RE: OSD

12 Mar 2004 - 3:55pm
10 years ago
12 replies
425 reads
Christian Simon
2003

>Not really. Bad code is bad code. When it is commented, all it allows a
>good coder to do is realize they need to toss it and rewrite the whole
>thing.

When I say bad code I hope you understand that I mean !perfect and not
wrong. This gets to the heart of my comment. The difference between design
and programming. Coding with wrong elements is fundamentally flawed and is
called a bug. Design needs the wrong answers because they are the basis for
understanding. A design that's "wrong" to some is edgy to others.

I'm not being glib here. That's why a design language is the closest thing
to open source design, But its not Design. design is the consideration of
these characteristics where the wrong answer is the right choice when the
constraints of the project and the understanding of the users is very
different.

Xtian

_________________________________________________________________
Christian Simon | www.christiansimon.com | San Francisco Bay Area
_________________________________________________________________

Comments

12 Mar 2004 - 4:36pm
vutpakdi
2003

--- Andrei Herasimchuk <andrei at adobe.com> wrote:
> On Mar 12, 2004, at 10:05 AM, Christian Simon wrote:
>
> > There is an understood value to the code a programmer can produce.
> > Some get
> > paid per line?
>
> Hardly ever. They get paid for end results.

My experience has been that programmers get paid for delivering something
that can, by some definition, be shipped. Whether or not there are any
repurcussions for what happens (in terms of quality and usability) after
the product goes out depends on the company and environment.

One of our senior developers once pointed out to our R&D manager that they
could fire all of the "quality" folks (quality assurance, documentation,
and usability) and still be able to ship *something* and that firing
everyone but the "quality" folks would result in nothing being shipped.

(The above was part of an exchange where he was complaining mostly about
not having enough resources and partly about having too much "quality" on a
project.)

>
> > And, well commented bad code is still very "useable".
>
> Not really. Bad code is bad code. When it is commented, all it allows a
> good coder to do is realize they need to toss it and rewrite the whole
> thing.

Only if the code is well commented and kept up to date. If the code is
well commented but the comments are out of date (ie, it no longer applies
to the underlying code), the comments can lead a coder down the wrong path
for a while before she realizes the problem.

Ron

=====
============================================================================
Ron Vutpakdi
vutpakdi at acm.org

__________________________________
Do you Yahoo!?
Yahoo! Search - Find what you’re looking for faster
http://search.yahoo.com

15 Mar 2004 - 11:39am
Christian Simon
2003

Ron-
I was interested to read this example of office culture and quality of work
product. A non-success story can be misleading because the reasons things
are not working can be deceptively complicated by issues unrelated to the
job. In light of Ben Hunt's entertaining defense of laziness in
design/programming, this example is well made.

In print design there is a idea that the quality of the design should equate
with the shelf life. Software designed not to perform beyond certain level
is a tenant of open source. How else would people be inspired to help?
<Ehehe>

Cheers,
-Xtian

_________________________________________________________________
Christian Simon | www.christiansimon.com | San Francisco Bay Area
_________________________________________________________________

16 Mar 2004 - 9:31am
CD Evans
2004

That's it then!

Quality in interaction design should be qualified as how long it lasts.
And, I'm not talking about the name of a product, i.e. illustrator, I
mean how long has one chunk of design lasted without review, update,
change, etc.

This is why we shouldn't look at Illustrator as a well designed UI.
It's the components that make up the design. This harks back to
Christina's post on languages, widgets etc, Jenifer's and Welie's
patterns and the whole subject about Tabs.

Tabs have lasted. That's quality. So is the OK button. Hopefully Adobe
and Macromedia don't have it out over that one too... ;)

Seriously though, patterns, languages, emergent design globs, flash
libraries, open sources, etc, whatever we call colab design chunks
we've got to set some standards and I think Quality should be one of
them. There are different types of quality though (anyone?) but length
of visibility is definitely one of them.

For the long term..

CD Evans

On 15 Mar 2004, at 16:39, Christian Simon wrote:

> In print design there is a idea that the quality of the design should
> equate
> with the shelf life. Software designed not to perform beyond certain
> level
> is a tenant of open source. How else would people be inspired to help?

16 Mar 2004 - 10:47am
vutpakdi
2003

--- CD Evans <clifton at infostyling.com> wrote:
> Quality in interaction design should be qualified as how long it lasts.
> And, I'm not talking about the name of a product, i.e. illustrator, I
> mean how long has one chunk of design lasted without review, update,
> change, etc.

I'm not sure that I'd agree with the longevity assertion, particularly in
vertical market applications. Sometimes, an interaction design lasts a
long time because of the functional niche it fills and the momentum the
design has.

For example, some of the interaction design of software used for airline
ticketing and gate operations is atrocious from a design perspective and
looks like it dates from the 70's. Tons of arcane keystrokes and
abbreviations. *If* (and that can be a big if) the overall computer system
does support a GUI interface, the actual application is still the same
character based system. The system is fairly fast for an experienced user,
but I suspect that the interaction design has persisted in spite of the
quality of the design because of other factors.

Ron

=====
============================================================================
Ron Vutpakdi
vutpakdi at acm.org

__________________________________
Do you Yahoo!?
Yahoo! Mail - More reliable, more storage, less spam
http://mail.yahoo.com

16 Mar 2004 - 12:31pm
Dave Collins
2004

>That's it then!
>
>Quality in interaction design should be qualified as how long it lasts.

>And, I'm not talking about the name of a product, i.e. illustrator, I
>mean how long has one chunk of design lasted without review, update,
>change, etc.

Well, some of the lousiest designs have lasted the longest merely
because of their penetration. The Windows File Manager model for
example. It is a non-intuitive mental model, difficult for users to get
their heads around, difficult to file and retrieve items, even for
advanced users. Yet it has persisted because it is so deeply ingrained.

16 Mar 2004 - 3:46pm
Dan Saffer
2003

On Mar 16, 2004, at 10:47 AM, Ron Vutpakdi wrote:
>
> I'm not sure that I'd agree with the longevity assertion, particularly
> in
> vertical market applications. Sometimes, an interaction design lasts a
> long time because of the functional niche it fills and the momentum the
> design has.
>

I'm not sure I agree either. You can have a great design that only has
a limited shelf life. Technology changes, cultures change, and,
importantly, the behavior you were trying to influence might change.

Think about Adriana Parcero's SEAMS Project, one of the coolest
interaction design products I've seen in a while:

http://www.seams.la
http://mdp.artcenter.edu/~parcero/m6/catalog/press_release.html

"The SEAMS collection maps sub-standard clothing production onto the
garments themselves. Embroidered details and innovative info-labels on
the clothes themselves describe the state of the garment industry in
four countries: Japan, the United States, China and Bangladesh. The
SEAMS catalogue both follows and subverts the usual forms by enveloping
an enticing vision of the clothes themselves in socially relevant
content."

If, say, this project was extremely successful and changed the garment
industry in the next year, would we say that it had failed because it
wasn't around a long time? I don't think so.

Dan

Dan Saffer
M.Des. Candidate, Interaction Design
Carnegie Mellon University
http://www.odannyboy.com

17 Mar 2004 - 9:41am
CD Evans
2004

I'd say Dave's comment sums up best why I'm wrong. There are tons of
crop designs that have just 'hung around'. But, to be honest, I'm not
sure that length of presence isn't important.

Windows File Manager is 'easy enough'. I've seen other approaches, and
I think we need a new one, but I haven't seen anything that could
honestly replace it. It's too simple of a design. It works really well
with ten objects! Think of your favorite folder, there's a perfect
example of it working. It's just not scalable, well, not very well
scalable at least. With replacing outdated design, perhaps
'sustainability' is the ticket to understanding acceptable designs... I
don't know.

I'm all for social clothing and intuitive airline systems as well, but
the length of the old systems being in place does say something about
the designs: "There isn't anything better". I'd love to design an
airline gate system, but not with the current set of patterns for UI,
there just isn't anything we have that says, "yes, definitely better".

In terms of the Socially Relevant clothing, I'm a big fan of any sort
of relevant design. I think your key word here, Dan, is the 'influence'
it has. How can influence be regarded as 'quality'? Certainly it isn't
time related, or perhaps it is. Perhaps a design is more susceptible to
influence the 'longer' it has expired? Maybe for 'relevance', 'quality'
is less important?

I still think it's 'quality' we're after right now, not quantity,
teehee, and I think in a lot of cases it is time related. Albeit, the
time maybe on different scales and in different contexts and with many
different influences, but I would still say that Time plays a major
part in what is accepted as Quality. I'd be happy to hear more about
how I'm wrong though, especially because I'm apt to define Quality much
more than to sit around thinking about the essence of time itself.

Any tips on Quality?

CD

On 16 Mar 2004, at 17:31, Dave Collins wrote:

> Well, some of the lousiest designs have lasted the longest merely
> because of their penetration. The Windows File Manager model for
> example. It is a non-intuitive mental model, difficult for users to get
> their heads around, difficult to file and retrieve items, even for
> advanced users. Yet it has persisted because it is so deeply ingrained.

11 Mar 2004 - 5:33pm
Andrew Otwell
2004

> Most americans live in a bauhaus / frank lloyd
> wright spin off architecture,

I wish! Walking through my Austin neighborhood, I see mostly Texas
cottages jacked up with nouveau-riche ornament like stupid gas lamps or
tile roofs. Prefab apartments don't equal bauhaus spin-offs.

11 Mar 2004 - 5:37pm
Andrew Otwell
2004

> Design is modular. Most americans live in a bauhaus / frank lloyd
> wright spin off architecture, just as most britons live in a medieval
> village / terrace dwelling architecture. These are designs. Sony made
> the first walkman in a rectangle that opened on the side and had
> buttons top or side, not much has changed. Again and again these simple
> modular designs are reused exhaustively. From balcony handrails to
> volume wheels, design is modular.

Ok, here's my non-flip response. What you're talking about isn't
modularity. These are not design systems that use interchangable parts
or conform to some guidlines. Just because "buttons on top" is a common
approach, it's not a "module" of a larger design system.

What you're talking about are typologies, or perhaps design patterns in
the Alexander sense. These are strategies, "best practices", or rules of
thumb that have evolved together. I suppose one could see them as
"modular" in that someone like Alexander can express an overall system
for grouping them.

ao

11 Mar 2004 - 6:33pm
Jenifer Tidwell
2003

On Thu, 11 Mar 2004, Andrew Otwell wrote:

> > Design is modular. Most americans live in a bauhaus / frank lloyd
> > wright spin off architecture, just as most britons live in a medieval
> > village / terrace dwelling architecture. These are designs. Sony made
> > the first walkman in a rectangle that opened on the side and had
> > buttons top or side, not much has changed. Again and again these simple
> > modular designs are reused exhaustively. From balcony handrails to
> > volume wheels, design is modular.
>
> Ok, here's my non-flip response. What you're talking about isn't
> modularity. These are not design systems that use interchangable parts
> or conform to some guidlines. Just because "buttons on top" is a common
> approach, it's not a "module" of a larger design system.
>
> What you're talking about are typologies, or perhaps design patterns in
> the Alexander sense. These are strategies, "best practices", or rules of
> thumb that have evolved together. ...

Idioms, perhaps? Sometimes patterns; depends on the level of
abstraction, I think. And "best practices" might apply if the
idiom in question actually *is* a good solution (and they
sometimes aren't).

When they come in groups, I'm not sure what to call them. But
they do "clump" sometimes. Newly-built American houses use a
bunch of recognizable design idioms, as do apartment buildings,
and British village houses. We can see that in different genres
of software too: corporate web pages, blogs, e-commerce sites,
"big apps" for Windows desktops, etc. Certain design idioms
and patterns are recycled within these genres, and help a user
recognize a site (or whatever) as belonging to one of these
genres.

Can we call these "emergent design languages?"

Certainly they're not well-designed "design systems," as Andrew
puts it, in the same sense that a single company might build
a family of products to conform to such. But they're related.

- Jenifer

--------------------------------------------
Jenifer Tidwell
w: jtidwell at mathworks.com
h: jtidwell at alum.mit.edu
http://jtidwell.net/

12 Mar 2004 - 1:05pm
Christian Simon
2003

>> :Jason said something that got me=
>> So when designing/building anything we have to make guesses -- the hope is
>> that
>> these are at least educated guesses.
>>
It's not practical to share guess work. Consider that design is also a
process of revision that has a eureka moment. This is different then the
programmer¹s a-ha after finding the snag in complicated hours of code.
ehehe. Is it because of working on several points of a design require very
connected and ongoing interactions? There is no easy place to stop.

Using Flash in conjunction with a component library (called Smart Clips)
requires some coding. It comes close, and it looks more like an end product.
Possible, though not many coders will accept a Flash file as extensible to
their part of the product.

There is an understood value to the code a programmer can produce. Some get
paid per line? And, well commented bad code is still very "useable". IxD
produces many different types of products in multiple formts. What part of
the IxD work process can be shared that makes coders and designers feel they
both r giving something valuable?

-Xtian

12 Mar 2004 - 1:52pm
Andrei Herasimchuk
2004

On Mar 12, 2004, at 10:05 AM, Christian Simon wrote:

> There is an understood value to the code a programmer can produce.
> Some get
> paid per line?

Hardly ever. They get paid for end results.

> And, well commented bad code is still very "useable".

Not really. Bad code is bad code. When it is commented, all it allows a
good coder to do is realize they need to toss it and rewrite the whole
thing.

Andrei Herasimchuk
andrei at adobe.com

work: http://www.adobe.com
personal: http://www.designbyfire.com

Syndicate content Get the feed