Matthew Wilson
Posted in reply to Andrew Edwards
| > Any hoo! I wanted to revisit the subject of The D Journal. Matthew, I know that with your book, multiple (?) columns in CUJ, and whatever else you
may
> have on your plate, you are a very busy man. Do you think however, that
you
> may be able to spare some time to resurrect this baby from it's year-long slumber?
>
> I think it's time we realized what we have in front of us and start
telling
> the world about it!
Well, it's certainly still a long-term intention to do this. D's getting a bit more coverage. There's my recent WDN C# perf series, which may have two more parts within the next few months. I've an article in next month's (October) CUJ discussing Identity and Equality in various languages (C++, C#, D, Java, Python and, for a laugh, VB.NET!), although I don't exactly say I agree with the D way of doing things - still, there's no such thing as bad publicity, eh? I'm aware of a couple of other things (by me and by others) brewing so expect to see quite a few things over the next six months.
As for the Journal itself right now, I can't see it getting to a first issue before next year. There're a number of factors involved:
1. there's a huge amount of logistical effort involved in the preparation and upkeep of the website. I do not have the skills, and will never have the time to do this. Alix Pexton was going to do this when we discussed it last year, but I've not heard from Alix for sometime, and one can hardly blame him for failing to stay in a state of high-preparedness for 18months without having anything to do. :/
2. there was a fair bit of interest when the idea was first mooted, and some good info put forth, but it was all from about four or five people. Writing's not easy, and people are naturally shy about doing so. I will say that when we do the journal, I'll be happy to expend efforts to helping people in this regard, but fundamentally it still requires people to put in the hard yards and knock out the words.
3. the financials. For my part, I have to say that at the moment I am truly living the life of the penurious author, so am subconsciously aimed towards (non-D) subjects that are likely to sell to a paying publication, and also if I get an idea for a really good article on D it's still likely to be aimed at DDJ or CUJ for the same reason. Until I get past the book writing and back on to some proper paid work (or the wife win's Lotto!) this is likely to continue to be my focus. Sorry if that sounds mercenary and/or selfish, but something's got to pay the mortgage! ;/
4. time. As you said so above, writing a book (not to mention all the other stuff I've insanely got myself doing) just eats up your whole life. (I have to remind myself every few days to go and do an hour on the bike just so I can remember what it used to be like to be fit!) Doing the D Journal will be a project in itself until the momentum builds, and I don't have the time to spare on it at this point. Note that I'm not suggesting that I'm the only bottleneck here; Walter's up to his neck working on the implementation, so it'd be a similar story trying to get him to invest serious time in the journal at this time. I'm sure everyone else involved in this ng has a real-life to distract them, so until such time as a concentrated few can put in a large effort to complete the structure, and provide material for the first few issues, it'll continue to languish.
5. the preparedness of the language, libraries, tools, and the lack of finished non-trivial D projects. This may be a point on which others would choose to disagree, but I think it's just too early. We need to see collections, properties, exception-handling, string/localisation (there's been no argument about this for a while ...) and perhaps generics done to a complete degree, and tried and tested a bit. We need some libraries. (Putting my money where my mouth is, I've committed to writing the Win32 registry library for Phobos as soon as the foreach stuff is done). We need exception-handling agreed and implemented (though I think we're nearly there). And I confess I have no idea where the templates/generics issue is at the moment, but from what people have said, we still don't have implicit instantiation, so to my mind there's still a hill of work to be done there. And then there's the fact that we've no serious completed & polished projects can't be ignored. (Again, when I've got the reg stuff done, I plan to write a layer for the registration of COM components, and then perhaps write a shell extension in D. It would be nice to have a suite of COM components, and perhaps some wizards, such that D was better at building COM components than C++/ATL.) And though DIDE and DIG and such things are great starts, they're not (AFAIK - please correct if I'm wrong) of a maturity where Walter'll be taking them to an expo in the next few months, I think.
I'm not being negative, and I hope I'm not coming accross as such, but I just think it's a case of patience. I know that I've absolutely no interest in writing any articles involving parts of the language which may yet undergo significant refinement, and I would assume most would agree.
So my position is, I very much look forward to a time, say 2 years hence, when the D Journal is running, and if I am in a position to be editing it (and my services as such are welcomed by the D community) that would be fantastic. For my part I am doing what I can to develop D (writing libs and spouting lots of opinions <g>) and to promote D (the articles, and featuring it as one of the languages in my upcoming CUJ integration column), and I can see no reason why, with similar modest commitments from all concerned (except Walter, of course, we need a massive commitment from him for the foreseeable future, methinks) we can't get to the mutually desired point. I just think that we're not at that point yet.
Feel free to talk me round ...
Cheers
Matthew
|