I missed a word or two out. I meant a constraint on a generic parameter to a class. E.g. class MYCLASS [G -> ARRAY [STRING]; H -> SOME_OTHER_CLASS [G]] end Here SOME_OTHER_CLASS [G] can have the constraint in it's own text, but MYCLASS has no way to express it. If we had true dependent types

Such a constraint goes to a precondition (or another assertions) rather than to a type declaration: foo (data: ARRAY [...]) require data.count = 7 Alternatively, one can have a descendant/client of an ARRAY with an invariant that there are exactly 7 elements. But, maybe, I'm

when it is the easiest path to follow, in whatever their current task is. - thomas

Another limitation I see with DbC in practice is that I can't add a constraint (e.g. ARRAY must have exactly seven elements) to a generic parameter. On Sun, 19 May 2019 at 10:48, Arslan, Volkan wrote: > Can someone familiar with "Dependent Types" [1] of mainly functional >

10 deaths a day. I had no idea it was so many. I’ve only ever caught the train once in Mumbai after spending the day with Freddy Mercury’s best friend as one does. I think now they are getting automatic doors. You could probably deter the Muslims from crossing the tracks by smearing the tracks

On 19 May 2019, at 19:19, Woland's Cat wrote: this is also how standards development at ISO and HL7 work: everything is consensus-based +loudest voice wins (well ISO also has bloc-voting like Eurovision - to create technical standards!). Problem is the consensus usually

Great (and sad) observations, all. To challenge another's belief system is typically futile. That change must come from personal realization, not rational discourse. Changing the behavior of a population, on the other hand, is much more feasible (for better of worse, as seen in recent elections,

That sounds quite interesting: to get an idea do you have an example for such a function?

On Sun, 19 May 2019 at 10:48, Arslan, Volkan wrote: > Can someone familiar with "Dependent Types" [1] of mainly functional > languages such as F* [2] and Idris [3] shortly compare / assess it with > "Eiffel's strong type system in connection with Design by Contact"? > I'm

I have given a presentation/demonstration to my company a few years ago, doing just that (to no avail - they are dropping Eiffel in favour of C# - not looking forward at all). I could try to repeat that here from memory, but it would be much better for you to watch this video by Edwin Brady, the

this is also how standards development at ISO and HL7 work: everything is consensus-based +loudest voice wins (well ISO also has bloc-voting like Eurovision - to create technical standards!). Problem is the consensus usually only has a weak relationship with truth or reality, unless it's only

Can someone familiar with "Dependent Types" [1] of mainly functional languages such as F* [2] and Idris [3] shortly compare / assess it with "Eiffel's strong type system in connection with Design by Contact"? Some consider "contracts as a restricted form of dependent types which can be checked

Hi all, It is an interesting discussion. I think that Eiffel fans are more professionals than marketers. To boost something, one needs an association with a buzz word, positioning the language as "the solution for XXX" C - fast and embedded, JAVA - virtual machines, Android - smart phones,

We should move to distributed production, not centralised. Problem is the centralised producers don’t want to lose their cash cows. As all computing people know, power (nice pun!) is built up by the combination of millions of little computations. We need to handle water and waste the same way.

Ian Joyner wrote: - I heard a radio program the other day researching how to change people’s opinions. This has certainly been more than demonstrated in this country yesterday with the election here. It showed that all the party that won had to do was to put around a lot of lies,

I have been following this thread with considerable interest. Thank you! This photo -- of a poster I saw in a developer cubicle of a client a few weeks ago -- epitomizes attitudes I heard you describe. I had to share it! I, too, would like to see Eiffel work on the full stack with other

On 19 May 2019, at 02:34, r...@amalasoft.com wrote: an area where things like garbage collection were seen as the spawn of Satan Countering widely held beliefs like this is difficult. Essential bounds checking for correctness and security put down as training wheels. Nefarious notions of

Where to start? Yes, it was the EWD611 paper I read before when I saw the title "On the fact that the Atlantic Ocean has two sides.” What I think is behind this paper is the analysis of why the message about system and software correctness was not getting through. Whether the paper is right or

Garbage collector accelerators are on their way. See IEEE Micro Vol 39 Issue 3 may-June 2019, p 38 From: eiffel...@googlegroups.com On Behalf Of Woland's Cat Sent: Sunday, 19 May 2019 03:12 To: eiffel...@googlegroups.com Subject: Re: [eiffel-users] Quora - What is

On 18/05/2019 17:34, r...@amalasoft.com wrote: All nicely put. As much as I'd love to be able to web-app my way to happiness with Eiffel, I worry that only a handful of the faithful would care. It might be a better use of resources to build on what Eiffel does best, to create conspicuous value

All nicely put. As much as I'd love to be able to web-app my way to happiness with Eiffel, I worry that only a handful of the faithful would care. It might be a better use of resources to build on what Eiffel does best, to create conspicuous value in contexts where that matters (objectively, and

Eric, Spot on! Apparently, in this on-going software tools environment, beta-test-by-consumer plus "mostly-works" is good enough. People have an emotional toleration for problems and issues as long as they see A) Immediate benefit B) Acceptable Work-arounds C) Bugs being steadily fixed and

On 18/05/2019 13:51, r...@amalasoft.com wrote: Thanks Ian. Good stuff. Perhaps we (and I very much include myself in this) techno-geezers need to step back from our walkers for a moment and smell the . That I believe is the scientifi

Thanks Ian. Good stuff. Perhaps we (and I very much include myself in this) techno-geezers need to step back from our walkers for a moment and smell the . The state of computing might have differed with geography and politics, at some

On 18 May 2019, at 01:06, Larry Rix wrote: Without a doubt, Eiffel is a European child! As I learnt from reading Dijkstra, there is a big difference between the two sides of the Atlantic in thinking about computing. American computing is more about hardware and