Where's it going?

classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

Where's it going?

Graham Lee-3
Hi folks,

I originally sent this message to Greg back in March but we haven't
found an opportunity to talk about it one-on-one since, I'm really
interested in answering these questions though so I'm opening it up for
discussion on the list. I appreciate any insights GNUstep collaborators
can provide.

as you know, I'm an on-off contributor to GNUstep, with a lot more "off"
than "on". I'd like to be more involved, and I'd like to convince others
to get involved, but I realised I'm not clear on what the direction is,
and there isn't a lot of guidance from the public materials. I find it
difficult to see what I or others _should_ be doing and easy to get
overwhelmed by the things that I think I _could_ be doing. I have a
friend who is interested in GNUstep (I will discuss our motivations at
the end, so this finishes a bit more upbeat) and who says the same: of
all the things to choose from, it's not clear what would be more valued.

I'm aware that what I'm about to launch into will come across as a rant.
Two things: yes it is, but overall I've been around GNUstep since about
2002, I'm glad to be around it and I'm trying to find a place and being
frustrated; secondly, I'm sending you this email privately as the chief
maintainer so that you know the problem exists and we can work together
to resolve it, which I hope is better than giving up (which I
occasionally do) or airing my gripes on a blog or in the list.

The public website at gnustep.org gives the impression that GNUstep is
baked: it's a "mature framework" with a "robust implementation" of the
libraryies and tools. There's no guide to where the project is going, no
request for help, no call to action for people who are interested.

The project page on Savannah similarly doesn't have any clear guidance,
the news is out of date, and the bug tracker doesn't have any clear
"themes" or steers from the maintainers, and seems not to get much
triage or interaction (there are open bugs from 2003). What are the
things you wish we had? What are the bugs that you would like people to
fix? What is the next thing that an upcoming contributor can, well,
contribute to? This situation is (temporarily, I appreciate) exacerbated
by having some patches in Savannah and some in GitHub.

The wiki does include a roadmap of sorts (http://wiki.gnustep.org/index 
.php/Roadmap) but again it's hard to pull _actionable_ tasks out of
this. For example, one item is "Gorm 1.4.x": if I just update the
version in git (currently at 1.2.23) are we done or is there a
collection of tasks that make Gorm 1.4.x ready? How do I know how far
along this goal is, and what I can do to enable it?

I said at the beginning that I would describe my motivations, which
match with those of a friend who also occasionally gets involved. I'd be
interested to discuss how well this aligns with what you see, with your
chief maintainer hat on, as the direction for the project.

My motivating goals for being involved are twofold:

1. having a cross-platform desktop app development environment that I
enjoy using. I am working on an app at the moment, and enjoy being able
to use ObjC away from a Mac. My main frustrations on this point are Gorm
and PC having bugs which mean that I have dropped them for my app
project, it's not clear how much time I would invest in fixing some of
the problems (e.g. the drag and drop NSFormatter problem I discovered
over a year ago in Gorm and still haven't addressed), and whether the
time invested is worthwhile or whether there will be a strategic move to
Ycode or another IDE.

2. providing a Free Software haven for Mac developers who get
disillusioned with Apple (as I did), or put off by the move to Swift, or
have other reasons to need homes for their code. From this perspective,
what's missing is an API "gap analysis" showing what you can do on macOS
10.x that you can't do on GS; importers or compatible tools for Xcode
projects, storyboards, autolayout UIs; clear guidance on whether I'm
expected to support gcc or whether it's OK to just use modern ObjC and
clang; and a priorities list to see the relative urgency and value of
'missing' cocoa features like updated gscoredata, webkit, pdfkit, and
others.

I'm sure there are other goals people could have for a framework like
GNUstep, I'd be interested in hearing your perspectives and finding out
some next actions I could help with.

Thanks,
Graham.


_______________________________________________
Discuss-gnustep mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep
Reply | Threaded
Open this post in threaded view
|

Re: Where's it going?

Graham Lee-3
Hi Riccardo, thanks for your reply.

> However I anticipate a conclusion: why do you need goals to contribute? 
There will be something that attracts you in GNUstep: talk to the 
contributor of that part and start using and providing patches. Or just 
work on the long buglist we have! 

I have not observed this to work. I have months-old pull requests and patches, and years-old bugs, that have not been triaged, even to say "we are not interested in this" or "this is bad and you are bad". Hence me asking, what sort of contributions are the project team looking for?

Thanks,
Graham.

From: Riccardo Mottola <[hidden email]>
To: Graham Lee <[hidden email]>, <[hidden email]>
Sent: 14/06/2018 2:52 PM
Subject: Re: Where's it going?

Hi Graham

these emails usually trigger an infinite discussion of different goals,
rants about not having a common on and critique of every kind.
Please not. Nevertheless, rereading your email I thought I'd give my
personal reply. I've been around GNUstep too since about 2002! And am
official GNUstep maintainer as well as of several apps and "de facto"
maintainer of even more stuff., so I hope my opinion is not totally
worthless.

However I anticipate a conclusion: why do you need goals to contribute?
There will be something that attracts you in GNUstep: talk to the
contributor of that part and start using and providing patches. Or just
work on the long buglist we have!
Maybe you are interested in Foundation? maybe in one or more Apps? go!
If you feel more for general help across everywhere... there are many
tasks from checking and updating packages for many distributions.

Graham Lee wrote:

> as you know, I'm an on-off contributor to GNUstep, with a lot more
> "off" than "on". I'd like to be more involved, and I'd like to
> convince others to get involved, but I realised I'm not clear on what
> the direction is, and there isn't a lot of guidance from the public
> materials. I find it difficult to see what I or others _should_ be
> doing and easy to get overwhelmed by the things that I think I _could_
> be doing. I have a friend who is interested in GNUstep (I will discuss
> our motivations at the end, so this finishes a bit more upbeat) and
> who says the same: of all the things to choose from, it's not clear
> what would be more valued.

I would value anything directed towards GUI and Apps.. but somebody else
could tell you WebObjects or Database backends.. or whaterver!
Don't forget to seek fun: this is OpenSource, nobody will give you a
raise.. and in many years just a few people wrote me thanks for what I
did, mostly people download your app and don't tell you, neither good
nor bad stuff, just discarding if they don't like it.

>
> The public website at gnustep.org gives the impression that GNUstep is
> baked: it's a "mature framework" with a "robust implementation" of the
> libraryies and tools. There's no guide to where the project is going,
> no request for help, no call to action for people who are interested.

Mature and robust mean that it is intended for production use, not just
as a proof of concept. It is complete enough to do useful work.
It doesn't mean that it is complete and there is nothing to do, also we
know that the GUI and Backend parts need more work than Base. I suppose
you are around long enough to know this!

However that statement refers only to GNUstep as its core libraries, if
you instead think also of its applications because you want some kind of
environment, the scenario is worse and "any help" appreciated!

>
> The project page on Savannah similarly doesn't have any clear
> guidance, the news is out of date, and the bug tracker doesn't have
> any clear "themes" or steers from the maintainers, and seems not to
> get much triage or interaction (there are open bugs from 2003). What
> are the things you wish we had? What are the bugs that you would like
> people to fix? What is the next thing that an upcoming contributor
> can, well, contribute to? This situation is (temporarily, I
> appreciate) exacerbated by having some patches in Savannah and some in
> GitHub.

The bug list on savannah is long enough to have fun for years. I suppose
every reporter will imply his bug is important!
I do not track much on github, I don't like the platform: can't even
find the bugs you are referring to it, only libobcj2 seems to have bugs.
But I don't want to spur any discussion here, time becomes more limited
and I like concentrating on fixing and improving stuff. Github was not
meant to be a definitive move, but it proves convenient enough and I
suppose it will just stay.

Do you really want a list of bugs I'd like to have fixed and couldn't
yet fix myself? IN that case I can send you privately a list :)

>
>
> I said at the beginning that I would describe my motivations, which
> match with those of a friend who also occasionally gets involved. I'd
> be interested to discuss how well this aligns with what you see, with
> your chief maintainer hat on, as the direction for the project.
>
> My motivating goals for being involved are twofold:
>
> 1. having a cross-platform desktop app development environment that I
> enjoy using. I am working on an app at the moment, and enjoy being
> able to use ObjC away from a Mac. My main frustrations on this point
> are Gorm and PC having bugs which mean that I have dropped them for my
> app project, it's not clear how much time I would invest in fixing
> some of the problems (e.g. the drag and drop NSFormatter problem I
> discovered over a year ago in Gorm and still haven't addressed), and
> whether the time invested is worthwhile or whether there will be a
> strategic move to Ycode or another IDE.

I can't speak for Greg, but unfortunately "time" is relative in
opensource, so if somebody has personal issues he might get slower and
this is especially noticeable in a small community. I hope he will get
back in strength and fix your bug!
As for PC, I invested countless hours fixing bugs (not yet released). Of
course it still has many and limitations. I intend continue working on
it, but of course, if you want to use another IDE, up to you, there
shall be no monopoly. In the last months though my attention is elsewhere.

>
> 2. providing a Free Software haven for Mac developers who get
> disillusioned with Apple (as I did), or put off by the move to Swift,
> or have other reasons to need homes for their code. From this
> perspective, what's missing is an API "gap analysis" showing what you
> can do on macOS 10.x that you can't do on GS; importers or compatible
> tools for Xcode projects, storyboards, autolayout UIs; clear guidance
> on whether I'm expected to support gcc or whether it's OK to just use
> modern ObjC and clang; and a priorities list to see the relative
> urgency and value of 'missing' cocoa features like updated gscoredata,
> webkit, pdfkit, and others.

For me the guideline about GCC vs Clang is very clear: I want all the
stuff I work on to work on GCC. I want that both for compatibility
reasons, convenience (and lately also just possibility: it works out of
the box BSD and Linux, while I am struggling to use libobjc2 and/or
clang on most platforms I try). Also, my personal taste is that I don't
like "modern" ObjC at all. Up to know I have not found a case where my
code wouldn't work on Clang/libobjc2 except when it was bad code and I
fixed it.

>
> I'm sure there are other goals people could have for a framework like
> GNUstep, I'd be interested in hearing your perspectives and finding
> out some next actions I could help with.
>

Well, I think we share both goal 1 & 2.

I could however phrase them differently:

1) I want eventually GNUstep to be a comfortable environment on my
Workstation and my Laptops. Have a comfortable environment, which is
pleasant to use and develop in. I don't like directions Windows and Mac
are taking, but neither the ever-changing desktop mess of Linux (think
of Ubuntu with Unity and now all again...). I'm not the only one who
wants an efficient, classic desktop, people disillusioned from Mac,
Window 10 as well as GNOME4 pop up, especially on BSD and certain Linux.
Most then flock to Xfce or Gnome 3 revival, however there you look only
at the use, I still prefer to use AppKit and Obj-C instead of GTK and C++!

In this sense, a "gap analysis" makes not much sense to me: I need
everything I need to use and code in, not everything apple has. Of
course, if something is missing, implement it the Apple way to maintain
compatibilty

In this goal I work on all the Applications: they give me the
environment, be it a simple battery monitor, Terminal, Workspace and
GNUMail! And also ProjectCenter.
I really use LaternaMagica to do some slideshows or to prepare some
Photo CDs!

2) a Framework where to code stuff I need and which is portable, from
the Raspberry to Windows. The Raspberry is the fun side, the Windows is
the platform certain apps are needed to be used beyond our small
community. Sadly, this part is lacking, I did invest time to improve our
situation, but with no real result.
That means on Windows I do not need to have a full environment, but be
able to run a GNUstep apps as smoothly as possible.


Riccardo

_______________________________________________
Discuss-gnustep mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep
Reply | Threaded
Open this post in threaded view
|

Re: Where's it going?

Fred Kiefer
Hi Graham,

I also prefere to stay out of such general discussions and will only reply to the one point below.

> Am 15.06.2018 um 09:28 schrieb Graham Lee <[hidden email]>:
>
> > However I anticipate a conclusion: why do you need goals to contribute?
> There will be something that attracts you in GNUstep: talk to the
> contributor of that part and start using and providing patches. Or just
> work on the long buglist we have!
>
> I have not observed this to work. I have months-old pull requests and patches, and years-old bugs, that have not been triaged, even to say "we are not interested in this" or "this is bad and you are bad". Hence me asking, what sort of contributions are the project team looking for?


Are you referring to the pull request on GitHub? We maintainers don’t get any notification from GitHub when a new pull request is created. The only way to find out about them is to go through all the different modules and check whether there is a new one. Not a very efficient system. I would suggest that you go through all of the GNUstep sub projects yourself to see how time consuming that process really is.
Today I found your pull request plus two others that add Japanes language support to base and gui. I merged all three of them. But in the future if you want your pull request to be handled faster, please add at least one GNUstep maintainer to it as reviewer, for base Richard and for gui and back Ivan or me. That way we will get a mail about it.

Cheers,
Fred
_______________________________________________
Discuss-gnustep mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep
Reply | Threaded
Open this post in threaded view
|

Re: Where's it going?

David Chisnall-7
On 15 Jun 2018, at 13:36, Fred Kiefer <[hidden email]> wrote:
>
> Are you referring to the pull request on GitHub? We maintainers don’t get any notification from GitHub when a new pull request is created. The only way to find out about them is to go through all the different modules and check whether there is a new one. Not a very efficient system. I would suggest that you go through all of the GNUstep sub projects yourself to see how time consuming that process really is.

I believe you must have explicitly opted not to get them for this to be the case and it is configurable.  GitHub will send email notifications for a variety of things, including new pull requests or issues for any project that you’re watching.  I certainly get emails when anyone files an issue or PR for libobjc2...

David


_______________________________________________
Discuss-gnustep mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep
Reply | Threaded
Open this post in threaded view
|

Re: Where's it going?

kanzlei

i want to be removed from this mailig list with all adresses of the domain txt.de. Can you help, please?

Kind regards

David Chisnall <[hidden email]> hat am 15. Juni 2018 um 18:01 geschrieben:

On 15 Jun 2018, at 13:36, Fred Kiefer <[hidden email]> wrote:

Are you referring to the pull request on GitHub? We maintainers don’t get any notification from GitHub when a new pull request is created. The only way to find out about them is to go through all the different modules and check whether there is a new one. Not a very efficient system. I would suggest that you go through all of the GNUstep sub projects yourself to see how time consuming that process really is.

I believe you must have explicitly opted not to get them for this to be the case and it is configurable. GitHub will send email notifications for a variety of things, including new pull requests or issues for any project that you’re watching. I certainly get emails when anyone files an issue or PR for libobjc2...

David

_______________________________________________
Discuss-gnustep mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep


_______________________________________________
Discuss-gnustep mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep
Reply | Threaded
Open this post in threaded view
|

Re: Where's it going?

Ivan Vučica-2
(list to bcc as this is not ontopic)

The link to do so is in the signature of every email. Individual users can unsubscribe using this link.

Readers of this list don't generally have administrative access. We generally cannot unregister other users, much less perform such large scale operations as 'remove all user of a domain'.

Not to mention that users cannot generally get on the mailing list except by signing up and confirming registration. It would not be okay to get removed just because one particular user has requested everyone else gets removed.

This is therefore a completely wrong venue to ask for such a change. I have no suggestion about where you could reach out instead.

On Fri, Jun 15, 2018 at 6:32 PM <[hidden email]> wrote:

i want to be removed from this mailig list with all adresses of the domain txt.de. Can you help, please?

Kind regards

David Chisnall <[hidden email]> hat am 15. Juni 2018 um 18:01 geschrieben:

On 15 Jun 2018, at 13:36, Fred Kiefer <[hidden email]> wrote:

Are you referring to the pull request on GitHub? We maintainers don’t get any notification from GitHub when a new pull request is created. The only way to find out about them is to go through all the different modules and check whether there is a new one. Not a very efficient system. I would suggest that you go through all of the GNUstep sub projects yourself to see how time consuming that process really is.

I believe you must have explicitly opted not to get them for this to be the case and it is configurable. GitHub will send email notifications for a variety of things, including new pull requests or issues for any project that you’re watching. I certainly get emails when anyone files an issue or PR for libobjc2...

David

_______________________________________________
Discuss-gnustep mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep

_______________________________________________
Discuss-gnustep mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep

_______________________________________________
Discuss-gnustep mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep
Reply | Threaded
Open this post in threaded view
|

Re: Where's it going?

Fred Kiefer
In reply to this post by David Chisnall-7

> Am 15.06.2018 um 18:01 schrieb David Chisnall <[hidden email]>:
>
> On 15 Jun 2018, at 13:36, Fred Kiefer <[hidden email]> wrote:
>>
>> Are you referring to the pull request on GitHub? We maintainers don’t get any notification from GitHub when a new pull request is created. The only way to find out about them is to go through all the different modules and check whether there is a new one. Not a very efficient system. I would suggest that you go through all of the GNUstep sub projects yourself to see how time consuming that process really is.
>
> I believe you must have explicitly opted not to get them for this to be the case and it is configurable.  GitHub will send email notifications for a variety of things, including new pull requests or issues for any project that you’re watching.  I certainly get emails when anyone files an issue or PR for libobjc2…

This may be true for repositories where you are the owner, but have you ever thought about the fact that you do not get this emails for the rest of GNUstep? I have now added myself as a watcher to the  four main GNUstep modules.

And I found a query for the PR filter to show all the relevant requests at once:

"is:open is:pr archived:false user:gnustep“

There where actually quite a few there. I was able to merge most of them but want to apologise to everyone for taking so long to do so. We never intended to use GitHub for change requests or issue tracking. Now they are happening there and we need to adjust to this.

Fred


_______________________________________________
Discuss-gnustep mailing list
[hidden email]
https://lists.gnu.org/mailman/listinfo/discuss-gnustep