Re: Re[2]: Is Smart Eiffel ready for large-scale application ?

Hi Lothar,

 A friend told me that VisualWorks has a license model where the tool can be used for 
free (full version, not a lite version) and when you deploy your app, you pay them a 
percentage. I think you can negotiate, and it depends on how much Smalltalk code you 
have (versus say how much in terns of C prims etc). I think I recall hearing 6%.

 In other words, maybe you can propose something similar to ISE ? Basically a "you make 
money if I make money". I agree that for people trying to bootstrap their small little 
company, US$5,000 is prohibitive. If you need more people to lobby, let me know, I can 
join the list.

 Using their free version just in the development phase (because you get incremental 
compilation, debugging etc) and SE for deployment does not work well because of the 
differences in both implementations (even if you use GOBO end EPOSIX). As you said, 
Thread etc is still problematic, plus even basic language features. Maybe if we had a 
standard and implementors followed it, but I don't feel optimistic about it. In some 
sense, it feels as if Eiffel was being invented now, with discrepancies across 
implementations because of a lack of agreement at the early phases. Again, maybe in a 
few years, with standards... Maybe...

 Besides, doing that may violate the terms of the agreement. Even if it were ok (legally 
speaking), I doubt that ISE would be happy to have people using their free version for 
development and then see it being dumped for SE towards the end of the cycle. They would 
make $0, and "They have to eat too". So, no real incentive to be compatible with SE. 
That's why I think a business model of "we make money if you do" fits well with 
individuals (Shareware developers) and/or small companies and start-ups. If you are 
hacking in your spare time trying to get something started, US$5K for the IDE is 
something hard to justify. For bigger companies (say a Telco), $5K is peanuts if you can 
demonstrate the increase in productivity. Most "Enterprise FOOBAR" IDEs cost a lot of 
money too.

 I recently installed ISE's IDE for the 3rd time, for a 3rd look, and this time it was a 
lot easier, thanks to their videos/tutorials online ( 
http://www.eiffel.com/developers/presentations/ ). But I noticed that in the free 
download you get no profiler and no BON. Hard to justify using it, when I can get so 
much for free around Eclipse, for example. Incremental development a la Smalltalk right 
there. Profiling, right there. And so on. For $0. If only GCJ worked properly, producing 
small&fast executables... <sigh>

 Back to the language options. There seems to be a lot more stuff for Eiffel and Modula-
3 than for D, I think. It would be hard to justify choosing D at this point (how many 
books ? how many people in the community ? Anything similar to GOBO, EPOSIX, etc *fully 
working* ?), Same sort of argument for Modula-3, I guess.

 Personally I think that Lothar's project is a great test case for the language & 
libraries & compiler. It is not some batch-mode program (compile this into that), but a 
fully interactive application that has to be competitive with state-of-the-art. Not just 
a command-line tool or some Mickey Mouse GUI quickly put together by command-line lovers 
(Linux is full of these). I strongly believe that you have to "eat your own dog food" as 
they say (SE team using SE to implement the compiler, etc), but you also need a 
few "real customers" in order to "keep you honest" in terms of your assessment of what 
really matters and how far off you are.

 And I am not even talking about something extensible via plug-ins like Eclipse. 
Lothar's IDEs do not aim at being as flexible and general-purpose as Eclipse. Doing his 
IDEs in Eiffel should be doable in a productive manner, suffering less than in other 
languages. Otherwise, you have a problem with your language or IDE or class libs. Which 
Eiffel is needed, however, is another story... ISE Eiffel has a neat multi-platform IDE 
written in ISE Eiffel. Eclipse is a very nice Java IDE written in Java (and works as IDE 
for many other languages too !!!). Again, Lothar's IDE is a great test case for the 
tools used, so I truly respect his insights. I've had a similar SE experience myself, 
although I was careful to "fail early" in my evaluation of the tool (started with the 
risky parts first, and saw how far I could push with SE). It's been years now, and the 
problems are still there. Sigh. For now, I keep wathcing...


marcio


On Apr 9 , Lothar Scholz <[email protected]> wrote:
> 
> Hello Marcio,
> 
> Saturday, April 9, 2005, 4:28:53 PM, you wrote:
> 
> MM> Hi,
> 
> MM>   Why are you considering SmartEiffel instead of Modula-3 then ? I'd be interested 
to
> MM> know.
> 
> MM>   Also: how about ISE Eiffel versus SmartEiffel ? Does it
> MM> have the same negative points, 
> MM> Lothar ? I mean, it has incremental compilation, Threads, an IDE with Debugging 
support,
> MM> etc. No ?
> 
> 5000 US$ per license :-(
> 
> Threads are AFAIK not working with shared memory, they use like TCL
> separated memory areas.
> 
> If threads and dll's are not a must have feature on your list then i
> would consider an ISE Eiffel environment for development and a SmartEiffel
> compiler run for deployment as optimal.
> 
> 
> -- 
> Best regards,
>  Lothar                            mailto:[email protected]
> 
>