May 14, 2006

Get lean. Stay hungry.


“The old Jetta was trim and compact, with chunky good proportions. The new one — 5.7 inches longer — is so big and amorphous they should have called it Jetta the Hutt. Every manufacturer engages in this incremental generation-to-generation size creep, and if it keeps up, eventually Shriners will drive 1996 Buick Roadmasters and we’ll laugh at their comical little cars from the observation decks of our Subaru Imprezas. Somebody, stop the madness.”
The NY Times auto section*

The same could be said about a lot of modern software, of course, and a decent backlash is underway. Throw a dead cat & you’ll hit some manifesto or other talking about how features don’t matter, shouldn’t be added, etc.**
Why is that? A few things come to mind:

  • Packing in tons of features makes software take forever to load, and/or makes it run slowly and consume tons of resources. Therefore everyone is penalized by stuff they’ll never use.
  • The existence of unused features makes it harder to get at the small percentage you actually care about. Locating the right command is like finding a needle in a stack of needles.
  • Being presented with a wall of options (especially if the previous set wasn’t well understood) makes people feel inadequate. The percentage anyone comprehends grows smaller as the app grows more vast.
  • New features give the impression of a neverending, ever longer learning curve. Rather than make things simpler, they risk adding confusion and redundancy, fatiguing the people they’re supposed to help.

If this is all true, then aren’t the critics right? Yes–if it’s true. But what if it weren’t? What if:

  • New software booted up faster than its predecessor on the same machine?
  • It ran faster, felt smoother, and produced better results, without requiring any additional learning from users?
  • The interface could grow simpler, more focused, more relevant to your needs (and your needs only)?

In short, if you could take away the pain that comes with a large and growing feature set, yet keep its benefits, would it cool the critics out? Would we then have permission (or blessings, even) to add whole new levels of power and capability?
As you might guess, we’re thinking about these issues all the time. In my view we need to define a fairly rigorous “Contract with the Customer” to ensure that before we move on to adding new layers of richness, we do the hard work of addressing the problems mentioned above.
We need your permission to take Photoshop in new directions, to add features that will blow people’s heads clean off. And to earn that permission, we need to show that we’re nailing the fundamentals. It’s not going to be an overnight thing, but I think we’re on the right track.
* VW can always take solace in having possibly the coolest parking structure ever. Oh, and once again, a fistful of great ads.
** To me, though, these critiques ring a little hollow–not unlike the great Onion article, “Report: 98 Percent Of U.S. Commuters Favor Public Transportation For Others.” That is, to some extent the critics are saying, Don’t add anything for anyone except me, that I personally don’t need right now.

Posted by John Nack at 10:40 PM on May 14, 2006


  • Mike Perry — 12:21 AM on May 15, 2006

    Go for it! At some point, Adobe should bite the bullet and clean up an interface that grows more confusing and inconsistent with each new version.
    I’m driven nuts by InDesign’s inconsistent menu on/off items–some add check marks when something is enabled (good) and some change what they say (bad). Make it one way.
    And don’t get me on that Story palette with just one item. That’s at least an easy fix. Optical Margin Alignment should be a paragraph attribute rather than a story attribute. Then it’d be on the paragraph palette where it belongs.
    Menus, control bars, tools with little menus, and palettes with long menus, not to mention all the secret commands known only to those with secret decoder rings–if present trends continue, CS 6 will be impossible for anyone without a genius-level geeky IQ.
    –Mike Perry, Inkling Books, Seattle

  • Martin — 3:29 AM on May 15, 2006

    Being a long time Fireworks user, as far as Photoshop goes, I would like to see you rip Imageready out of it and throw its decaying corpse into the dustbin. That would certainly slim it down a bit. Then you can make FW and PS play nice together and everyone is happy (well, nearly everyone I suppose.)

  • Matthias Lilke — 4:25 AM on May 15, 2006

    I hear you, Jack! I really don’t mean to start a flame-war but I can attest to this in regards to Mac OSX and my aging 450MHz Powermac. I started with 10.1 and the version increased in speed and features up to 10.4 which I’m using today.
    I read the article on Scott Byer’s blog about the old code in Photoshop being a hinderance in the Intel-move. Maybe this is a good time to look at the code and to a bit of spring cleaning. I don’t think you’ll get a better chance than right now. And even the windows side will profit from this.

  • George Penston — 9:58 AM on May 15, 2006

    The interface could grow simpler, more focused, more relevant to your needs (and your needs only)?

    This is what I find the most interesting in your what if’s. Although many software companies have done some interesting things when it comes to user experience in the past few years, I can’t think of any real popular software that has really attempted to tackle this.
    I added my suggestion along these lines to your post on future features to Photoshop. But in a nutshell, it would be interesting if the new Workspaces along with Keyboard Shortcuts & Menus evolved to accommodate tools and palettes. Then as I switch tools in Photoshop, it would bring up the relevant palettes and highlight the relevant menus for me. So if I selected the Slice tool, Photoshop in essence could switch to something more along the lines of ImageReady. That’s the best example I can think of. Apple’s iPhoto does this a bit when you bounce between reviewing your photos and editing them. The edit tools don’t appear until you need them.
    Another thing I find interesting is that Apple seemed to actually reduce the amount of menus in iPhoto through the version upgrades while adding more features. I’m always impressed by this sort of thing. Sure, many of the new features show up in the main window that you work with the most but it just goes to show that not every new feature needs to fill up the application’s menus as well.

  • Patrick T — 2:55 PM on May 15, 2006

    I can very much recommend the book About Face 2.0, by Alan Cooper and Robert Reimann.
    It deals with different classes of users and gives an allround fresh view on interface and interaction design (focussing on the latter)
    And Photoshop gets mentioned a few times as well, in a positive way even… :)

  • BWJones — 3:07 PM on May 15, 2006

    Of course one way to do this with products like Photoshop is to target specific markets (like science) who use and benefit from Photoshop. Perhaps Photoshop could function as a suite that loads up specifically used functionality and the attendant interfaces that are dependent upon the tasks desired. A morphing Photoshop if you will….

  • Brian Garrison — 4:38 PM on May 15, 2006

    I think the hard thing here would be deciding what is relevant to whom. Photoshop is used by so many different people in so many different ways, that I don’t see how you could take things out without teeing SOMEBODY off.
    Maybe if you could decide what features you wanted to use in some sort of preferences file, and only load those on startup…
    Creating different versions with application-specific functionality is a bad idea, from my point of view. I use almost everything in Photoshop at one time or another, and I don’t want any of it going away.
    I’m afraid there isn’t an easy answer to this question. I’m glad I’m not the one who has to figure it out!

  • J. Peterson — 5:25 PM on May 15, 2006

    See also Joel’s famous essay on the “80/20 Myth”. Maybe 80% of the people only use 20% of the features, but they never use the same 20%.

  • Skyler Kline — 1:09 AM on May 16, 2006

    I had a vision while on a bike ride today for the creative suite of the future.
    Rather that have seperate programs, you could have modules much like Lightroom that switched automatically when you changed to a different type of layer. You could have a raster layer, a vector layer, a type layer, a placed layer, a grouping layer, etc. They could all have raster, vector, and/or type masks
    It would save out pdfs natively, and open or place most any graphics document. You could do pdf style editing when working on a grouping type layer (folders). Everything in the selected folder could be exported with similar functionality to todays Acrobat.
    For interface functionality, I think Luxology’s Modo deserves some serious attention from the folks at Adobe. I think these ideas combined with those used in Lightroom would make an amazing interface for the above mention creative suite.

  • Michael — 9:19 AM on May 17, 2006

    Very interesting topic. Adobe potentially runs into the same problem that razor manufacturers have – the basic function of the razor was resolved years ago, but economic survival demands new and improved. Which is why we see 5 blade, dual lubricating strip, battery powered razors on the market.
    [On that note: –J.]
    Photoshop’s core image manipulation functions – selection tools, color correction, text creation, printing capabilities, etc. etc. currently perform at a very high level. It seems Adobe’s more or less solved the problem of the razor. You can tweak the core functions, redesign the interface, and add a degree of automation – but (to extend the metaphor) I’ll still be able to shave with a Bic. (by version 15, Photoshop 7 might seem like a Bic.)
    Additions like liquify and cloning in perspective fall into more of a more gee-whiz “5-blade” catagory (like a lot of 3rd party plug-ins), fun to play around with, but I’ve almost never used them for professional work. Unfortunately I still have to load them when I run the program.
    [Actually, on that last point, you don’t. Many (most) chunks of Photoshop don’t get initialized until you load them. If you never use Liquify or Vanishing Point, they never affect you more than any other 8MB of data on your hard drive. But because Photoshop startup isn’t as quick as anyone would like, it leads to the perception that new, specialized functionality is a drag on everyone. Therefore the answer is, I think, to work on speeding up launch time, and to make meat and potatoes functionality faster and smoother throughout the app. –J.]

  • BJ — 9:36 PM on July 14, 2006

    Have you seen 37signals’s “Getting Real” book? It talks a great deal about this.
    [I’ve heard of it, but I haven’t read it (still trying to educate myself about the ol’ Middle East before spending yet more time reading about software ;-)).]

  • Rhys Dippie — 2:28 AM on April 21, 2008

    I actually disagree that photoshop has perfected the razor – as Michael states.
    There are aspects of photoshop – like color correction that are absolutely dreadful and antiquated. Let us look at the curves control for example. Why are we limited to a tiny tiny box to edit a fiddly little curve in? Why can’t the curve be resized to cover half the screen? Why do we have to put up with ridiculous weird interpolation? (yes I know, it’s to try and stop newbies from putting kinks in their curves) well, sometimes I WANT kinks in my curves and I don’t want to micromanage 6 points to do this. Look at almost any compositor out there (apart from after effects of course which suffers from the same affliction) Because compositing applications have competition, they wouldn’t last 5 minutes in the marketplace with such user hostile tools.
    Also as a side note, Photoshop needs a good 3 way colour corrector – it is intuitive and fast to use to develop grading looks and again it’s very simple to code.
    One other area that I find I always have trouble explaining to new photoshop users are how the channels work. It is STUPIDLY long winded and fiddly to take a mask and put it into a mask slot for a layer, or to take a mask and make it a layer. How about this. If I take a layer and drag it into the layermask slot of another layer – it copies and converts it to a mask? is that really so hard?
    Also when I drag it out, it should allow me to drop it as a standard layer in the stack.
    This simple interface tweak would save MAJOR time in certain workflows and I cannot see any downside at all.
    Thanks for running this excellent blog – It’s great to hear a friendly voice from a large corporation.

Copyright © 2020 Adobe Systems Incorporated. All rights reserved.
Terms of Use | Privacy Policy and Cookies (Updated)