On Thursday, 26 March 2015 at 22:05:27 UTC, Andrei Alexandrescu
wrote:
On 3/26/15 2:23 PM, Jake The Baker wrote:
But until
people stop acting like little children and think seriously
about
the problems, goals, and solutions then who the heck knows what
to do?
http://wiki.dlang.org/Vision/2015H1 -- Andrei
Yes, but you had to pull that up for me. Suppose there was a one
stop D shop that had all this stuff available, direct, and
immediate? A place that one could get there D fix in many ways.
A "facebook" for D coders, so to speak. All integrated into an
IDE proper. (collaboration, development, documentation, core
development, etc)
My main point about the "child" remark is that there are very few
people here(you, Walter, and a handful of others) actually doing
all the heavy lifting and most others(including myself) just talk
or do minimal work that ultimately won't lead anywhere(At least
any time soon).
What good is D if no one uses it? What good is it if all the
libraries used have poor or hidden documentation? What good is it
if the average noob gets board with it and gives up trying to get
it to work? (Not just D, which is, of course, easy to setup if
you don't mind doing a little text editing and file manipulation)
This is what I imagine. I know it can be done, that is not the
issue. This is all standard programming stuff. That is not the
point. The point is, to discuss ideas freely so the best
solution(Which is usually a combination of all relevant inputs).
1. Get up in the morning
2. Turn on computer.
3. Run the famous "D IDE"
4. Immediately on loading on presented with the main
window(eclipse, visual studio, etc...)
5. Apon loading I notice I have a msg from Joe. Seems Joe found a
bug and he thinks maybe I can fix it.
6. I download his project(all handled by the IDE, as it has a
build in "messenger" system).
7. Load up the probject, find the bug, can quickly match the bug
to the D internals since it's all there. (can jump to core source
code and look at it from any version. All the details of
versioning is handed behind the scenes. All I have to do is
"click on a button" to select which version I want to use.
8. So I write small patch quickly(just needed a null check or
what ever). I can cross reference the bug's(find out if anyone
else has the issue, the discussions involved, all at a drop of a
click.
9. I can submit the patch. Internally the IDE handles creating
the patch report, all the info(source code, test project,
info(line numbers, file names, etc)
10. Now, I'm done with the patching. I want to see what currently
has been done already in the last week. I can easily see the
"whats new" and extract relevant info by filtering. (I could
filter only on core topics, forum posts, etc)
11. By doing so, I find out that the bug I reported on as similar
bugs. I can further investigate that and see how it all relates.
I can pin, follow, start, (potentially) merge, etc topics, bugs,
patches, etc.
12. I can jump over to Walter's "D blog" and see what he is up to.
13. I can check out the main "Vision" which is interactive.
(click the year see the topics for that year, see the individual
progress(would be great to have a sort of project manager build
in).
14. After I have all that fun(With the world of D at my finger
tips in the IDE) I can simply open up some project I've been
working on.
15. The IDE has complete cross-referencing of documentation built
in and very fast. It has full intelligent support. All the
standard IDE stuff.
16. The only thing really difference from this point on is that I
can easily send people projects(simply "send project o "Joe,
Mike, Other" where Joe and Mike are my "buddies" that have joined
my project)... all standard collaboration stuff but quick access.
They can even "remote view" the project if I allow them.
17. e.g., imagine how much easier it would be for people to solve
problems if they could easily "remote view" into a project?
etc.. etc.. etc..
18. Of course, If I wanted to switch to C++ or (D++?) I could do
so at the drop of a hat. The IDE should be somewhat language
agnostic. (this would allow you to do benchmarks and comparisons
easily between different languages... easy Go and D.
19. Not only that, Suppose I have a bug in my code. The IDE can
automatically cross-reference the error code with all bug
reports, forum posts or whatever. I can show anyone who is
working on the patches, allow me to join the discussion or
whatever. Similar to Git in many ways. The point is "quick". I
don't have to open a browser, go to my favorite search engine,
sift through hundreds of irrelevant posts just to find that Bob
had the same bug and no one is working on it. Might take me 10
mins, or 10 hours, or even 10 months to figure out what is going
on or to fix it.
The point being, we know that having this so called "perfect IDE"
would solve many problems.
The real question is it worth doing? Will having such a thing
ultimately bind all the "looseness" that the D community has into
one well oiled machine?
If done well, I believe it would. (I'm not saying my vision of
the IDE is best, but I'm simply talking about a more organized
and unified way to "interact" with D/community.
I'm very particular to the one-stop-shop type of IDE simply
because I find browsers to be very limited, too generic, and very
inefficient to do what I'm talking about. (Those that think
wasting 10 mins or so on something that should take 1 min is not
a big deal... add up all the time you have wasted in your life
waiting for the little spinner/hour glass and I think you'll find
you've wasted a lot of time. (At least a vacations worth... and
how many of us would turn down a free vacation?)
