Re: [Liberty-eiffel] PellesC for Adler

[Paolo Redaelli]

2013/11/12 Cyril ADRIAN <address@hidden>

Hans,

2013/11/11 H. Zwakenberg | Ocean Consulting GmbH <address@hidden>

I monitor SE TEST results on a regular basis and I really did not expect anything like a RC anytime soon. I suppose I expected any Eiffel community to have the 'pretense' that 'the rigor of Eiffel' would be followed in a release plan, i.e. that a release will only happen if a minimum standard was met.  Since I saw a lot of errors existing in the core language tests (like in the memory manager) and in some core libs as well (math and others) I was surprised by the RC announcement.

I would also like to point out that we do not have data about the releases of SmartEiffel. I bet a beer that the actual number of not-passed tests were higher.

Hans, striving for complete correctness is of course our target, but we must also take in account effectiveness and efficiency of the development. Until Google notice us and pay us full time to work on our wonderful compiler, libraries and wrappers (pun intended, expecially for the wrappers part 8-) we are working in our free time. This is of course not an excuse for doing a bad work, but as Cyril correctly wrote we will never release anything with a such tight requisites. Let's pick up the most subtle bugs, those in the garbage collector. We could release even with them because they refers to the new GC yet we still have the "old" gc that is working and is bug-free (AFAIK). I suspect that you are now labelling me as sketchy and imprecise (I can you shaking the head muttering "ah, italian are all like this!"; erhm. I'm Lombard 8-). Waiting to have polished it all, making every single test successful in all the compilation mode is by far much more that most languages could ever hope. It would be correct, but it won't be useful. I claim that this way of allocating the available resources is currently the best to achieve our goal: until we attract more people (i.e. volunteer work force) we cannot strive for complete correctness. A couple of quotes will explain this a lot better than my meaningless complaints: (a) `In contrast, one of my favorite mantras is "perfect is the enemy of good" [1] and the idea is that "good enough" is actually a lot more flexible than some idealized perfection.'         Linus Torvald [3] (b) "Had the first versions of other imperfect specifications such as TCP/IP waited ten years for those in disagreement to work out their differences, the Web might still be hiding somewhere in the back of inventor Tim Berners-Lee's mind" [2]         David Berlind August 7, 2003 on techupdate.zdnet.com Hans, please, keep asking for correctness though. I personally do appreciate your contribution, especially as constant reminds and recalls to correctness... [1] Actually the mantra is attribuited to Soviet Admiral S.G. Gorshkov. [2] http://techupdate.zdnet.com/techupdate/stories/main/0,14179,2914447,00.html [3] http://os.newsforge.com/article.pl?sid=05/06/09/2128249 "Linus compares Linux and BSDs"