On Jul 22, 2:46 am, llothar <llot...@[EMAIL PROTECTED]
> wrote:
> It was about the EiffelDoc or how they call it. It has nothing to do
> with the runtime speed of compiled apps.
Oh. I did see this. It seems to me that EiffelDoc (a) performs the
same checks asthe compiler and (b) documents all of the SmartEiffel
code and tools as well. So, I can understand the slowdown if this is
neww functionality.
> Yes, but i'm not sure if this is marketing for Eiffel or bad marketing
> for my chaotic
> development style, my tight time schedules and my lack of money. I can
> only say that the
> ability to get the SmallEiffel stack traces from customers (i had to
> make a lot of patches
> to the exception handling to get this right) was the hugest benefit of
> all and it pays back for all the other problems i had from Eiffel.
> Sure a lot of it depends on the
> DBC in the runtime libraries and that errors are re****ted
> immediately.
Excellent. And those issues you mentioned are common development
problems that ALL projects deal with. Trust me. I have seen
some very poor, well funded, large development efforts!
> So yes, i find Eiffel still great, thats why i'm reading this
> newsgroup and still using it. I like it so much that from time to time
> i'm thinking about producing my own ECMA Eiffel compiler (i certainly
> would do it if SmallEiffel had a BSD License). But then the real
> worlds comes in and remembers me to make money to pay my bills on each
> first of a month.
>
> I like writing compilers and i have a lot of ideas what could be done.
> Maybe if a new generation of CPU with fine grained multithreading
> comes
> it is a chance because for this everything needs to be rewritten.
What about the GOBO complier? Several people have told me how well
it is written and how well it works.
> Read only. Don't forget this. And - yes, its ba****ng - between 2003
> and 2005
> every single decision they made was IMHO wrong. There aren't many
> software
> projects where i can say this.
So, have they done better since 2005? :-)
> Frameworks are different then libraries. First you need the libraries
> and even this is something not yet available. Don't take my words for
> EiffelStudio - i'm commenting on SmallEiffel, but i heared that
> EiffelStudio
> still has the problems. Look at Python and then look back into Eiffel
> world.
I agree that you need libraries first, but as you posted early last
year (i.e,
3/2006), people (i.e., users) need to help with the development. I
have seen
the SmartEiffel team and SmartEiffel users get frustrated that no one
wants to invest the time. Having 4 kids, I know that time is
precious. That's
why I have been encouraged to see Enterprise SmartEiffel and the
Eiffel
Wrapper Libraries Collection projects. As for EiffelStudio, I can see
some ongoing projects, but most people creating those projects seem to
be using GOBO. I have thougt of that too, but I anticipate needing
agents, tuples, and a few other features.
Lee
|
