On Thu, 19 Oct 2006 13:10:31 -0700, Georg Wrede <georg.wrede@nospam.org> wrote:

> John Reimer wrote:
>> Lionello Lunesu
>>> I'd prefer sourceforge, where it will be seen by more people.
>>>
>>> dsource is nice and has certainly done a lot for D, but it makes us look  like some sort of cult.
>>  What?!!!
>
> LOL!
>
> I'm not entirely sure he doesn't have a point.
>
> (At least partially.)


Just maybe... but I was hoping not. :D

-JJR

John Reimer wrote:
> On Thu, 19 Oct 2006 13:10:31 -0700, Georg Wrede <georg.wrede@nospam.org>  wrote:
> 
>> John Reimer wrote:
>>
>>> Lionello Lunesu
>>>
>>>> I'd prefer sourceforge, where it will be seen by more people.
>>>>
>>>> dsource is nice and has certainly done a lot for D, but it makes us  look  like some sort of cult.
>>>
>>>  What?!!!
>>
>>
>> LOL!
>>
>> I'm not entirely sure he doesn't have a point.
>>
>> (At least partially.)
> 
> 
> 
> Just maybe... but I was hoping not. :D

The thought never occurred to me that it was cult-like.  My thoughts were
1) cool! I can find (most) all of D related projects in one place
2) cool! These guys are serious enough about this to setup their own sourceforge
3) cool! It's not s***forge!

Sourceforge is popular (and I use it a lot) but there's very little excellence there.  It's free and well-known.  That's about it all it's got going for it.  Various services are frequently down or unusably slow.  And it would be hard to imagine worse web interfaces than the ones they've designed for just about every aspect of project management and distribution.

--bb

John Reimer wrote:
> On Thu, 19 Oct 2006 13:10:31 -0700, Georg Wrede <georg.wrede@nospam.org>  wrote:
> 
>> John Reimer wrote:
>>
>>> Lionello Lunesu
>>>
>>>> I'd prefer sourceforge, where it will be seen by more people.
>>>>
>>>> dsource is nice and has certainly done a lot for D, but it makes us  look  like some sort of cult.
>>>
>>>  What?!!!
>>
>>
>> LOL!
>>
>> I'm not entirely sure he doesn't have a point.
>>
>> (At least partially.)
> 
> 
> 
> Just maybe... but I was hoping not. :D
> 
> -JJR

Oh, also pooling code is also something that seems to be popular with other easy-to-use languages.  Perl has CPAN, Python has the "cheese-shop", Ruby has Rubyforge, ... but I guess those are all cults now that I think of it.  :-P

--bb

Bill Baxter wrote:

> You generally flow down the steps from top to bottom, but at *any* stage you can get loops back up higher.

Are you sure?

Assume an idealized situation where you take over a team of 10 designers and 490 coders.

Your team has projects A and B in the pipeline.

Project A is in the beginning of the design phase, planned duration two years, and project B is in the beginning of the coding phase, planned duration also two years.

your remaining budgets:
for designing project A   4,000,000 $
for coding project B    147,000,000 $

This means every month delay will cost you at least 6,000,000 $

Now one of your 490 coders comes to you saying: "I am unable to implement this because of erroneous design."

Are you willing to "sell" this detection to your management, your sponsor or your loan officer?

And if you change positions: are you willing to "buy" such statements from yourself?

"Dear Sir, I have a plan to produce software that only costs you 150,000,000$. My main production plan consists of repeated designing and coding, because every design is flawed. But although every design is flawed I am sure I do not need more money than what I said before."

rm wrote:

> Do you know *1* project where no backtracking was ever needed?

Several. They died immediately on first try to backtrack.

Karen Lanrap wrote:
> Bill Baxter wrote:
> 
> 
>>You generally flow down the steps from top to bottom, but at
>>*any* stage you can get loops back up higher.
> 
> 
> Are you sure?
> 
> Assume an idealized situation where you take over a team of 10 designers and 490 coders.
> 
> Your team has projects A and B in the pipeline.
> 
> Project A is in the beginning of the design phase, planned duration two years, and project B is in the beginning of the coding phase, planned duration also two years.
> 
> your remaining budgets:
> for designing project A   4,000,000 $
> for coding project B    147,000,000 $
> 
> This means every month delay will cost you at least 6,000,000 $
> 
> Now one of your 490 coders comes to you saying: "I am unable to implement this because of erroneous design."
> 
> Are you willing to "sell" this detection to your management, your sponsor or your loan officer?
> 
> And if you change positions: are you willing to "buy" such statements from yourself?
> 
> "Dear Sir, I have a plan to produce software that only costs you 150,000,000$. My main production plan consists of repeated designing and coding, because every design is flawed. But although every design is flawed I am sure I do not need more money than what I said before." 

It sounds like we're just talking about things on a completely different scale.  We're talking about *iterations* on one design.  You seem to be talking about scrapping the initial design completely and starting over.  Although there is a school of thought that says you should always plan to "throw the first one away" -- i.e. build a prototype, learn from it, then do the real thing -- I will agree that in the real-world that's almost always disastrous.

But we're not talking about complete redesign from scratch and starting over with fresh code.  We're talking about design iterations. Incremental improvements to the original design.

--bb

Karen Lanrap wrote:
> Bill Baxter wrote:
> 
>> You generally flow down the steps from top to bottom, but at
>> *any* stage you can get loops back up higher.
> 
> Are you sure?
> 
> Assume an idealized situation where you take over a team of 10 designers and 490 coders.
> 
> Your team has projects A and B in the pipeline.
> 
> Project A is in the beginning of the design phase, planned duration two years, and project B is in the beginning of the coding phase, planned duration also two years.
> 
> your remaining budgets:
> for designing project A   4,000,000 $
> for coding project B    147,000,000 $
> 
> This means every month delay will cost you at least 6,000,000 $
> 
> Now one of your 490 coders comes to you saying: "I am unable to implement this because of erroneous design."
> 
> Are you willing to "sell" this detection to your management, your sponsor or your loan officer?

The typical response is to the coder to figure it out, ship a product with structural issues, and let the maintenance team deal with it.  And ten years down the road the product is so filled with such compromises that even theoretically trivial changes take days to fix, and often introduce additional bugs in the process.  But that's the coder's fault for not being capable or efficient, so you reprimand him and hire more programmers to pick up the slack.  Eventually a competing product is released by another company and you simply can't compete on a feature/cost basis.  New sales drop off a bit and you lose a few existing customers to the competitor, but encourage others to stick around with fancy service contracts and promises of a redesign.  And so it goes.  ;-)


Sean

Karen Lanrap wrote:
> Bill Baxter wrote:
> 
>> You generally flow down the steps from top to bottom, but at *any* stage you can get loops back up higher.
> 
> Are you sure?
> 
> Assume an idealized situation where you take over a team of 10 designers and 490 coders.
> 
> Your team has projects A and B in the pipeline.
> 
> Project A is in the beginning of the design phase, planned duration two years, and project B is in the beginning of the coding phase, planned duration also two years.
> 
> your remaining budgets:
> for designing project A   4,000,000 $
> for coding project B    147,000,000 $
> 
> This means every month delay will cost you at least 6,000,000 $
> 
> Now one of your 490 coders comes to you saying: "I am unable to implement this because of erroneous design."
> 
> Are you willing to "sell" this detection to your management, your sponsor or your loan officer?
> 
> And if you change positions: are you willing to "buy" such statements from yourself?
> 
> "Dear Sir, I have a plan to produce software that only costs you 150,000,000$. My main production plan consists of repeated designing and coding, because every design is flawed. But although every design is flawed I am sure I do not need more money than what I said before."

please tell me which firm you're acquainted with, it's certainly a criterion to take into account whenever I've dealings with them.

I've seen the thing you propose as the most normal business in the world first hand. You know what happens? People do it, but leave after a year or so, the software is a mess, and non-maintainable, non-extendable. The farther you're in the cycle the more it costs to backtrack, and sometimes you do not have a choice, but with every hack you put in place, you subtract some of the lifetime of your software, and I'm not talking about D. D, C++, Python all can be used in whatever scenario you want.

roel

Bill Baxter wrote:

> It sounds like we're just talking about things on a completely different scale.

Maybe you overread my statement on "very tiny set of coworkers".


> We're talking about design
> iterations. Incremental improvements to the original design.

But then you never started to code---and that seems to be a promising way. Have a look at the ASM design approach.

Sean Kelly wrote:

>  And so it goes.  ;-)

Quite true, but only if you had the right touch.