Java Virtual Machine

koolmansam375 · Post by **koolmansam375** » Fri 11 Mar, 2005 12:11 am

Patori wrote:
CoBB wrote:The hardness of asm is a myth. In fact, it's the simplest language you can imagine on a computer.
Yeah, it is very simple. Try it. IF you need help, ask us.

you know that pinball game for the z80 calcs? well CoBB wrote that and its in assembly too (i dunno if that post was sarcastic or not but anyways...)

CoBB · Post by **CoBB** » Fri 11 Mar, 2005 7:15 am

Well, Tijl, if you could manage it this way, it only proves you're much brighter than the majority. That doesn't come as a big surprise to me.

Unfortunately most people just can't cope so easily.

As for present day software building, most of the projects rolling out are a total mess. If you look at the skills of those who code the remaining ones you'll rarely find people who are uneasy with asm or C even if they don't use them actively any more. And what you're talking about is software design, not coding. Entirely different sets of skills. A designer doesn't need to care about object lifetime management for that matter, but a programmer who can't distinguish class instances from the references to these instances won't be able to produce good code even if handed the most brilliant design. I agree that the designer doesn't need to know much about the machine, they just need to decompose the problem into sensible modules and organise these parts into a sensible structure. This is not a even a CS related skill, it is basically what engineering is about. On the other hand, coding is a technical task and it requires practically no planning once the specs are available.

Seriously, I love declarative languages. I love the abstraction level they provide. Still, when I'm coding in Prolog, I'm grateful for knowing how my program will be translated to WAM instructions and how the WAM itself is implemented in terms of machine language primitives. Same goes for functional langauges and the implementation of lazy structures. Yes, I could actually program without this knowledge, but it would take significantly longer time to solve problems that come up during the development of non-trivial applications. I love low-level languages too, because they don't conceal anything from me. This is a disadvantage too, since this also means that I need to care about little implementation details just everywhere. No wonder they are not used in production environments, but they should never ever be left out from the learning phase, as they give you invaluable experience.

Post by **Patori** » Fri 11 Mar, 2005 2:03 pm

wait... Tjil Coosemans? Who has that screenname? I kind of wanted to talk to him, and thank him for his port of tetris to the 83+

Post by **kv83** » Fri 11 Mar, 2005 2:11 pm

Timendus

Post by **Patori** » Fri 11 Mar, 2005 2:14 pm

wait... *missed something*...

Post by **kv83** » Fri 11 Mar, 2005 2:15 pm

Patori wrote:wait... *missed something*...

Timendus = Tijl Coosemans... Timendus is his screen name... Tijl his real name

Post by **Patori** » Fri 11 Mar, 2005 2:27 pm

kv83 wrote:
Patori wrote:wait... *missed something*...
Timendus = Tijl Coosemans... Timendus is his screen name... Tijl his real name

I got that much, I just didn't know that he was he..... thanks for this valuable intel...

Timedeus, thank you for your port of ZTETRIS!!! Keeps me alive in math class....

Kalimero · Post by **Kalimero** » Fri 11 Mar, 2005 4:34 pm

CoBB wrote:As for present day software building, most of the projects rolling out are a total mess.

I wonder if this is because of bad programming or bad designing. The pressure lies to much on the coding. "Show me what you've done" "Wait, you don't have a gui yet?" Same with the OSS community. Its motto is "release early, release often". That's completely wrong in my opinion.

You know architects can design a house on a computer and show you a 3D rendering. It would be nice if you could design a software project this way (not a 3D rendering of course, but something equivalent). You could then even test some parts of the design without actually building it. I guess I'm dreaming...

kv83 wrote:Timendus = Tijl Coosemans... Timendus is his screen name... Tijl his real name

Cool, my name is Tijl Coosemans too and I thought that name was unique in the world.

Post by **kv83** » Fri 11 Mar, 2005 4:43 pm

Kalimero wrote:
CoBB wrote:As for present day software building, most of the projects rolling out are a total mess.
I wonder if this is because of bad programming or bad designing. The pressure lies to much on the coding. "Show me what you've done" "Wait, you don't have a gui yet?" Same with the OSS community. Its motto is "release early, release often". That's completely wrong in my opinion.

You know architects can design a house on a computer and show you a 3D rendering. It would be nice if you could design a software project this way (not a 3D rendering of course, but something equivalent). You could then even test some parts of the design without actually building it. I guess I'm dreaming...

kv83 wrote:Timendus = Tijl Coosemans... Timendus is his screen name... Tijl his real name
Cool, my name is Tijl Coosemans too and I thought that name was unique in the world.

Oops... I was already afriad that I mixed that up... ok... you may kill me now

CoBB · Post by **CoBB** » Fri 11 Mar, 2005 4:45 pm

Well, completely visual development is of course possible, I just doubt that it helps you explain your cause to non-technical people. You'd just have big obscure drawings versus little obscure drawings instead of many versus few LOC. Most people can't see behind the interface and can't comprehend that it comprises only a little fraction of the work done. No magic technology will change that.

Kalimero · Post by **Kalimero** » Fri 11 Mar, 2005 6:39 pm

CoBB wrote:Well, completely visual development is of course possible, I just doubt that it helps you explain your cause to non-technical people. You'd just have big obscure drawings versus little obscure drawings instead of many versus few LOC. Most people can't see behind the interface and can't comprehend that it comprises only a little fraction of the work done. No magic technology will change that.

True, but I think it could shift the focus from coding to designing

CoBB · Post by **CoBB** » Fri 11 Mar, 2005 7:32 pm

There is another thing technology won't change, and that would be the demand for creating products of high standards. We have the necessary means to create quality software in our disposition for decades. It's all a matter of organisational skills to use them properly. Technology won't solve human sloppiness issues either, unless the machine can automatically verify if a design is feasible the same way it can halt on syntax errors today.

Food for thought question: what do you think how a compile time error would look like in a visual design tool that generates code automatically from the hand-drawn specs? In other words, what would be the equivalent?

Post by **Patori** » Fri 11 Mar, 2005 8:00 pm

kalimero wrote:
kv83 wrote:Timendus = Tijl Coosemans... Timendus is his screen name... Tijl his real name
Cool, my name is Tijl Coosemans too and I thought that name was unique in the world.

What? *even more confused*

Timendus · Post by **Timendus** » Sat 12 Mar, 2005 1:56 am

Patori wrote:
kalimero wrote:
kv83 wrote:Timendus = Tijl Coosemans... Timendus is his screen name... Tijl his real name
Cool, my name is Tijl Coosemans too and I thought that name was unique in the world.
What? *even more confused*

Hehehe... that's a funny misunderstanding
Kalimero = Tijl Coosemans
Timendus = Tim Franssen
I know, it's difficult with all those Dutch names...

So no; I didn't port Tetris to the 83+...

the_unknown_one · Post by **the_unknown_one** » Sat 12 Mar, 2005 9:00 am

I already thought that kv was wrong

To add more confusion:
the_unknown_one = Bram Tant (i'm not dutch, but belgian, close enough)

MaxCoderz

Java Virtual Machine

lol