www.digitalmars.com         C & C++   DMDScript  

digitalmars.D - Full D IDE in the browser - something to pursue?

reply WebFreak001 <d.forum webfreak.org> writes:
https://github.com/Pure-D/code-d/issues/393

what do you think of this idea? worth checking out in the future? 
not important at all or high priority?

With [hatf0 showing DMD-FE being able to run in 
WebAssembly](https://forum.dlang.org/thread/uccwuheorntxtashq
nt forum.dlang.org) I think this could be a very real possibility and we could
be pretty close to it being doable already.

Instead of trying to port DCD could also try to immediately use 
DMD for auto completion, though that's a whole other project that 
would bring other benefits as well.
Jan 11 2022
next sibling parent reply russhy <russhy.rsy gmail.com> writes:
Other than syntax coloring for vscode-web i don't think having 
the full suite is necessary

I'd rather see more investment being made in improving DCD, there 
still lot of areas where it just is bad..

- ``auto data = cast(Data*) myPtr;`` this is basic code and yet 
DCD can't figure out what ``data`` is


- ``T get_x(T)(); auto x = get_x!float;`` DCD can't figure out 
it's a float


- ``const(int) myInt;`` DCD can't tell me it's a const

There are many more little things like that that just doesn't work

For a language that rely heavily on these features, it should be 
supported

And i don't think the D audience is using web based IDE, it's a 
system language


That would be a smarter time investment, getting basic things 
supported is long overdue


Because right now it's pretty rough compared to the competition 
(rust language server with analyzer)

We need retain people in the language, tooling is still rough 
decades later, that drives people away..
Jan 11 2022
parent reply Dennis <dkorpel gmail.com> writes:
On Wednesday, 12 January 2022 at 00:08:02 UTC, russhy wrote:

 I'd rather see more investment being made in improving DCD, 
 there still lot of areas where it just is bad..


What's stopping you from working on it?


 And i don't think the D audience is using web based IDE, it's a 
 system language


I agree, I wouldn't be using it myself, though it would be a cool 
way to try out the language online, and it would be a cool 
demonstration of D being capable of targeting WebAssembly.
Jan 12 2022
parent reply WebFreak001 <d.forum webfreak.org> writes:
On Wednesday, 12 January 2022 at 09:09:32 UTC, Dennis wrote:

 On Wednesday, 12 January 2022 at 00:08:02 UTC, russhy wrote:

 I'd rather see more investment being made in improving DCD, 
 there still lot of areas where it just is bad..


 What's stopping you from working on it?

 [...]


well russhy did plenty of improvements to DCD, I think it's 
mostly nobody reviewing them or having time for it is delaying 
the contributions quite a bit.

I think it would be best to focus on making a DMD frontend based 
DCD replacement though.
Jan 12 2022
parent reply Dennis <dkorpel gmail.com> writes:
On Wednesday, 12 January 2022 at 09:38:28 UTC, WebFreak001 wrote:

 well russhy did plenty of improvements to DCD


That's great to hear! Btw, russhy, sorry for the implication that 
you weren't working on it. I'm genuinely curious what the barrier 
is for D users to implement the DCD features they want (I haven't 
looked much at the code myself).


 I think it's mostly nobody reviewing them or having time for it 
 is delaying the contributions quite a bit.


That's unfortunate. I started watching GitHub activity on the 
pinned dlang-community repos, intending to get more involved 
there.
Jan 12 2022
next sibling parent rikki cattermole <rikki cattermole.co.nz> writes:
On 12/01/2022 11:02 PM, Dennis wrote:

 I think it's mostly nobody reviewing them or having time for it is 
 delaying the contributions quite a bit.


 
 That's unfortunate. I started watching GitHub activity on the pinned 
 dlang-community repos, intending to get more involved there.


The biggest problem we have over on dlang-community is having people 
capable of reviewing stuff.

While I can pull PR's and grant rights, unless I know a PR is good (e.g. 
simple or I know the code base) and know the person is good for the 
rights, I can't go round doing these things.

The more people that get involved the better. Even if all they are doing 
is fixing spelling mistakes in comments is good.
Jan 12 2022
prev sibling parent russhy <russhy.rsy gmail.com> writes:
On Wednesday, 12 January 2022 at 10:02:10 UTC, Dennis wrote:

 On Wednesday, 12 January 2022 at 09:38:28 UTC, WebFreak001 
 wrote:

 well russhy did plenty of improvements to DCD


 That's great to hear! Btw, russhy, sorry for the implication 
 that you weren't working on it. I'm genuinely curious what the 
 barrier is for D users to implement the DCD features they want 
 (I haven't looked much at the code myself).


 I think it's mostly nobody reviewing them or having time for 
 it is delaying the contributions quite a bit.


 That's unfortunate. I started watching GitHub activity on the 
 pinned dlang-community repos, intending to get more involved 
 there.


it's ok, i'd have said the same if someone kept getting negative, 
but i'm trying to do better

the main issue is the project is way too scattered away


- dsymbol
- libdparse
- dcd
- containers
- allocators


everything has its own git repo, every project reference some 
random version from each other

it makes implementing fix/feature that touches many repo at the 
same time a pain to manage, and also i guess a pain for reviewers

i suggest the following:

- merge everything as a single entity in a single repo, and get 
rid of dependencies we don't need (containers? we have std, 
allocators? when will that get out of experimental?)

that'll be a start making the project easier to use and to 
understand! contributor friendly!
Jan 12 2022
prev sibling parent max haughton <maxhaton gmail.com> writes:
On Tuesday, 11 January 2022 at 12:29:25 UTC, WebFreak001 wrote:

 https://github.com/Pure-D/code-d/issues/393

 what do you think of this idea? worth checking out in the 
 future? not important at all or high priority?

 With [hatf0 showing DMD-FE being able to run in 
 WebAssembly](https://forum.dlang.org/thread/uccwuheorntxtashq
nt forum.dlang.org) I think this could be a very real possibility and we could
be pretty close to it being doable already.

 Instead of trying to port DCD could also try to immediately use 
 DMD for auto completion, though that's a whole other project 
 that would bring other benefits as well.



Brute force solution is to just host some codeserver instances. I 
will do some at some point, for trying out new pull requests etc
Jan 12 2022