26

Re: The best language on light

Hello, kochetkov.vladimir, you wrote: KV> It is not necessary so, about colleagues One of these "blockheads" after JetBrains was arranged to us and, for ~1.5 years of the operation, no censures to its efficiency arose. From these one and a half years, months ~8 I was it , .. I know, about what I speak. It thus that the project over which it all this time works, in places is more difficult than the compiler or language  will be. Yes, Stas the programmer from god. A little bit, only, on requests of users spits (from a series "from me a packet left") and, unfortunately, recently to me at all does not help. If over the project worked would like it we already finished it. KV> It turns out, what quite to itself the adult person kicked three years , and then suddenly became reasonable, yes so, what the second year cannot already release? "I do not trust!" () Yes at them constant idea that they on our place all would make that not so, and it is a hundred times easier and better. It if you will plunge, all complexities of tasks are clear. And always it is easy to argue on another's operation from outside. We it and the country would control is better. And in football would play (despite beer stomachs) is better. KV> Result - in repository Nitra and it definitely differs from zero on some orders. Thanks for support. KV> if speech about Nemerle it (how much I remember) the JetBrains interested only, as means, but not the purpose. The purpose was development by guys  which could, on the one hand - to occupy a niche fashionable now Haxe-like , and with another - to give the chance to JetBrains to build around all it an infrastructure on the basis of already available products. They them did not interest at all. They once looked at its demonstration. I showed that all their prejudices concerning macroes are groundless, but interest there not especially arose.  comparing with the language decisions was interesting (author Kotlina) only. He to me told about plans to make support of multiple inheritance in Kotline. I doubted simplicity of implementation of this business. He told that make, but and did not make. And we here, with Stasom, made. In Nitra for nuclear heating plant multiple inheritance is supported. Very conveniently it turned out. KV> the reasons on which all happens how happens - set and guys concern far not all from them, IMO. So. Likely we not always gave all the best on 100. But three years to give all the best on 100 % simply is not real on end. And likely we threatened on too big task, underestimating its complexity. But I can prove in details what to make the correct expanded language on another practically not probably. Well, unless with much  a command and  the MSEC.

27

Re: The best language on light

Hello, kochetkov.vladimir, you wrote: KV> I to that development by them first of all integrations with VS (beaten Nemerle nails to an ecosystem.NET) was an error. The first that it was necessary to make is to hammer on Polish proof-of-concept and to untie language from.NET (from a word absolutely). And only after that to do integration with IDE and not the fact that VS would be the best choice. Result of this miscalculation was loss of the considerable mass of potential users (the part from which, quite could and adjoin the project, in due course) and stagnation in one platform on ten years. But then all here actively heated for.NET and the heresy written above, at the best, simply would deride. Integration now lives in server process. The client code is minimum. To transfer it on another IDE, let even based on Java or JavaScript it is already completely not difficult. Itself  (languages based on it are more exact) it is possible to compile under Core what to provide portability on all reasonable platforms. KV> it not so. Concept Nitra (then still it was called N2) started to study long before the beginning of history from JetBrains. And guys initially there took on development language , instead of Nemerle. : https://blog.jetbrains.com/dotnet/2012/ … d-nemerle/ so. And the concept of Nitra has been adapted so that just to interest any office as just Nemerl anybody did not want to support. Personally I was put to write qualitative expanded language to this project. Whether Nemerl it or SharpaEks will be important. For today we can generally at once a little  assort in uniform nuclear heating plant with uniform typification. So there are no problems to make support in one language and expanded Sharpa, and Nemerla 2, and even absolutely a modern language if it is necessary to someone. Actually we solved all technological problems that were at us in Nemerle. The only thing with what can be it turned out not so well is language of display. It not so intuitive and explicitly excessive for  tasks. But this problem it would be possible to eliminate  implementations of that expanded language. Simply to make   macroes when syntax is set round nuclear heating plant, instead of it is displayed on nuclear heating plant. Unfortunately it not the general-purpose decision, but for 80 % of its cases will be enough. So also this problem it is known how to solve. From the point of view of portability it is necessary to do two steps: 1. To translate Nemerl to Nitra in that volume in which Nitra uses Nemerl. 2. To do correct . First of all it makes sense to try to use Roslyn-ovsky  for code generation. It gives the chance to code generation for.Net Core and portability. At the expense of the server we can hill any IDE in reasonable periods. Further it is possible to make  for Java and-or LLVM. Then there will be generally the full portability. Even on mysterious platforms. But all is forces and time. KV> this article has been published right after how guys appeared under a JetBrains wing. Over Nemerle it there anybody and never paid operation. From guys waited for a bit different result and to reproach them that they aspired to achieve it is at least strange All so. Simply IT wants to be right, to abuse us and thus not to help. It would be fairly simple to tell that it does not have time and desire on this project. But what for justifications are necessary. Here also any mythical purposes of DzhetBrejns and zero result are invented.

28

Re: The best language on light

Hello, Qbit86, you wrote: KV>> But then all here actively heated for.NET and the heresy written above, at the best, simply would deride. Q> And now heat, and correctly do. I during the last years heard time and again that.NET for our project - a minus not giving to use it in someone's operation.

29

Re: The best language on light

Hello, kochetkov.vladimir, you wrote: KV> Losing thus for Nemerle where the big user mass, than ever was round all ecosystem.NET. What in it of correct? It can and is correct, but it not so is simple. It demands forces. And to make it also technically correctly without making all that we made now - even more difficult. The probability to receive as a result  a product like Nemerla 1 is great.

30

Re: The best language on light

Hello, kochetkov.vladimir, you wrote: KV> And to whom it is now necessary Nemerle? You seriously believe, what it is still capable to turn something? I am serious, without a dirty trick I ask. For those 5-7 years since the moment when the last assumptions about a demand and mission of this language became, all remaining world as though on a place did not stand. I am assured that language with expanded syntax will be claimed more and more. Complexity of a software increases and simply to keep a tight rein on it it is necessary DSL-it decisions. A minus of Nemerla that it differs from Sharpa and on it  it is necessary more forces. Plus a huge question as implementation. KV> here, right now, it is watched rather strong  round microservices,  and isomorphic applications,  with smart contracts and machine training. Poor (to measures Nemerle) JavaScript already now climbed through in all it, except unless ML and quite successfully there it is used. What now Nemerle can oppose to this language, even if receives made ? Yes all too most only is more convenient and with possibility to solve specific targets making DSL-chik. Only in Nemerle for this purpose direct means were not. In Nitra we just thought over this question. For DSL it is possible to do specialized nuclear heating plant which by the possibilities does not differ from built in in Nemerl 1.. More precisely in Nitra this same. Means of the description of nuclear heating plant of Nitra allow to describe any language constructions and extensions. The language construction can be described as it is and  receiving the first class essence.