rather than asking the developer to improve his product for a potentially larger market.

Not really, or at least, that's not the intention. The suggestion that Mellel urgently needs to add features is not so much because existing users want them (although we do, of course) but because that way Mellel would reach a market of technical writers, academics and others who want to do (potentially) long document production with good typesetting and light page layout without resorting to LaTeX. That market certainly exists but is not served by any OS X application.

Another niche that Mellel has only partly reached is the niche for a really capable multilingual word processor. Again, there is no such thing on OS X.

[staypuffinpc]

I just don't see the burning need.

We'll just have to agree that we have entirely different workflows. If I am writing, I'll get into the flow of things and clean up afterward. Also, your scenario assumes that you're the one creating the document. In many cases, I inherit a document that I would then like to add styles to.

Your other suggestions — GUI improvements and consistency; integrated help — strike me as much more important. But I would prefer the functionality that I need — smart handling of captions; figure styles; dynamic TOC — introduced before UI niceties.

While I agree that there are many features that would really help the program, "niceties" stop the whole thing from moving forward. Simply adding new features w/o improving on the experience of what's already there leads to feature bloat, something many on here argue against. All I'm asking is that, before moving on to add more features, why not perfect the implementation of those that already exist?

BTW, all those things you mention are things that I'd like. In fact, I needed a dynamic TOC feature just yesterday, as I wanted specific TOCs for each section of the document. To me, though, that's more an issue of refining what's already there rather than adding a new feature (e.g., adding a "start at" and "end at" in a TOC insert dialog).

[rickl]

After all, nobody would doubt the validity of concentrating on niche-markets as a business-model in food production anymore, so why would not the same go for application development?

I couldn't agree more. The idea of appealing to a broad market usually means producing something that isn't really satisfying to anyone, arguably what Word has done. I think replacing Framemaker is a noble and extremely worthwhile goal.