I agree that enabling indented soft wrapping is more difficult than it should be. In other text editors (such as Xcode) it is either on by default or just a single command away, while TextMate requires setting up a custom bundle and defining a regular expression. I would like to work on a patch to improve this situation, but I’m not even sure how it should be architected. For example, if a default indentedSoftWrap setting is added to the Text and Source bundles, how can it be turned on and off? Is there some existing setting that works like this? Any guidance from the core TextMate devs would be appreciated, thanks! Trevor On Jan 29, 2014, at 2:52 AM, Tom Wardrop <tom(a)tomwardrop.com> wrote:

On 31 Jan 2014, at 4:00 pm, Allan Odgaard <mailinglist(a)textmate.org> wrote: FWIW, here's an idea: Would it be possible to expose an API for bundles to provide their own "preference pane", a la System Preferences? I'm imagining an interface something like Growl's Applications tab—a page in TextMate's Preferences window with a list of bundles in the sidebar, and an area on the right with settings for that bundle. It wouldn't enable anything you can't already do with the Bundle Editor, but it would be much more *discoverable* for those who don't want to be required to hack on their bundles. –Adam

On 1 Feb 2014, at 8:50, Matt Neuburg wrote: The current Bundles → Select Bundle Item… (⌃⌘T) window is a query functionality that I would like to extend to allow querying settings and key bindings outside bundles. Though it’s a challenge to get all the options squeezed into the window and still keep it simple.

On 3 Feb 2014, at 0:49, Kyle Hanson wrote: The issue is that “soft wrap” is not a global boolean choice. For example I like comments to be soft wrapped in source code, where soft wrap is otherwise disabled. We also soft wrap list items in Markdown documents, etc., which is why these settings are in bundles, since they are sort of language specific. That said, how to disable soft wrapped comments is one of the most frequently asked questions, so it’s likely a bad default that I’m considering changing.

On Feb 2, 2014, at 6:41 PM, Philippe Huibonhoa <phuibonhoa(a)gmail.com> wrote: There are two problems I'm having: * A setting that causes a phenomenon I'm seeing could come from many different places (including a theme or a setting, as well as a language grammar) in a bundle different from the one I think I'm in. * And then to confuse matters *still more* there are all these global settings, plus the .tm_properties files. If we did what Philippe suggests, it would be *even more* confusing than it is already: "I'm saying to wrap so why aren't we wrapping?" One more undiscoverable thing off in one more undiscoverable location (the view menu or whatever). It makes me want to argue for exactly the opposite of what Philippe is saying: there should be NO global settings, NO view menu settings. Everything should be in files that I can actually find, in order to govern and understand what's going on. But at the same time there needs to be a way, as I suggested earlier, to ask TM2 to *tell* me what's going on - for any given phenomenon, *what* file is causing this behavior / phenomenon, including how it overshadows any other files that are trying to influence the same behavior. It seems to me that this can't really be that hard to do, because TM2 must *have* a knowledge of this. It has a logic that *causes* any given aspect of what I'm experiencing. So it should surely be able to *tell* me that logic. m. -- matt neuburg, phd = matt(a)tidbits.com, http://www.apeth.net/matt/ pantes anthropoi tou eidenai oregontai phusei Programming iOS 7! http://shop.oreilly.com/product/0636920031017.do iOS 7 Fundamentals! http://shop.oreilly.com/product/0636920032465.do RubyFrontier! http://www.apeth.com/RubyFrontierDocs/default.html TidBITS, Mac news and reviews since 1990, http://www.tidbits.com

This a good discussion, but getting back to the original question, what is the current “best practice” for an end user wanting to enable intended soft wrap globally? I remember asking this question years ago, and the response was to create my own personal bundle (e.g., “Trevor’s Bundle”) and add to it a setting like the following: { indentedSoftWrap = { match = '\A[ \t]*'; format = '$0\t'; }; } I certainly agree this procedure is not very user-friendly, but is it still the way to go? Or have things changed in TextMate 2? Thanks, Trevor On Feb 2, 2014, at 7:21 PM, Philippe Huibonhoa <phuibonhoa(a)gmail.com> wrote:

Hmm. I disagree. User-friendliness is not, as many self-proclaimed power users seem to think, a newbie feature. It is a matching of the control over the capabilities to the user’s workflow, no matter the level of user and/or complexity of workflow. Saying “we don’t need user-friendliness” is throwing the baby out with the bath water, and belies a lack of user experience design, IMO. For any workflow, even, no, *especially* power user workflows, there is a sweet spot of providing control over that workflow in ways that make that control simple, elegant, and effective. Power user != just throw me into the guts of a system. If it were that, the only true power users would be flipping bits directly. Now, you can say that user-friendliness is not a strongly required feature for a particular application or system. That’s a systems design choice. But to say that power users don’t need user-friendliness? Naw, that’s just taking the easy way out. UX design is hard, absolutely, but to say “oh, we don’t need that, we’re *power users*” quickly leads to the preference panels in many Linux apps that look like the dashboard of a 747. Everything-is-an-option is chaos. Instead, TM2 should be looking at ways of making the common tasks of power users simpler and more direct. I think it does a remarkable job of this as it is. At the risk of contradicting myself above, perhaps a “Global preferences take precedence” switch. Turn it on, the pref you set at the global level trumps all. Turn it off, and the pref you set at the global level is the default, to be overridden by bundles. I can easily see scenarios where either is rational and reasonable, and neither jumps out as a clear winner. Therefore, a power user workflow decision is promoted to a discoverable, clear checkbox. Voila. User friendly for the power user. And just as a datapoint? I finally ran screaming from emacs back in the great 19.27/28 schism of ‘94, and haven’t looked back. Got tired of dealing with the ongoing breakage of basic power user scripts. That may not be the best model to aspire to, from a user perspective. :) I take the description of TM2 as “emacs for Macs” to mean that it has the Mac usability baked in, and has the same level of attention to detail and design as other Mac apps. Otherwise, it’s just “emacs with Python and Ruby” and I don’t think anyone wants that. We can do *way* better than that, and I think Allan has. On Feb 2, 2014, at 8:27 PM, Matt Neuburg <matt(a)tidbits.com> wrote:

On 3 Feb 2014, at 12:52, Tom Wardrop wrote: I think for soft wrap, one could split it into “soft wrap” and “semantic soft wrap”. The latter would turn all the bundle soft wrap settings on/off, but even that is not really perfect, as one user may dislike the wrapped comments, but would still want list items to be wrapped in Markdown. The solution is probably just to disable the default wrapping of comments, as this is what people generally want to disable. Matt was actually talking about “behavior”, which is a little bit more abstract. Like “why does it indent after a open brace?”. But as I previously replied, the current Select Bundle Item… tool can be extended to allow querying more things, which would go a long way.