Hello Firas,
I agree with you in the points you are suggesting. However, I might have made a mistake to link to that Workstation specifications, because now it looks like I want to change the core applications for the Workstation. On the contrary.
I would like to take those Workstation core applications as specimen of possible core applications (in my understanding, these should be apps that are thoroughly tested and that solve a very basic and important use of an operating system.) I am also totally fine with Gnome applications that are currently used for it.
What I am not very satisfied with the following points:
- For Gnome Workstation, all applications are tested and require basic functionality, however we do not define exactly what basic functionality is. Therefore, the results might vary from a person to person. Also, I believe that the core applications should be tested more than just for basic functionality, if basic functionality means that they start and do something.
- There are spins, where not all of the core applications (mentioned in the list for GW) are present in the spin. Here, I do not think about particular Gnome equivalents, but there was a spin where I could not find a terminal emulator. At least, it was not part of the menu. For less advanced users, this could be frustrating. So, I would like that all spins would have their own equivalents of core applications and we would check that they are in the menus and that they start. I understand, that we will not block on them, but it would increase the usability of the spins with minimum effort.
I totally agree, that the main focus should stay on the core of Fedora, which is Gnome Workstation. If we stopped blocking on the whole KDE and just checked for the core applications, we would win the time to check for core applications on every spin.
Thank you for your views.
Lukas
I would really appreciate your thoughts on my points Lukas.
- Workstation in Fedora is automatically referring to the gnome desktop environment. Other environments are called '<XYZ> Desktop'.
- The links that you provided are the technical specifications for the Fedora Workstation. While test-cases should be high-level, specifications should be precise and specific. And because Fedora Workstation is a Gnome environment, it is natural to exactly specify what package is the core app for a specific service. Coming from a company that deals with medical devices I think reducing detail level of the specification would not improve anything.
- A basic principle of quality: it is not possible to achieve a high level on all 6 quality characteristics. Starting to spend resources on every desktop environment dilutes the quality of the core product. Even though I prefer XFCE, Fedora Workstation is a gnome environment and this should have all the focus. (I am not saying that we should ignore the spins, embrace them and help the community, but stay on the core path!)
Kind regardsFiras D. Nuwayhid (nomos)
From: Lukas Ruzicka <lruzicka@xxxxxxxxxx>
Sent: Wednesday, November 7, 2018 4:38 PM
To: For testing and quality assurance of Fedora releases
Subject: Re: [fedora-qa] Issue #569: Proposal to redefine core applications.Actually, this might be a misunderstanding. The testcase is called *Workstation* core apps, and the technical specification wiki is in the *Workstation* namespace. The Workstation SIG have defined their core apps, and we have a test case for them. There are no other core apps. So sure, workstation core apps are tested on workstation images, and nowhere else, because that wouldn't make sense :)
By core apps, I am not talking about Firefox, gnome-terminal, and such. I am talking about terminal emulator, web browser ...
On the contrary, core apps make a lot sense for all spins. I do not see why there should be a spin made without a terminal, for instance. It does not have to be gnome-terminal, but some emulator should be present. The same goes for other apps that I believe are at core of a computer system, therefore I call them CORE applications.
[trimmed]
core apps, we can define KDE core apps, XFCE core apps, Server core apps (which we somewhat already have, just in a different structure, in our existing criteria), ...
We will probably just block on what we block now, but it could be a nice signal to the spin groups that something like that is "a nice to have". If they do not want to follow it, we will not be able to do anything about it, but I see it as a logical thing. If, for example, there is no text editor in LXDE, the user experience is somehow limited. And as far as I know, Fedora promotes some of the spins as suitable for older laptops. Sure, but why not to push for some better quality of that spins, although we do not block on that?
Workstation WG will want to have virtualization front-end, while KDE won't. Text-only spins won't want to a web browser. Etc.
Virtualization frontend should go from core apps, IMHO. At least according to what I believe is a core app. If it is in the spin itself, ok. But this is nothing that would be needed by everyone and a crucial part of the system.
I don't understand which apps and which functionality you talk about here. We already require basic functionality for all default desktop apps, and our criteria include required functionality in many tools/apps outside of the basic scope. What do you mean here?
I believe that OS is a good OS, when you can do something with it. There are some minimal tasks that it should be able to do for you. For example, it has to let you install packages. On the other hand, it does not have show you a route from one place to another. What I am trying to say here is, that for those core applications, we should probably focus more on the functionality than just basic functionality and for the rest we might be more tolerant.
Example, during the blocker review meeting, we were discussing if Maps had a blocker bug, finally the WG decided that it was not a blocker bug. I agree it was not. But ... If gnome-terminal had a similar rendering issue ... would that be a blocker? why or why not? If this was among the core application, I would say that it would be definitely a blocker to me. So basically, I am suggesting to make the situation a little less messy and the guidelines a little clearer.
I'm missing one thing and perhaps I misunderstood some of this because of that. What is your main motivation here? What exactly are you trying to improve? Thanks.
My motivation is:- realize what is really important that we test -> let's discuss that
- if we realize, that there is something "not as important" -> test for the basics and not spend time too much on such things - who is really interested about the functionality of ABRT reporter?- if we realize that there is something "very important" -> think how to test it more thoroughly (I believe those core apps are a good example of apps that should be tested thoroughly.)
Therefore, I wanted to start the discussion. If you think, that we do not need any improvement here, it is a valid opinion. Perhaps, it's just me who supposes that we could improve the routines a little bit. If so, I gladly give up on trying and we'll be good as gold.
--
Lukáš Růžička
FEDORA QE, RHCE
Purkyňova 115
612 45 Brno - Královo Pole
![]()
--
Lukáš Růžička
FEDORA QE, RHCE
Purkyňova 115
612 45 Brno - Královo Pole
_______________________________________________ test mailing list -- test@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe send an email to test-leave@xxxxxxxxxxxxxxxxxxxxxxx Fedora Code of Conduct: https://getfedora.org/code-of-conduct.html List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines List Archives: https://lists.fedoraproject.org/archives/list/test@xxxxxxxxxxxxxxxxxxxxxxx
