[Clipart] Way OT: templates, developers, and stuff

[Jonadab the Unsightly One]

Alan Horkan <horkana at maths.tcd.ie> writes:

> I think I'm probably taking a much broader view of what templates
> are.  I am thinking of clipart and stock images too, my templates
> might be little more than a document that combines some some clipart
> a nice border and some standard lettering to give you a Birthday
> card.

Yeah, but even then I wouldn't want to use the template, because it
wouldn't be exactly what I want.  I might like the border on one
template, but I might prefer the other images from another, and I'd
definitely have specific ideas about what lettering I want, what font
face I want it in, where I want it, how large, and so on.

I'm not against the existence of templates, mind you; I just don't
personally have any use for them.

> I think that is what I already am doing by contributing to
> OpenClipart.org adding to what will become a very large collection
> of resources.

Agreed.  Having the library of clipart is useful whether you want to
roll it up into prefab templates, or whether you prefer to just keep a
library of clipart and make your own documents on the fly.

> If you want to do something quickly and are more a user than a
> developer they are brilliant...  When you just want to stick up a
> notice saying Room to Rent, Lost Kitten, or Bicycle for Sale
> templates are useful.

Not to me.  When I want to stick up a notice saying Bicycle for Sale,
I'm going to have specific ideas about what font it should have, how
large, where, and so on, ideas that won't match any of the templates.
Give me a blank document and I'll whip up the sign in two minutes;
make me fight with a template and it'll take me twice as long or
worse.

Yeah, I know, I'm particular about how I want things.  I don't buy
computers either, because I have specific ideas about what I want for
each component, and the computer with enough RAM that I can stand to
use Gimp and Mozilla at the same time will have a goofy gamer-type
video card I don't want, instead of a nice Matrox card, and it'll have
junky onboard sound with no (or unsupported in Linux) hardware MIDI, a
case that's a pain to work in, and like as not a software modem.  So I
always build my own, because then I get what I want.

This is how geeks think, and that's why we don't care about templates.

> If I want a template I expect to choose one, either from disk or by
> using New from Template.

Good, we agree on something :-)

> I find interfaces that use Wizards/Assistants a pain in the ass if
> you have used the application much at all and the do get in the way
> of getting things done quickly.  That is not really what I have in
> mind in terms of templates although there are ways to make that kind
> of thing work without it getting in the way of more experienced
> users.

Actually, I think we agree on quite a bit.

> I'm surprised you tolerated MSWorks at all, 

At the time, it was what I had.  (This was circa 1998; StarOffice was
not very good at that point, if it even existed yet.)

> having had MSOffice 

I didn't have money for MS Office.  Heck, I don't feel like spending
$650 for MS Office *now*, much less in 1998 when I still had college
debts to pay and was working fast food.

I did get exposed to Office in college, but that was a version that
ran on Windows 3.1.  (Word was version 6.0; I'm not sure what versions
the other components were.)

> I never touched Works with a bargepole after my initial attempt to
> use it.  (I guess I have never given it a fair evaluation I was so
> revolted on my first attempt almost a decade ago).

Actually, the database in Works was really nifty.  It was strictly
one-table-per-database, but for consumer-grade purposes that's plenty.
It could do some things in 1998 that the frontends available for the
major RDBMSs today still don't do AFAIK.  (We're talking UI things,
like using different fonts for different columns, but still.)

These days of course I use DBI and whip up CGI frontends.

> I like to take a template and make a few modifications to make it
> look a little bit different when I'm constrained for time.  I dont
> much like doing redundant work that wont be reused so I am
> reasonably interested in creating templates that others might
> appreciate, the hard part is figuring out what might be useful to
> others.

>From what I've seen of people who do use templates, I think that
you're on the right track with posters and greeting cards.  I'm not
really someone to ask about this though, since like I said, I don't
use them.

> Once an application has a substantial userbase it should be possible
> to get users to submit some of the files they have created using the
> application and make them into slightly more generic templates.  

That's an approach I can see working.

> (I have a few stray gnumeric files lying around waiting for me to
> make something useful out them.

Heh.  I don't think my spreadsheets would make good templates.  (I've
got a really great personal finance manager in Calc, but it positively
reeks of the unusual and geeky way I do things, and I don't think most
people would appreciate its eccentricities.)

> I have a very broad idea of what a developer is.  

The line between power user and developer is just as blurry as the
line between end user and power user.  

I tend to define them in terms of the following traits:

 1.  An end user more often than not only wants to know how to do the
     specific task that he wants to get done right now; he does not
     want to learn general principles or understand why things
     work, and he doesn't want to experiment.

 2.  The poweruser wants to understand the user interface, so that he
     can be more efficient using the computer.  He wants to learn
     things that will save him time in the long run.  He'll experiment
     with software to see how it works and what it can do for him.

 3.  The developer doesn't just want to learn how to use the computer;
     he wants to *change* how it works, so that *it* does what *he*
     wants.  He doesn't just want to understand the interface; he
     wants to understand all the implications of everything that's
     going on under the hood, and he'll tinker to find out.
     
Of course, there are developers and then there are developers.  On the
above scale, a lot of web content developers would be powerusers.
Personally, I fall somewhere in the vicinity of the line between
poweruser and developer.  I want to change how stuff works so that it
suits me, but there are limits to how far under the hood I'll go to
accomplish that.  I think in Perl, but I never got much of a handle on
C; it's too low-level for me.  I read RFCs for protocols like POP3 and
SMTP, but I don't have any inclination to learn how ethernet works.

In the commercial world, you have managers, whose job is to talk
developers out of doing whatever they want (mostly stuff end users
would never appreciate) and into doing stuff that will make users
(or the marketing department) happy, stuff which the developers
themselves probably feel is unimportant.

> some think only programmer, but really vital projects like Inkscape
> make developers out of translators, documentation, writers and
> artists, usability people, bug finders, testers and I'm sure there
> are other roles I'm not thinking of.

Sure, yeah.  Like I said, there are developers and then there are
developers.  And then there are the really hardcore developers, the
people who rewrite entire kernel subsystems and submit compiler
patches before breakfast.  Those people are just a little scary.

Powerusers, incidentally, make great beta testers.  Usability QA is
best done by powerusers who spend a lot of time observing end users;
end users don't understand enough to get the developers to respect
what they are saying, and powerusers who spend most of their time with
other powerusers or with developers won't know what the end users
want.

Documentation is best divided.  The people who write the code need to
document it at some level, but tutorials and end-user documentation
are better written by powerusers, because they understand the end
users better.

-- 
$;=sub{$/};@;=map{my($a,$b)=($_,$;);$;=sub{$a.$b->()}}
split//,"ten.thgirb\@badanoj$/ --";$\=$ ;-> ();print$/