From: Ludovic Brenta on
Yannick Duchêne wrote on comp.lang.ada:
> Ludovic Brenta a écrit:
>>> Sorry, I know Ada is about content and the image it radiates,
>>> esthetically, maybe the least of your concern (I don't point you
>>> directly Ludovic) but the web site of ADAIC is a time machine to
>>> ~1993.  I came across that site often and I left it as many times
>>> because I though this was an obsolete site!
>>>
>>> Seriously, it needs refreshing!
>>
>> I agree.  Maybe that's the problem with Ada people: they spend too much
>> time writing quality software and not enough doing marketing.  I try to
>> do my part in the technical marketing department but I'm not a webmaster
>> or a graphics designer.  I think you'd get the same kind of anwser from
>> most people on this newsgroup.
>
> If may give my opinion.
>
> 1) Most of time, people who says “software designers are poor graphic  
> designers”, are themselves poor software designer. They stop on graphic  
> design, because this is the sole thing whose quality they are able to  
> evaluate (does not only applies to Ada, this also applies with web  
> applications written in JavaScript and the like).

I differ.

I use emacs in full-screen mode with no menu bar, no button bar and I
rarely ever touch the mouse. My "desktop environment" is
ratpoison[1].
I wrote a few web pages in the past: in emacs and hand-written, with
no
JavaScript or CSS. When I browse in firefox, I have a habit of doing
View > Page Style > No style to *disable* any fancy (and usually bad)
CSS layout that the "web designer" at the other end thinks is "cool"
but, to me, just gets in the way. I like my web pages simple and to
the point, just like the software I use or write. I am not a graphics
designer and definitely not a web designer but I think I am a good
software engineer. Maybe I *could* become a good web designer but
I'm just not interested. Back in the days when I worked on a large
web application, I avoided the "web" part and concentrated on the
back-end (server) part. I like Plain Text. I prefer command-line
tools over GUI tools, so when I write a utility for my own use, I
never give it a GUI.

[1] http://www.nongnu.org/ratpoison/

> 2) People seeking for documentation, should be aware that, when seeking  
> for documentation, documentation should come first! Not graphical design!  
> What can you expect from someones who run away from a graphical design and  
> prefer to go elsewhere, with may be a better graphical design, but lower  
> quality documentation. Look at Jean Pierre Rosen's AdaControl : may some  
> others sites has a better graphical design than his one, but none has such  
> an application as AdaControl. A choice must be made and every one is  
> responsible for his/her choice.

This would seem to indicate Jean-Pierre and I are of the same breed:
excellent software engineers, lousy web designers :)

> 3) AdaIC has a graphical identity, if it change it, pretty sure many body  
> gonna ask themselves what happened (“is it still the same AdaIC ?”)
> 4) AdaIC design is not so bad, it is even good (well balanced and  
> constant). Old fashioned ? What was nice 15 years ago is ugly now ? So  
> people 15 years ago all had bad taste ? Something is silly in this  
> assertion, and even some famous web designers agree that this is a silly  
> point of view (some lough out loud when they here about so called “Web 2.0  
> graphical style”).

I too like the style of AdaIC, except for the hardcoded "column"
layout
that wastes precious horizontal space.

> 5) If you want to differentiate from others, don't copy everyones else  
> graphical identity, get yours.

Right; don't be a lemming.

--
Ludovic Brenta.
From: Georg Bauhaus on
On 6/13/10 7:26 PM, zeta_no wrote:

> To be sure you understand my point here, I see this type of tutorial
> as something that can be quite long and involving. Rarely you can
> find that type of a document, that does not qualify as a book nor as a
> quick and dirty tutorial, as being organized halfway between a master
> thesis and a final year thesis. It treats a known subject (compared
> with the master thesis), with solid knowledge and neat understanding
> (compared with some final year thesis). These documents are great
> starters to light up curiosity and understanding to then permit
> austere books to sink in.

Indeed, there is an opportunity to set up an Ada teaching project in
the much needed style of Touch of Class (ETH Z�rich) or similar.
Much needed if it is true that CS education is no longer a sufficient
credential for some industries. A teaching project could extend
across several semesters, can include virtual production
scenarios, require teams, will start from a larger and serious
piece of software (to be written by university staff) and
include a collection of non-accidental hardware.
This can be fun (model trains is an example, I imagine there are
others, equally realistic and equally affordable). I mean, we
get to read about pressure sensors and thermometers all the time.
But are the real things part of CS proper?

As a modest example, I guess, it will be both possible and
instructive to teach a sorting algorithm that runs on a
microcontroller. But you cannot, I think, teach that in one week!

(Thinking further, it might be necessary to not be too specific
in the project, since being specific means being exclusive
and you loose your audience:

- doesn't have to be control software requiring 2 years of soldering
and quantum effects;
- a banking scenario offers enough to engage students for years,
you might kill the project, though, by asking for 1 year experience
in enterprise accounting;
- if there is a number crunching department, then there might be
modules that need to be written and that can be written without
a degree in math, but whose effect is seen in a larger computing
effort (it is demonstrably difficult to find mathematicians who
can even imagine that students from other departments find math
more difficult than they do, so the tasks would have to be
chosen to meet either need))

Do professors like it? Will they be allowed to do it?
I can't say whether university management can consider anything but
the latest style of measuring university reputation, such as
number of publications with "impact". If it turns out that
industry as a whole will like successful teaching as much as
anything else, this should encourage larger teaching projects.

The purpose of the Bologna process in Europe is to establish
a "European degree". I'd think that a long term CS project
offers an interesting teaching instrument. It even achieves
measurable success, and can be re-adjusted when needed,
without anyone needing to feel bad.
From: Nasser M. Abbasi on
On 6/14/2010 1:03 AM, Ludovic Brenta wrote:

> I prefer command-line
> tools over GUI tools, so when I write a utility for my own use, I
> never give it a GUI.
>

May be this is getting off-topic a little...But I think there is a place
for both?

Many things are really hard to do with just a command line. One needs a
good GUI for that.

And somethings are easy to do with a command line.

An example: I use rsync command to do my backup. a simple command. A gui
might actually get in the way here.

But when I want to move a file from one folder to another, rename files,
etc... I find it easier to use a file manager (a GUI) and the mouse,
rather than use the command line for this.

When I want to use ftp, I use FileZilla, a GUI frontend to ftp. Much
easier to use that using ftp commands. I forgot even the ftp commands by
now.

I think these days, in the web era, langauges with little GUI support
build into them can cause these language not to become popular.

Ada lack of good GUI integration is a problem. The other night I played
with GTKada. Wanted to learn how to make GUI with Ada. Downloaded
examples from libre web site,

http://www.adapower.com/launch.php?URL=http%3A%2F%2Flibre.act-europe.fr%2FGtkAda%2F

read the documents, and build them OK on windows. I found all the
examples to be trivial, a single button that come up and such.

Other GTKAda small programs that I found on the net (so few out there),
either do not build OK, or if they do, they crashed when ran.

These days, one needs a good language (like Ada), but also needs good
support for GUI build-into the language. The area I am interested in,
scientific applications, the ability to make plots and write GUI
front-ends for an application to interface to the algorithm is as
important as the algorithm itself. Need both.

This is one reason Matlab is popular. The language is not so good
actually, but it is so easy to make GUI's and do plots with it.

Not trying to compare matlab with Ada here, apples and oranges, but just
pointing out that the ability to make GUI in a language is important for
it to become "popular" these days.

my 2-cents.

--Nasser
From: Dmitry A. Kazakov on
On Mon, 14 Jun 2010 00:35:31 +0200, Yannick Duchêne (Hibou57) wrote:

> 3) AdaIC has a graphical identity, if it change it, pretty sure many body
> gonna ask themselves what happened (“is it still the same AdaIC ?”)
> 4) AdaIC design is not so bad, it is even good (well balanced and
> constant). Old fashioned ? What was nice 15 years ago is ugly now ? So
> people 15 years ago all had bad taste ?

It is not a question of taste or fashion. There are some objective
ergonomic rules, when violated make most people consider the page odd.
Three examples from AdaIC. it has two bars one vertical, one horizontal
with overlapping between, where is what? The colors: never use more than
one pure color, even if you wanted a shrilling visual effect (which AdaIC
did not). Never mix pictogram and text labels.

> 5) If you want to differentiate from others, don't copy everyones else
> graphical identity, get yours.

.... choose your identity wisely.

--
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de
From: Dmitry A. Kazakov on
On Mon, 14 Jun 2010 01:47:52 -0700, Nasser M. Abbasi wrote:

> On 6/14/2010 1:03 AM, Ludovic Brenta wrote:
>
>> I prefer command-line
>> tools over GUI tools, so when I write a utility for my own use, I
>> never give it a GUI.
>
> May be this is getting off-topic a little...But I think there is a place
> for both?

No, all people are genetically separated into Emacs users and others. (:-))

> Ada lack of good GUI integration is a problem.

Yes. No binding to crappy C-library X will solve this.

> The other night I played
> with GTKada. Wanted to learn how to make GUI with Ada. Downloaded
> examples from libre web site,
>
> http://www.adapower.com/launch.php?URL=http%3A%2F%2Flibre.act-europe.fr%2FGtkAda%2F
>
> read the documents, and build them OK on windows. I found all the
> examples to be trivial, a single button that come up and such.
>
> Other GTKAda small programs that I found on the net (so few out there),
> either do not build OK, or if they do, they crashed when ran.

Well, that depends on the compiler/GtkAda you use. Unless you are using
GNAT GPL 2009, presently, you will have troubles.

--
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de