Flutterby™! : Flexible software 2010-05-13 23:38:46.569057+00

Flexible software

2010-05-13 23:38:46.569057+00 by Dan Lyke 8 comments

Oh. My. Deity. This is precious in an "aaaigh!" sort of way: Screenshots of a Windows bulk rename program. The things people will do to avoid learning regular expressions and a little bit of code...

Eric, (Laughing) thats extreme, even for me. Most of my interfaces are a lot cleaner than they used to be. I have to admit I said "that looks like one of mine" when I saw that. My last big hurdle is there are too many ways to do the same thing in much of my systems. Powerful, but confusing to some end users.

We have picked a tabbed design paradigm, putting few options under each tab. but i still see people click on all the tabs looking for stuff. at least you know where the options are when they are all on one screen.

I think the answer you're giving me is that the "point and click" interface is what people want, even though it leads to monstrosities like this.

Halfway. The answer I'm giving you is that point-and-click is what people want, but it does *not* have to lead to monstrosities like this.

(I am married to someone who, as a large part of what she does for a living, studies how people actually want to use a website or application, and then recommends design based on it, and her timeless rant is how so few clients seem to see the importance of this sort of usability study, or of clean interface design. So do forgive me if some of it seeps through occasionally!)

I've been reading Everything is Miscellaneous from David Weinberger. When I think about how I organize my information in a computer nowadays, filenames are no longer a significant factor. Folders are only loosely. I keep my pictures in a folder named "pictures" and my music in a folder named "MP3". I use Picasa to find pictures - never the filename. Picasa indexes my pictures by data, tags and even faces. Same with music - MediaMonkey tamed my ID3 tags to the point where I can actually navigate my music collection reasonably.

I keep everything related to any content I create in one big "Documents" folder. That folder structure has gotten so complex that it might as well all be in one folder. I continue using sub-folders just to keep the list of things that show up when I need to insert an figure into a document or presentation. I've become more and more dependent on Google Desktop Search.

The point? I don't need awk or regex to deal with system-level objects (like files). And I'd hate to go back to when I did need them...

I think the answer you're giving me is that the "point and ~~grunt~~ click" interface is what people want, even though it leads to monstrosities like this.

Eh, I dunno that I count that among the evils of Microsoft. The cygwin tools do not act in a particularly Windows-like way (or at least didn't the last time I looked at them, which was several years back). I seem to remember that I installed them, they annoyed me, and I uninstalled them. And if they did bundle them or something similar with the OS, only a tiny handful of users would bother to find them or use them.

Windows is never going to be *nix. The question is, why should Windows bother to make itself more useful to the people who like *nix? Those people will just use *nix. And if they're forced to use Windows, they generally have the know-how to find, use, or even create the productivity tools they need.

dexev, sorry about the linebreaks, the history of this system involves pulling messages from email where linebreaks were removed and
was used when that was really meant.

Todd, it is one of the evils of Microsoft that Windows systems don't ship with Cygnus tools by default.

#Comment Re: made: 2010-05-14 17:30:18.707948+00 by: dexev [edit history]

An example in straight sh. find, sed, and grep are also standard to to POSIX, AFAIK.

j=1 for i in $(ls /mnt/camera/IMG[0-9][0-9][0-9][0-9].jpg); do new_name=$(printf "trip_to_spain_%04d".jpg" $j); let j=j+1; mv $i $new_name; done

But you know - putting aside the fact that this program was obviously designed by someone whose primary imperative was that the code never use more than one interface screen - normal users cannot casually sling regular expressions around in Windows *even if they know how to use them*.

Where is the Windows equivalent of awk or sed? You are certainly not going to get it from a DOS-style command line. Getting a useful copy of Perl running in Windows is so annoying that I have simply stopped doing it. This leaves people who want to do any reasonably common renaming operation in Windows up a tree.

Fortunately, the files I most often want to bulk rename are image files, and my photo editor/cataloguer software has a fairly sophisticated bulk-rename interface. And there are a few decent pieces of software devoted to the task. But it's always been something that you need SOME kind of external software to do.

Even in *nix, the kind of regexp processing you're talking about does not come standard to the OS - you've just been on so many systems that had a regexp tool of some kind already installed that you're used to it being there.

Frankly, given what I have seen of the habits of normal users, I think I'd almost rather teach them to use a horrible window like that than try to teach them to use raw regular expressions. However, most tools for this are not that horrible.