Eiffel needs more then a good open source compiler

[llothar]

Two month after the ISE deal it seems that the people who thought
"Eiffel just needs a good open source compiler" are really wrong.

Instead of seeing more newbie questions this newsgroup is more silent
then ever. A full month of threads on one groups.google page is less
then usual.

So is eiffel doomed or already buried ?

[Georg Bauhaus]

Or is usenet gradually loosing sex appeal?
Or are newcomers just writing Eiffel exercises, drawing upon
years of successful Eiffel FAQs?
Or is it that the people of Paris have finally decided to
tear down the Eiffel tower, so better not use that language?

What license have people been favoring when asking for an
open source Eiffel compiler? GPL?

FUD

Georg

[llothar]

> Or is usenet gradually loosing sex appeal?
> Or are newcomers just writing Eiffel exercises, drawing upon
> years of successful Eiffel FAQs?
> Or is it that the people of Paris have finally decided to
> tear down the Eiffel tower, so better not use that language?
>
> What license have people been favoring when asking for an
> open source Eiffel compiler? GPL?

Georg you sound like a Nazi in the Fuehrerbunker in Januar 1945.
We will win, we don't need to change anything. We don't need to know
why because we are the Uebermensch. Or to quote Hermann Goering "It
wasn't 1000 years, but the 12 were nevertheless very good".

Sorry but i think that an analysis of why eiffel is constantly failing
with its ideas to improve the situation is really necessary - i did
hear or see nothing about this when EiffelStudio became GPL virused.

I want to hear and discuss opinions from the people who have some
influence and could help in this situation a bit more. Especially
Daniel Moisset or someone else of the SmartEiffel Team (maybe - i told
them i will never talk with this fools again), Emanuell Stapf or
someone else from ISE/Eiffel Software.

Well maybe Eiffel is really not more then an item for software
archeologists. But i'm not convinced because at least i'm using a fork
of this language still 8 hours a day.

[John Perry]

> Well maybe Eiffel is really not more then an item for software
> archeologists. But i'm not convinced because at least i'm using a fork
> of this language still 8 hours a day.

In that case, maybe you could help out with the question I posted earlier?

thanks
jack perry

[llothar]

Sorry i can't.

I'm very conservative in this case. I try to keep generics as simple as
possible and my rule is never use any inherited relationship in the
generics argument, so i learned (also because of some bugs in
SmallEiffel) to never use an ARRAY[NUMBERIC] and pass an ARRAY[INTEGER]
to it.

And in my case i never found a situation in real life where i had a
need for it.

[Georg Bauhaus]

llothar wrote:

> Georg you sound like a Nazi

I'm not trying to fight a pro Eiffel war. Are you?

Also, please always be careful with allegations of Nazism.

May I draw your attention to some improved pieces of Eiffel,
and to what in my view is still missing. For example,

- more steps towards LiskovSP (ECMA "only") is an improvement
(this includes concurrency, which is still somewhat
experimental, and system dependent, AFAICT.)

- lifting clusters to the language level is still missing: no
subsystems in the source code.

> Sorry but i think that an analysis of why eiffel is constantly failing

I think it's difficult to discuss a vague claim of constant failure
if no reference to points of failure has ever been given except
whining about normal compiler prices, and that Eiffel doesn't reign
the mass market.

> EiffelStudio became GPL virused.

Get the non-GPL version, then, if your programs require closed source.

> I want to hear and discuss opinions from the people who have some
> influence and could help in this situation a bit more.

I guess it will be most helpful if you identify what is needed
exactly, by way of a sufficiently precise enumeration of missing
items. How else could compiler makers, and library makes learn what
exactly their prospective customers need?
(I know this is difficult with SE, as they have a fairly strict
view of others' needs. A shareware author will need to be creative
when using SE if by shareware we also refer to a shareware budget
for development tools.)

The market has brought a number of "commercial" languages to a halt.
This happens, but apparently it isn't happening to all of Eiffel.

Georg

[peter_...@hotmail.com]

llothar wrote:
> Instead of seeing more newbie questions this newsgroup is more silent
> then ever.

Because the action's all happening on the EiffelStudio Origo mailing
lists, related to the EiffelStudio 5.7 open-source development.

Most EiffelStudio and EiffelEnvision discussion moved from this
newsgroup to http://groups.yahoo.com/group/eiffel_software -- years
ago! Unfortunately, that mailing list doesn't show up in Google.

- Peter Gummer

[gmc...@gmail.com]

Here's what I've noticed:

The number of subscribers to the ISE Wiki is over 120 people. I hadn't
realized that there were that many Eiffel programmers still around
(even allowing that many of them may be more Eiffel lurkers than Eiffel
workers).

That indeed, most of the dialog about Eiffel has been outside of
Usenet. The community's not large enough to support a more than a few
channels.

That I've actually seen a lot of interesting innovation occuring in the
Open Source project. Major improvements in the GUI, the project
support, and more thorough testing are all underway.

I've heard second-hand that ISE is very pleased with the result of the
Open Source release.

I'd say that instead of being dead, that it's too early to decide and
that there are some promising signs. Eiffel has always been a good
language crippled by inadequate tools and libraries. It will take much
longer than two months to get these deficiencies corrected.

Greg C

[Jens Knaack]

Hi,

two months? Thats nothing. We need to be patient and will see Eiffel
booming.

Jens

[peter_...@hotmail.com]

Greg C wrote:

> That I've actually seen a lot of interesting innovation occuring in the
> Open Source project. Major improvements in the GUI, the project
> support, and more thorough testing are all underway.

Yes indeed.

For example, a few weeks ago they released an experimental preview of
docking in the EiffelStudio IDE. They tried to do this in EiffelStudio
5.5 and 5.6, but it was really clunky and I found it useless. The
experimental preview, by contrast, was fantastic. There were bugs, as
you would expect of an experiment, but overall its docking worked
really, really well; so well, in fact, that I will be very disappointed
if it doesn't make it into EiffelStudio 5.7.

There's lots of other good stuff going on too: search in the IDE across
classes with regular expressions; a much improved system for organising
projects (sadly using XML gibberish: Lace is dead); ECMA expanded
semantics; fixes to some irritating .NET limitations; Unicode and
internationalisation; and someone's working on a Mac OS X Carbon port
of EiffelVision.

- Peter Gummer

[Scott Wyatt]

peter_...@hotmail.com wrote:
> semantics; fixes to some irritating .NET limitations; Unicode and
> internationalisation; and someone's working on a Mac OS X Carbon port
> of EiffelVision.

I'm not sure I want Carbon as much as Cocoa, but anything native OS X
will improve the chances of me working with something other than
Objective-C.

- CSW

[llothar]

> I'm not sure I want Carbon as much as Cocoa, but anything native OS X
> will improve the chances of me working with something other than
> Objective-C.

If Carbon is done right you will see absolutely no difference in an
Eiffel binding,
Carbon is just the C language binding for the Coca framework.

[Roger Browne]

Hi llothar,

> Instead of seeing more newbie questions this newsgroup is more silent
> then ever.

Yeah, everyone seems to have moved over to comp.lang.ecma-eiffel and
comp.lang.smarteiffel :-)

There's also the TeamEiffel blog of course:
http://teameiffel.blogspot.com/

where I have also commented on your points:
http://teameiffel.blogspot.com/2006/06/whither-complangeiffel.html

Regards,
Roger Browne

[Daniel F Moisset]

Hi,

llothar wrote:
> Georg you sound like a Nazi in the Fuehrerbunker in Januar 1945.

And by doing nazi comparisons, you sound like a troll in Usenet in the
1990s.

> I want to hear and discuss opinions from the people who have some
> influence and could help in this situation a bit more. Especially
> Daniel Moisset or someone else of the SmartEiffel Team (maybe - i told
> them i will never talk with this fools again), Emanuell Stapf or
> someone else from ISE/Eiffel Software.

What makes you think I am influential anywhere? I am one of the newest
members of the community...

If you want MHO, I have said it before, in several places. Eiffel is
not dying. It is being killed by petty political struggles by different
groups that are harming the language in the name of doing what it best
for it. They care more about the name of a class or a syntax construct
than in building a community, which is more valuable for a language
today. Of course, there are still a handful of people above that and
doing really helpful things, so I think it is not dead yet

Anyway, the simple and most heavy argument to say that eiffel is not
dead yet, is that you are asking this and stirring a discussion. How
many flames (or replies) can you get posting that "Prograph is dead"?
When that happens for Eiffel (probably at that time you won't even
remember to ask the question, anyway), it will be dead.

Meanwhile, even despite the petty people harming the future of the
language, Eiffel is still alive. As long as stupid people like me keep
reading these newsgroups and biting your trolls.

Regards,
D.

[Grant Rettke]

What .NET improvements have they made?

[peter_...@hotmail.com]

Grant Rettke wrote:
> What .NET improvements have they made?

According to the EiffelStudio 5.7 documentation's release notes,
"Removed limitation to inherit an Eiffel class which inherited from a
.NET class." That's been a big limitation in the past.

I've also noticed that there's been work on .NET properties, enums and
constructors. I don't know the details.

All should be clearer once 5.7 is closer to release.

- Peter Gummer

[Raphael Simon [ES]]

There will be more information available in the release notes but to
summarize (Peter's list is spot on):

* Every Eiffel attribute now generates a .NET property
* Every Eiffel query coupled with an assigner generates a .NET property
* It is now possible to use the 'inspect' clause on a .NET enum
* It is now possible to generate .NET constructors that will call Eiffel
creation routines. Use the 'dotnet_constructors' indexing clause to list
the name of the creation routines that should be called by constructors.

Also the consumer now shows an icon in the systray with some information
about what it is consuming. If you recompile it in .NET 2.0 (it will be
.NET 1.1 in the delivery for backward compatibility) you will have a
nice balloon that shows more information and whose content can be
customized via an environment variable [1].

--
Raphaël.

[1]
http://eiffelsoftware.origo.ethz.ch/index.php/MDC_InformationTips

[Ulrich Windl]

"llothar" <llo...@web.de> writes:

> Two month after the ISE deal it seems that the people who thought
> "Eiffel just needs a good open source compiler" are really wrong.

Is ISE 5.7 actually an open source compiler? I could not find the sources.

>
> Instead of seeing more newbie questions this newsgroup is more silent
> then ever. A full month of threads on one groups.google page is less
> then usual.

Wait, I'm going to make some noise soon ;-)

[Martin Piskernig]

Ulrich Windl wrote:
> Is ISE 5.7 actually an open source compiler? I could not find the sources.

Take a look at http://eiffelsoftware.origo.ethz.ch/index.php/Main_Page

Martin