That's my point though -- I don't care about the file structure. It just happens to more-or-less resemble the class division in Java, but even that's not always helpful if you adopt some functional style with all the anonymous inner classes into your code. So what really counts is just what the bubbles approach lets you do (in a more obvious manner, as you noted) -- define or discover meaningful selections of code and use them as first-class citizens of the development process (persist them, share them, attach meta-data to them, etc.).. Contrast that with the current crop of IDE's that do let you discover bits of code that are connected, but it's tough to isolate, say a method from one class with a method from another class with some instance variable, etc. etc. -- all in one "unit". What's ironic, is that while a typical IDE will let you quickly pop up a bunch of tabs, each jumping to the relevant line, if you want to see 2 snippets from one file together, it's tough to do if they're not close together, if I'm not mistaken (I'm not entirely up to speed on the IDE state of the art).
________________________________ From: Peter Becker <[email protected]> To: [email protected] Sent: Thu, March 11, 2010 3:43:41 PM Subject: Re: [The Java Posse] Re: Code Bubbles: A really weird new IDE. (Posse: Interview this guy!) The current generation of IDEs already supports this navigation approach, just not the visualization. I hardly ever go through the file hierachy to find a file to open, I use the shortcuts to open types or resources, I use the shortcut to go into a method that's called, I find all callers via another shortcut or hit a key to see the hierarchy for an object. And then there is the object browsing, hot code replacement and all the other cool stuff in the debugger. If you are still using vi/emacs/whatever you should probably go back and check out the keyboard reference chart of a proper IDE. I only recently converted a hard-core vi user to Eclipse (or at least an Eclipse/vi combo) -- it is pretty easy to see an IDE just as a glorified text editor, in which case sticking with vi makes sense if you already know it. But an IDE is much more than that, but it is not all that obvious. The code bubbles just make it obvious, but that by itself is a major achievement. Peter On 12/03/10 02:46, Alexey Zinger wrote: > >I >gotta say, this is the first IDE I want to use. Yes, I know, I haven't >actually tried it, so it's a bit premature for such judgments, but I >can't help but feel enthralled. I'm a little concerned about what >it'll feel like to work on longer methods, where vertical scrolling is >necessary. I guess in that situation your bubble takes up as much >vertical real estate as you can give it and then if the user would >probably start moving its sibling bubbles to the side to give it >maximum height. I guess it could work... > >>As far as a paradigm shift, I don't see it. Both in IDE's and in plain >text editors, it seems most people spend a lot of time referencing API >docs, other snippets of relevant code, often looking at the same set of >snippets of code for any given problem, etc. I think this thing >matches the current paradigm perfect, but with a better execution than >what's on the market presently. > > > > ________________________________ From: >Brian Ehmann <[email protected]> >To: The Java Posse ><[email protected]> >Sent: Thu, March 11, >2010 11:31:35 AM >Subject: [The Java >Posse] Re: Code Bubbles: A really weird new IDE. (Posse: Interview this >guy!) > >>I really like its approach. The UI appears to intuitively allow a >>developer to narrow their focus to the parts of the code that are >>absolutely necessary in order to accomplish a given task. Also, the >>ability to send a serialized copy of a given section of my workspace >>to another developer is another plus. Its strikes me as the next >>evolutionary step from pastbin since you can build out the context of >>the message you are trying to get across with the appropriate code >>fragments, notes, and flags. > >>Unfortunately, going from seeing entire source files in a project tree >>to functions in a bubble is such a radical change that I wonder if it >>will be difficult for developers to adapt their practices in order to >>take advantage of the features of Code Bubbles. > >>- Brian > > >>On Mar 11, 6:00 am, Johannes Thönes <[email protected]> >>wrote: >>> Yes I agreed. It is a very interesting approach. And I would love >to >>> hear an interview about this. >>> >>> On Thu, Mar 11, 2010 at 11:55 AM, Reinier Zwitserloot >>> >>> >>> >>> >>> >>> <[email protected]> wrote: >>> > I stumbled on this video of Code Bubbles in action: >> >http://www.youtube.com/watch?v=PsPX0nElJ0k >>> >>> > and was quite amazed. I'm not sure if its pragmatic to have >so little >>> > signal-to-noise when actually typing new code, but there's >easy >>> > solutions around this. Basically, your IDE is an effectively >boundless >>> > plane and the granularity of editing anything isn't per-file >but per- >>> > method/class, with the method/classes actual location just >metadata, >>> > with the IDE sorting it all into the appropriate files >internally. >>> > Navigating anywhere is primarily via a search-in-everything >keyboard >>> > box, and code appears in bubbles on this infinite plane. If >you do >>> > things like "visit declaration", the declaration opens, but >in a new >>> > bubble, visible together with the original code, instead of >what most >>> > current IDEs do, which is to open a new 'tab' and replace the >view >>> > entirely. It would also be an _amazing_ pair programming / >code review >>> > IDE if only you could all work in one plane, each having >their own >>> > little section in it, with you able to freely travel to >someone else's >>> > space. Unfortunately from the video it seems like all you can >do is >>> > email workspace layouts around, but that too could be >addressed, I >>> > guess. >>> >>> > Later on in the video a debug session is started which really >does >>> > look amazing (for code editing I'm not entirely convinced >yet, but >>> > that looks like a fantastic debugger!) >>> >>> > I know discussions about "Why are code editors still a >glorified dumb >>> > terminal" show up from time to time and this is certainly >something >>> > new. >>> >>> > There isn't a download yet; more info is here: >> >http://www.cs.brown.edu/people/acb/codebubbles_site.htm >>> >>> > I wouldn't mind seeing an interview of Andrew Bragdon about >this :) >>> >>> > -- >>> > You received this message because you are subscribed to the >Google Groups "The Java Posse" group. >>> > To post to this group, send email to [email protected]. >>> > To unsubscribe from this group, send email to >>> > [email protected]. >> > For more options, visit this group >> > athttp://groups.google.com/group/javaposse?hl=en. >>> >>> -- >>> Johannes Thönes >>> johannes.thoenes[at]googlemail.com > >>-- >>You received this message because you are subscribed to the Google >Groups "The Java Posse" group. >>To post to this group, send email to [email protected]. >>To unsubscribe from this group, send email to >>[email protected]. >>For more options, visit this group at >>http://groups.google.com/group/javaposse?hl=en. > > >>-- >>You received this message because you are subscribed to the Google >Groups "The Java Posse" group. >>To post to this group, send email to [email protected]. >>To unsubscribe from this group, send email to >[email protected]. >>For more options, visit this group at >http://groups.google.com/group/javaposse?hl=en. > -- You received this message because you are subscribed to the Google Groups "The Java Posse" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/javaposse?hl=en. -- You received this message because you are subscribed to the Google Groups "The Java Posse" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/javaposse?hl=en.
