This year, DConf has an extra day tacked on for problem solving in the form of a hackathon. The intent is to work on issues people find frustrating in the D ecosystem. While there will be time given at the event for proposals, and those involving third-party projects are welcome, it will help speed things along for the organizers to come in with a list of big issues to get the ball rolling.

To help in compiling the list, what are some major issues from the ecosystem that you'd like to see fixed?

On Thursday, 27 April 2017 at 14:53:02 UTC, Mike Parker wrote:
> To help in compiling the list, what are some major issues from the ecosystem that you'd like to see fixed?

* multi-threaded parsing in DMD
* https://github.com/dlang/dub/issues/838
* finishing std.experimental.xml https://github.com/dlang/phobos/pull/4741
* Docs which allow people to go back and see docs for previous versions. Huge headache when using GDC or LDC

On Thursday, 27 April 2017 at 14:53:02 UTC, Mike Parker wrote:
> This year, DConf has an extra day tacked on for problem solving in the form of a hackathon. The intent is to work on issues people find frustrating in the D ecosystem. While there will be time given at the event for proposals, and those involving third-party projects are welcome, it will help speed things along for the organizers to come in with a list of big issues to get the ball rolling.
>
> To help in compiling the list, what are some major issues from the ecosystem that you'd like to see fixed?

For starters:
- https://github.com/dlang/dub/issues/104
- https://github.com/dlang/dub-registry/issues/117
I don't think the following ones are feasible to deal with in that context, but they are ones that irk me in the ecosystem:
- ABI compatibility between D compilers with the same frontend version.
- More than one official package description language (json and sdlang). I honestly don't care which it ends up being, but please pick *one* (I am aware of the previous discussions on the topic, but the current state of supporting both is just one more point of friction for newcomers)

> - More than one official package description language (json and sdlang). I honestly don't care which it ends up being, but please pick *one* (I am aware of the previous discussions on the topic, but the current state of supporting both is just one more point of friction for newcomers)

SDL should be dropped. for starters it's poorly supported. just go to https://sdlang.org/ and scroll to the links at the bottom. official homepage and reference are both broken urls (as http://sdl.ikayzo.org/ is offline) and needs mirrors. Then there's the fact that there's not very good support for SDL in other languages, sure the website says there are implementations in other languages but has anyone tried using them? I did recently, the Java version (actually written by ikayzo - https://github.com/ikayzo/SDL) will not even parse all the examples on the homepage! not to mention it being a dead project with no response to issues (https://github.com/ikayzo/SDL/issues). As far as I can tell the only reason dub defaults to sdl now is because s-ludwig maintains the sdlang.org website. at least json is universally known and can already be handled by just about any tool. If we really want to have an alternative it would at least make sense to use something like YAML that is also heavily used and supported by all IDE's. It would save me the hassle of writing a lexer for SDL - which of course has no support in my tool of choice because...

NOBODY USES IT!

On Thursday, 27 April 2017 at 16:33:02 UTC, singingbush wrote:
> As far as I can tell the only reason dub defaults to sdl now

It did in the past, but hasn't done so for several minor versions (it defaults to json).

On Thursday, 27 April 2017 at 14:53:02 UTC, Mike Parker wrote:
> This year, DConf has an extra day tacked on for problem solving in the form of a hackathon. The intent is to work on issues people find frustrating in the D ecosystem. While there will be time given at the event for proposals, and those involving third-party projects are welcome, it will help speed things along for the organizers to come in with a list of big issues to get the ball rolling.
>
> To help in compiling the list, what are some major issues from the ecosystem that you'd like to see fixed?

I would vote for getting the std.decimal proposal from the review queue into a good shape for std.experimental.


Another big issue for me is using dub in a company. Big companies do not want to use the official dub repository due to security issues. They want to run their own dub repository with dub packages they have done code checks. That means also the git projects are mirrored in git repositories within the company. First step into this direction is to have the possibility to specify the dub repository address within the dub.json.

Kind regards
Andre

On Thursday, 27 April 2017 at 17:06:39 UTC, Andre Pany wrote:
>
> Another big issue for me is using dub in a company. Big companies do not want to use the official dub repository due to security issues. They want to run their own dub repository with dub packages they have done code checks. That means also the git projects are mirrored in git repositories within the company. First step into this direction is to have the possibility to specify the dub repository address within the dub.json.

Is there a problem with just using the dub command line arguments for this?
# cat > /usr/local/bin/company-dub
#! /bin/sh
/usr/bin/env dub --skip-registry=standard --registry=YOUR_COMPANYS_REGISTRY

The pending pull requests.  In person is a great high-bandwidth way to work through the massive backlog.


On 4/27/2017 7:53 AM, Mike Parker via Digitalmars-d wrote:
> This year, DConf has an extra day tacked on for problem solving in the form of a hackathon. The intent is to work on issues people find frustrating in the D ecosystem. While there will be time given at the event for proposals, and those involving third-party projects are welcome, it will help speed things along for the organizers to come in with a list of big issues to get the ball rolling.
>
> To help in compiling the list, what are some major issues from the ecosystem that you'd like to see fixed?

On Thu, Apr 27, 2017 at 11:22:06AM -0700, Brad Roberts via Digitalmars-d wrote:
> The pending pull requests.  In person is a great high-bandwidth way to work through the massive backlog.

+1. Phobos at one time was down to about 30-40 PRs, but now it has clogged back up to around 100. We need to get it back under control so that valuable contributions aren't just sitting there bitrotting.


T

-- 
Не дорог подарок, дорога любовь.

On 04/27/2017 04:53 PM, Mike Parker wrote:
> This year, DConf has an extra day tacked on for problem solving in the
> form of a hackathon. The intent is to work on issues people find
> frustrating in the D ecosystem. While there will be time given at the
> event for proposals, and those involving third-party projects are
> welcome, it will help speed things along for the organizers to come in
> with a list of big issues to get the ball rolling.
>
> To help in compiling the list, what are some major issues from the
> ecosystem that you'd like to see fixed?

I'm most frustrated by regressions and wrong-code bugs in dmd/druntime/phobos (with a focus on dmd). Regressions are annoying, because that stuff used to work, dammit! And wrong-code bugs are annoying, because they can be hard to track down and work around.

I don't really have a list of them sorted by importance or impact, but I guess the one's I have reported are, on average, more dear to me than others [1][2].

I haven't reported this one, but it has a special place in my heart, because the miscompiled code is so simple:
https://issues.dlang.org/show_bug.cgi?id=15538

Generally, just sorting the issue list by severity and going from the top would probably have a good impact:

https://issues.dlang.org/buglist.cgi?bug_status=__open__&columnlist=component%2Cassigned_to%2Cbug_status%2Cbug_severity%2Cshort_desc&content=&no_redirect=1&order=bug_severity%2Cbug_status%2Cpriority%2Cassigned_to%2Cbug_id&product=&query_based_on=&query_format=specific

Instead of trying to fix those bugs during DConf, maybe experienced devs could help interested folks get into it. Look at an issue together, lay out what needs to be done about it. I'm thinking about dmd in particular, because it seems to be understaffed. Only works if there are people who want to get into fixing compiler bugs, of course.



[1] My regressions: https://issues.dlang.org/buglist.cgi?bug_severity=regression&email1=ag0aep6g&emailreporter1=1&emailtype1=substring&list_id=214636&query_format=advanced&resolution=---
[2] My wrong-code bugs: https://issues.dlang.org/buglist.cgi?email1=ag0aep6g&emailreporter1=1&emailtype1=substring&keywords=wrong-code&keywords_type=allwords&list_id=214637&query_format=advanced&resolution=---