October 23, 2003 Re: phobos docs | ||||
---|---|---|---|---|
| ||||
Posted in reply to Matthew Wilson | Matthew, I take your point about contractions in general, but I think lib is so ubiquitous in other languages it has become engrained in programmers minds as meaning library. My first choice is "dlib," with the C library called "clib." My second choice would be "dsl," with the C library called "csl." My third choice is an uppercase "D", with the C library called "C." I think the single lowercase character "d" is asking for trouble. I don't think this will prevent compilation, but dlib itself ;-) uses variables called "d" in places (date.d, math.d, math2.d, path.d). It would seem very arbitrary to me if I saw a line of code like "int a, b, c, e, f" and I think it would flummox beginners to find that you have to do this. From, Rupert "Matthew Wilson" <matthew-hat@-stlsoft-dot.-org> wrote in message news:bn7ujf$2nij$1@digitaldaemon.com... > But its a contraction, rather than an initialism. Why not dli? or dlibr? See > what I mean? > > I'd rather dsl, for D standard library, than that, but I'd rather just "d" > > "Helmut Leitner" <helmut.leitner@chello.at> wrote in message news:3F976C76.5BCF6E5C@chello.at... > > > > > > J C Calvarese wrote: > > > > > > Matthew Wilson wrote: > > > > Ok. Here's where I see the situation. We have several options > > > > > > > > 1. D > > > > 2. d > > > > 3. phobos > > > > 4. std > > > > 5. stdd > > > > 6. lang > > > > > > 7. dee > > > 8. drt (D RunTime) > > > 9. dlib <==== > > > > That's my favorite, because that's what it is, > > the "D library", in short. > > > > -- > > Helmut Leitner leitner@hls.via.at > > Graz, Austria www.hls-software.com > > |
October 23, 2003 Re: phobos docs | ||||
---|---|---|---|---|
| ||||
Posted in reply to Matthew Wilson | > > > 9. dlib <==== > > > > That's my favorite, because that's what it is, > > the "D library", in short. Matthew Wilson wrote: > > But its a contraction, rather than an initialism. Why not dli? or dlibr? See what I mean? No, there are two ways to argue for "dlib". First "lib" is the typical library file extension / abbreviation, so its the shortest way to get all the semantics in. Second, just test programmer expectations. Ask ten people which abbreviation feels most natural for "D libary" and offer them your choices, for example - dli - dlib - dlibr - std - stl or whatever you like. Make them use the name in sentences like "Where can I download the new ... version?" "Is this a ... module?" "Is there an online ... reference?" "Is ... covered in the book 'D programming'?" before they make their decision. What is easiest to talk about? What feels most natural? Don't let them make quick decisions. Let them think about it, ask them for their final decision after a day. Then follow the majority of programmer expectations, they will be right. I'm pretty sure that at least 7/10 will vote for "dlib" and no other choice will get more that 2 votes. Anyway it seems quite clear to me, that neither "d" nor "D" are valid choices. Because "Where can I download the new D version?" "Is this a D module?" "Is there an online D reference?" "Is D covered in the book 'D programming'?" doesn't clearly refer to the library. You would always have to talk around it, like "Is the D-library covered in the book 'D programming'?". Here again comes "dlib" as the most natural abbreviation. Of course "dlibrary" would also be good, but I know that Walter would never, never, never accept such a long name. -- Helmut Leitner leitner@hls.via.at Graz, Austria www.hls-software.com |
October 23, 2003 Re: phobos docs | ||||
---|---|---|---|---|
| ||||
Posted in reply to Helmut Leitner | "Helmut Leitner" <leitner@hls.via.at> wrote in message news:3F978DCF.8218ED54@hls.via.at... > > > > > > 9. dlib <==== > > > > > > That's my favorite, because that's what it is, > > > the "D library", in short. > > Matthew Wilson wrote: > > > > But its a contraction, rather than an initialism. Why not dli? or dlibr? See > > what I mean? > > No, there are two ways to argue for "dlib". > > First "lib" is the typical library file extension / abbreviation, so its the shortest way to get all the semantics in. > > Second, just test programmer expectations. Ask ten people which abbreviation feels most natural for "D libary" and offer them your choices, for example Well, I was using reductio ad absurdum, so was not suggesting dli or dlibr as they are both unambiguously offensive. ;) > > - dli > - dlib > - dlibr > - std > - stl > > or whatever you like. Make them use the name in sentences > like > "Where can I download the new ... version?" > "Is this a ... module?" > "Is there an online ... reference?" > "Is ... covered in the book 'D programming'?" > before they make their decision. > What is easiest to talk about? > What feels most natural? > Don't let them make quick decisions. Let them think about it, > ask them for their final decision after a day. > > Then follow the majority of programmer expectations, > they will be right. I'm pretty sure that at least 7/10 > will vote for "dlib" and no other choice will get more > that 2 votes. I just hate dlib. zlib, xmllib, etc. are all specific libraries. dlib just stands out. Yes, I know libc is the C std lib on UNIX. > Anyway it seems quite clear to me, that neither "d" nor "D" > are valid choices. Because > "Where can I download the new D version?" > "Is this a D module?" > "Is there an online D reference?" > "Is D covered in the book 'D programming'?" > doesn't clearly refer to the library. > You would always have to talk around it, like > "Is the D-library covered in the book 'D programming'?". > Here again comes "dlib" as the most natural abbreviation. The d/D.* libraries we are talking about are not just standard libraries of D, it/they also contain(s) aspects of the language itself, in the same way that the C runtime library, the C++ standard library do, and the basic Java libraries do. Hence, your characterisation is somewhat flawed. Maybe what's been bandied about as the standard library should be explicitly segregated into language core and libraries, although we may find that's not as easy to do as it sounds. Gah! It's all so inconsistent. I'm afraid that I still think the language is called D, so the module should be d, just as with Java/java. Sorry. It still seems a minor imposition to prevent people using d and D as identifiers in a language called D. > Of course "dlibrary" would also be good, but I know that Walter would never, never, never accept such a long name. Maybe we should get him to change the language name back to mars, and use that. :) |
October 23, 2003 Re: phobos docs | ||||
---|---|---|---|---|
| ||||
Posted in reply to Matthew Wilson | Who knows somebody in marketing? Ask them!! I vote for dsl (D Standard Library). Sean "Matthew Wilson" <matthew-hat@-stlsoft-dot.-org> wrote in message news:bn75am$1lou$1@digitaldaemon.com... > Ok. Here's where I see the situation. We have several options > > 1. D > 2. d > 3. phobos > 4. std > 5. stdd > 6. lang > 7. some other multi-letter prefix |
October 23, 2003 Re: phobos docs | ||||
---|---|---|---|---|
| ||||
Posted in reply to Patrick Down | Great minds think alike. ;) Sean "Patrick Down" <pat@codemoon.com> wrote in message news:Xns941CDE50916BApatcodemooncom@63.105.9.61... > 10. dsl ( D standard library ) |
October 23, 2003 Re: phobos docs | ||||
---|---|---|---|---|
| ||||
Posted in reply to Matthew Wilson | "Matthew Wilson" <matthew-hat@-stlsoft-dot.-org> wrote in message news:bn831f$2u7f$1@digitaldaemon.com... > > Then follow the majority of programmer expectations, > > they will be right. I'm pretty sure that at least 7/10 > > will vote for "dlib" and no other choice will get more > > that 2 votes. > > I just hate dlib. zlib, xmllib, etc. are all specific libraries. dlib just stands out. Yes, I know libc is the C std lib on UNIX. You're right... it seems there's precedent for calling it "libd". > I'm afraid that I still think the language is called D, so the module should > be d, just as with Java/java. Sorry. It still seems a minor imposition to prevent people using d and D as identifiers in a language called D. I would hate to lose d and D as identifiers. Well they wouldn't be lost, as you can do .d.win32.registry, but they would be a rather common source of confusion. > Maybe we should get him to change the language name back to mars, and use that. :) Then we could leave the library named phobos, at least. ;) Sean |
October 23, 2003 Re: phobos docs | ||||
---|---|---|---|---|
| ||||
Posted in reply to Charles Sanders | "Charles Sanders" <sanders-consulting@comcast.net> wrote in message news:bn75j8$1m5u$1@digitaldaemon.com... > Also, what is win32 / linux specifc stuff doing in the library ? Shouldn't > phobos just contain platform indepenent functions, like C's ? That's a really good point. I agree and think it'd be best to keep the win32 and linux modules in some other namespace besides the standard library. "platform.win32.*" and "platform.linux.*". Maybe also could have "vendor.digitalmars.*" etc for compiler-specific or vendor-specific functions. It'd be best to keep the core standard library unpolluted with platform-specific functions. Either do least common denominator, emulate the functionality, or leave it to other libs. Sean |
October 23, 2003 Re: phobos docs | ||||
---|---|---|---|---|
| ||||
Posted in reply to Matthew Wilson | "Matthew Wilson" <matthew-hat@-stlsoft-dot.-org> wrote in message news:bn781a$1pal$1@digitaldaemon.com... > Now maybe COM is not the best example, because the likelihood of our doing a > Linux version is pretty slim (although maybe that's not the case for Mac!?), > but I'm sure you can extrapolate to other things. Networking, for example. We'd be idiotic to not support using IO Completion ports on Win32 implementations, but this doesn't work for UNIX. I fully expect the runtime library to be using os-specific functions under the hood. > > Shouldn't phobos just contain platform indepenent functions, like C's ? > > Do you really want to hold up the C runtime library as an example of the best libraries there can be? Surely not. Nobody said "the best", did they? But they did some things right. Sean |
October 23, 2003 Re: phobos docs | ||||
---|---|---|---|---|
| ||||
Posted in reply to Sean L. Palmer | > > Now maybe COM is not the best example, because the likelihood of our doing > a > > Linux version is pretty slim (although maybe that's not the case for > Mac!?), > > but I'm sure you can extrapolate to other things. Networking, for example. > > We'd be idiotic to not support using IO Completion ports on Win32 implementations, but this doesn't work for UNIX. > > I fully expect the runtime library to be using os-specific functions under the hood. But what if those under-the-hood things are also to be provided as external interfaces? |
October 23, 2003 Re: phobos docs | ||||
---|---|---|---|---|
| ||||
Posted in reply to Matthew Wilson | I prefer D over d, and over most of the other acronyms. Plese not anything on the "sdl", "dsl", and so on theme. If you go that way, acronyms would have a much higher recycling rate than code!
What i find neat is "lang" or something alike. :)
-eye
Matthew Wilson wrote:
> As you can see from the group, everyone prefers d to D, irrespective of
> whether they prefer mutli-letter over d/D.
>
> Can we start by changing D.win32.registry to d.win32.registry for 0.75, and
> move from there unless and until we encounter any bumps in the road?
|
Copyright © 1999-2021 by the D Language Foundation