Please dont abandon this project over time #934
Comments
|
Hello @zh54tg234f and thanks for your concern about the future of Material Shell. I understand your feeling about the project and I will share my current plans / vision about it while trying to answer your questions/concern. Is the project abandoned?The short answer is no. We need to differentiate Why is there less evolution to Material Shell and why it's still unstable?Material Shell is probably one of the Gnome Extension that alter the most the original Gnome Shell through the Extension API and it's mostly a compilation of hack that is combined in order to provide the Material Shell workflow. Every Gnome Shell alteration can be broken at any Gnome Shell update and need to be maintain / repair / redeveloped regularly. But the most annoying things is that even if we provide clean and safe code to hack a feature we reveal bugs that are inside Gnome Shell and face unexpected crash that we can't handle on our side. The overall feeling is that we are building upon a house of cards and adding new features increase the need of work exponentially. Why Material Shell ?With all the above explanations you will probably think that it was terrible idea to build Material Shell this way and before start talking about the future let's talk a bit about it's history. The Material Shell project began a bit more than 4 years ago. Everything started with the idea of using the natural sens of direction to organize windows and the desire to have a beautiful interface following the Material Design guidelines. I had no experience at all in Linux or OSS and I tinkered the first Proof Of Concept using AwesomeWM called Material Awesome.
After proven the POC it was time to prove the Proof Of Interest and therefore I decided to start over and make Material Shell. In order to make it available to the many and I chose the most distributed DE aka Gnome as foundation since it had an easy install extension framework and animation capabilities. It took me a full year to be able to provide the first beta and it kept growing piece by piece for another year After investing 2 years in the project and with it's growing popularity I hoped get enough funding from the community in order to allow me working on it. After all those investment it's complicated to see that the project is still unstable and all the time I found and invest in the project it's to fix issues that didn't existed the last month and came with the last gnome shell update. What is the future of Material Shell ?With all the issues and difficulties we currently have we can (with the 7000 github stars and ~60000 installation) accord the success of the Proof Of Interest for the workflow. I can't decide for the future of Material Shell alone but It's could be the right time to take a new step and get rid of the constraints of Gnome Shell and Mutter Window Manager and start our own. Another reason to start a new journey is that the current status of Material Shell doesn't reflect my full vision of the perfect workflow (because lack of time and technical limitation) But since I'm an happy father since 2 month (which also explain why it's calm lately) I can't spend another full year starting over alone or without any revenue. My current plan is to take time to write down all missing features specification and global workflow vision documentation and then either start a crowdfunding to gather a budget before start working on it or reach bigger linux corporation to get some financial support. Be reassured that I'm the first user of Material Shell and I'm currently doing my best for the project. Regarding
We already have a discord channel that you should joined ! I will draft a proper plan for the future of Material Shell soon so stay tuned and don't hesitate to get involved ! |
New Features are lowest priority since the basic functionality is not stable and result in many uninstalls and lack of interest.
Thank you for that but i miss the fundamental stability this projects needs as well as multi monitor support.
not bad for this beta but basically 0 in great terms
i understand this project is huge in terms of altering the gnome shell
seems to be the best option. i tipped you something for the multi monitor issue :)
When and where will we be able to notice? (estimate)
I can see your point since most tiling extensions some what suck but personally i think the smoothness of the reorganizing of windows seem unclean and laggy. But cosmic's tiling WM is fine, i don't need the double speed i need stability and flexibility and cosmic provides that. (its ugly but it works) |
|
Thanks for the tipped! What issue are you referring to exactly?
System76 are developing their own WM because they are also tired to hack gnome like us PaperWM is a fork which help a bit but still annoying |
|
using a 4k main display and full hd second display with material shell is an absolute nightmare |
|
@zh54tg234f Can you draft a proper issue about your monitor issue? I'll take a look |
Any updates? I will check if i can find next weekend. |
|
It's currently in progress. It's a bit complicated because I have several ideas and project. |
|
@PapyElGringo I found out about your project on r/unixporn. I've wanted to try it for a long time and recently got around to it. I love the tiling (mostly the idea of maximizing the window and forcing it to go full screen) and your solution looked nice and eye-catching (I love ricing). At the time, it was written in awesomewm, which I really liked. When I recently got my hands on your project, I was disappointed to learn that it is already on Gnome, which I hated (and the Redhat company too) probably starting from the 3rd version. But time goes on and things change. I decided to give Gnome and Wayland a third (not second) chance. When I installed the extension and went to configure it for my shortcuts, only then did I understand the main feature of the project - moving around the "grid" of windows/workspaces using the usual I also liked the way you did the split mode. The transition animation really enhances the experience. Moving the screen around a grid that fits n windows is a good idea. As well as a simple and quick adjustment of the number of windows on the top right. In fact, I would suggest that you improve this and allow the use of the number 1. A kind of combination of maximize and split. And that's where the good ends. Like my attempt at using wayland, gnome and your extension. In fact, there was a large paragraph of text about wayland and gnome. I was really disappointed, angry and worried about the future of desktop Linux. But almost after writing all this, I realized that you somehow had nothing to do with this and I simply deleted everything. There are whole articles on the Internet on the topic of wayland from people who, unlike me, not only used but also tried to fix things. Here is one of the fresh ones, good ones. And here's another one about the bad influence of gnome.
After all these years looking back, was it worth it? After all, you have lost something important - the possibility of customization for the user. And it seems to have significantly complicated the development itself. For animation, there was a lot to come up with. At least remake picom.
Endless rewriting of the extension for new versions so that it works adequately at least in most cases. And this is until the moment when the developers of the gnome want to throw out a certain API that you need. Sorry, but I'm pessimistic here. And as the last 10 years of gnome's development have shown, I was not particularly wrong in my predictions. I myself am switching to awesomewm(and void linux) with a fork of your work. But this is a temporary solution. I understand that xorg will be killed sooner or later, and wayland is just a disgusting solution. I'm looking at the arcan which seems like a really great thing. On this engine from the author himself there is a very original environment. I think if at one time you had switched to arcan as the basis of your DE, then at least the question of animations would not have arisen. I want to provide a couple of examples of bad work either by gnome or your extension. Might help you fix some problems. I didn't record anything from the gnome on wayland because it really annoyed me and the screen recording didn't work well. All of the following made in Xorg session: Win-A then Win-Dout.webmJust a random bug, I don't remember how it happenedout.webmCant open first workspace, idk whyout.webmAfter exiting fullscreen app
Also, after waking up from hibernation and entering the password, I consistently lost all windows from all workspaces after the first one. Like the workspaces themselves, except for the first one, they were lost. And I couldn't open other workspaces. Only opening a new previously unopened application "fixed" the problem and all the lost windows appeared on the first workspace. |
|
@Jipok Hello and thanks for all those feedbacks and to take time to come here and share your thoughts!
It's an original idea of mine and it's what motivated me in creating Material Shell. The material design guidelines did helped as they spur to visualize digital interfaces using physical concepts and that's what enlighten me. I'm currently in the process of formalizing the concept and idea and the current WIP is the following:
Don't hesitate to share your thought about it since I'm not an native English speaker ! I can understand your disappointment about Gnome but for me it was just a tool to propagate my workflow vision to the many. I have designed Material Shell to be the best desktop experience for everyone and not only for linux power users. Honestly if at the beginning it was doable under Windows OS I would have probably done it for it to make it available to the most of people. So to answer
I do think it was a dumb idea but still worth it since it allowed me to refined the workflow concept and make it available to many people. I'm now convinced that Material Shell does need to make it's own compositor and window manager because we cannot unleashed it's full potential otherwise. And of course I would have preferred started by that but at those time I wasn't even aware about all the technical stacks. Now choosing which is best way to do is another concerned because I'm not much opinionated about it. Xorg, Wayland or even maybe arcan (I'm not sure to understand what is it). I just want things to work properly. It's seem that all innovations are stacking up around wayland nonetheless. Gnome can only behave like they are because they have most of the users like Windows with Internet explorer once ago. But they also have most of the users because they offer the a well integrated environment and easy to access which is a GOOD things for linux community. Every open source project is navel-gazing because no one have resources to spend for others and that's because there is no money available in the open source desktop ecosystem. I would love to be more supported by the community, people enjoying the concept because we could do amazing things. |
Me too, but i chose material shell because of the nice gnome features being "implemented" within the wm, like calendar, notification, weather etc the side panel for icons, possibility to install other addons all would be gone if we ditch gnome and would require some good amount of work to make it beautiful again. believe me i don't like gnome, but their feature set for production / professional usage is the most visually pleasing and complete at the time.
Does not sound humble , since it implies we need much money. (would not harm) Microsoft has a lot of money and a lot of retarded software. Passion pays no bills but keeps the user base around.
This project is dying in terms of momentum regarding popularity because we wait for your ultra wide monitor. Get a cheap old hmdi 50$ ebay monitor and donate it to friends / the library / church when your done if you dont want to waste money. The current progress of popularity is like this: Power User sees material shell, thinks it looks awesome, installs it, bugs everywhere, uninstall, no support, no ultra wide monitor , repeat cycle with next user hoping we find enough people helping Question is what is your decision and do you have the resources for a transition or will it be a rebirth ?
Would be nice too see a community vote on this from all code contributors since they have a more realistic opinion |
|
@rizajur the only active code contributor is @HalfVoxel
I'm not waiting to get an ultrawide monitor to keep working on Material Shell or fix bugs, I'm just currently working on the bigger picture to resolve in-depth issues which the main is unlocking more workforce make the work spend in the project more efficient. Material Shell is pretty stable on my setup but not only it's also stable for many people that are not using tier extensions or exotic distribution and it's also available in the layout manager of Manjaro. One of your issues look like you used If we want to focus on a specific gnome shell version the best move we should make is to fork this specific version and update it properly instead of altering it at runtime. But this won't change that Mutter is not adapted for what we want to do. I'm tired to hack my way around to provide the features I got in mind and because I respect my user base I want to provide them the first class software experience and this come from reorganizing my resources. My first child is born 3 month ago which also take a lot of time.
The first thing to do in my opinion is to gather funds that would allow me and all other potential developers to work with dignity. And this meant either start a crowdfunding collect or find some partners and that what I'm currently working on. I want to organize community brainstorming events for all those subjects. |
said it regarding the multiple monitor setup. honestly its stable for me but im stuck to one display which makes the project kinda broken
Not using anything beside soft brightness and this never caused any issues.
Congratulation
multi monitor not working for 4 years - looks like disrespect to me |
|
I am not sure if it is a solution for u all, but porting the concept to a new WM, might be the best/only option to get this project stable. |
i have no obligation but it would be good if we could implement the current gnome shell applications like calendar , weaher, world clock, tray icons etc But this sound like a huge rewrite which could end up being like dwm where the user has to implement and hassle everything for himself in order to get a proper workflow. I dont like gnome, but version 43 as soon as stable seems to be for me the most realistic approach to make material shell stable and feature freeze complete for gnome 43 first than we can start working on a rewrite for a smooth transition otherwise we have no progress for a long time for the current user base. |
|
If there is a move away from GNOME, there could be many features lost :-/ I am currently using the following features from GNOME (non-exhaustive list), with Material Shell:
I guess other people use other parts of GNOME, or other extensions, etc. It looks like reimplementing everything would be complex and the first GNOME-less releases would be very limited regarding these features. Would it be worth the effort? |
|
@tiramiseb The answer depend on the timeline we are projecting. But I'm currently unable to make half of the features I planned for Material Shell workflow because of incompatibility with the Gnome Window Manager and Gnome ecosystem. So in the long terms we don't really have the choices. But I'm curious, for you @tiramiseb what is the cherry on the cake? Either having all the features you mentioned on top of Material Shell workflow or either having Material Shell on top of those features? If you had to keep one things what would it be? Anyway thanks everyone for sharing your thoughts here ! |
|
I am now addicted to Material Shell. Therefore, any improvement to the Material Shell workflow is a plus. If GNOME is blocking the workflow, I agree there is no choice, this approach would clearly be the cherry on the cake - it would allow avoiding the occasional bugs we have talked about here and in other issues. I came here because of tiling, after a few months with Pop! Shell. I have tried ion, but it too complex to configure. And animations are great (I am a visual person, and I understand what I'm doing better when I see workspaces slide for instance, even if I know I have just used the keyboard shortcut for it). And now, I have been trapped in the Material Shell workflow world 😉 I imagine most of the things I have listed (especially configuration stuff) could be done with some commands, but I don't know how: I have stopped looking at commands a decade ago, I want something that "just works" now, I don't have the energy to fiddle with the system too much. But there are things that cannot be done with command lines, especially these extensions that change some display or behaviour of menus (system monitor, appindicator, bluetooth, places...) ; some kind of extension system would be great (or having this stuff by default in Material Shell would of course be okay for me, but I can imagine other people have other needs, so...) |
|
I did a little bit of research and I found, that you could replace the wm of xfce with i3 maybe other wm as well... so this could be a good option I guess because it combines the config gui of xfce with the configurability of i3: xmone & xfce seam to be a even better fit: But you could replace the gnome wm with i3 too: |
|
How about a fork of gnome, and especially mutter? That's essentially what cinnamon did; they started as a set of gnome extensions, then decided that wasn't flexible enough and forked gnome. Recently they merged a bunch of stuff from mutter back into their window manager, but they waited a decade or so to do that. They also forked all of the desktop apps (like nautilus -> nemo), but I don't see much need for that here. It would be easier to just have a gnome-shell/mutter/gnome-control-center fork. |
|
Hello, first of all, this is an awesome extension, way better than gnome-xmonad. Thank you. What if there was a price tag on each request? E.g. I am ready to pay for displaying the workspace number on the left pane so that we can switch with the keyboard without having to think. |
|
First of all - thank you so much @PapyElGringo for creating Material Shell. It's by far the most promising interface I've worked with and it would be my only interface if it were more stable. After reading your comments above, I realize this instability is almost unfixable because of the 'house of cards' you're building upon - Gnome / Mutter / Wayland. As for becoming a dad - I know first hand how taxing this can be. Congrats and enjoy this very special period in your life - it'll be over before you know it :-) Keep up the amazing work and I'm looking forward to any updates on this project. If there's anything I as a non-programmer can do to help, please ask. |
|
@iGadget @ospyros Thanks for sharing your thoughts about the project! @ospyros Project developments are pretty slowed down lately on the project because I spend all my resources in planing the project transition and while any kind of income would really help buying some dev times we need to find some long-terms income to make the project viable. @iGadget We will have the same kind of issues with other projects and if we want to have sturdy foundations we need to make our own tailored Window Manager. But it doesn't mean reinvent the wheel either, it's exist project that help prevent that like https://github.com/Smithay/smithay. It's probably take the same amount of time porting Material Shell to Kwin (I'm not sure it's even doable) than making our own Window Manager.
@iGadget Thanks ! I'm currently composing a plan for the future of the project that will involve Crowdfunding and communications and every support will be vitals ! |
|
@PapyElGringo I came to the issues section because a recent update on Manjaro has broken the sizing of windows on MaterialShell. When I saw this post, I instantly read through it all because I can't live without Material Shell now. No other window manager will do. I moved from i3 to Material Shell and even moved my distribution to Manjaro for the built in Material Shell in layout manager to save me having to be careful about breaking updates of Gnome. If you want to move away from being a Gnome extension (I always thought it should be its own DE to be honest), I wouldn't mind getting involved. I can code (Python - Django, js and css) but my main skills are in commercial sides of software companies. I've built startups before and led global sales and tech teams for major vendors. Maybe I can do something commercially for you and try to get some revenues somehow to kick off the building of a standalone DE of Material Shell. If you fancy having a chat, give me a shout and we can have a sit down on a call. My time is pretty much my own at the moment so I wouldn't mind trying to get something kicked off. |
|
@speightashley I'm highly interested ! Reach me on LinkedIn or mail me 😁 |
|
@PapyElGringo E-mail sent to your e-mail address on your own GH profile. :) |
|
@PapyElGringo people are waiting. Debian 12 just released and they are shipping Gnome 43.4 in stable. I thought about it for a while. We cant rewrite within close time a panel to replace networking, Bluetooth, notifications, various gnome app api integration (weather, webdav) , night light, brightness control , media playback control, calendar events.
@PapyElGringo please make a proper RoadMap. |
|
Make a second project where we can start working on the prototype without wasting our time and make a final road map with steps we need to pass in order to have a product. than we can organize the framework requirements in MD files and have a more fluent work flow in the future when we know / outlined where the way shall lead to. |
|
@githubphind Hey there and sorry for the delay I know it's frustrating ! Let me give some insight of the progress so far:
All of this require some orchestrations and obviously any help is highly welcome. Shout out any questions or concerns ! |
|
Sorry about the voice of concern, it was not meant as accusation but we lose momentum in terms of people following the project. Thats why i want to focus on pushing things forwards and having a repo to start working on things instead of waiting. It would be very kind from you if you would add ReadMe etc for VeShell so others can start to try understanding the red line and than draw it out in the process together. I want to donate one specific oss-project - Material-Shell / VeShell. I think we should at least partially work on getting a working demo / - work in progress prototype in order to achieve more funding and code contribution. Please setup funding methods for the new project. VeShell <- How much is the first stream goal in usd $ to reach ? |
Link your twitch somewhere maybe? September???
wouldn't funds speed up the process or what is keeping you back? i would understand that i needs time to prepare things but could we cut the time horizon ? |
|
@githubphind
I'm unavailable for the 3 first August weeks and like you said I need to setup multiple things before being able to shout out about the incoming changes.
My goal is to get around 60k $ to being able to focus the full next year onto it
The momentum is already gone and that's why it's important to plan properly before activating all the community again in order to bring back light on the crowdfunding campaign
This is part of the setup of the
Currently having issue with my account but it's will be part of the large communication process
I need to write all the specifications which take time but it's planned I'm also available on our Discord if you guys want to be involved ! |
|
@githubphind
Is Open Collective better? |
dreams, we need less and this does not need to be a full time job. ublue came far with less expectations.
please keep things separated and don't mix them up. i want to support what i donate for, nothing else. Material Shell 2022-2023: |
|
@githubphind
There is plenty of dying projects around because they are all maintained by tired peoples that have to spend all their free time on project instead with their family to make them barely survive for people like you with very high expectation. Material Shell was not built on my free time: 2 hours on the week-end. I spent all my time for 2 years working on it using all my unemployment benefit and most of my savings. So now either I achieved to raise a substantial amount of funds for being able to keep working of it or either I will have to spent my time to work for this capitalized world.
I want to create a strong brand to lift the project up and that require a bit more ambition. The Free Explorers project serve as long term vision, donation jar and community of ultras that believe in my work.
What did you donate exactly? |
|
The argument from me was about lets focus on the near to mid term future on what we can do instead of setting very high goals. We need a bit ambition each at a time to grow on ideas, roadmap and code. Going with a more moderate approach seems to be better from my point of view than building a community of ultras... A strong brand comes from a strong product. We have neither so lets start with one product instead of multiple.
The amount of people living from their work around foss is very small so if it works out I'm very happy for you but be realistic we live in a capitalized world, so do the work as we all do, so we all can contribute as far as we are able and committed to the project. |
|
Dear Material Shell users, I would like to provide additional context regarding recent changes to our project and community. At present, most mainstream computer environments adhere to the desktop metaphor, envisioning the screen as the top of a writing desk where applications are placed, much like paper documents. Material Shell, however, diverges from this metaphor, introducing an innovative workflow that seamlessly aligns with our natural sense of direction, facilitating more intuitive navigation and organization of applications. Four years ago, I made the decision to develop Material Shell as a Gnome Shell extension, believing it to be the best way to make it accessible and discoverable for a broad audience due to Gnome's widespread popularity. However, implementing such a significant overhaul using the Gnome Shell Extension system presented considerable challenges. For those unfamiliar, Material Shell involves modifying Gnome Shell components through the application of monkey patches. These patches entail dynamic alterations of a class or module at runtime, driven by the intention to patch existing third-party code as a workaround for features that do not behave as desired. This approach led to the following constraints:
The lack of stability in Material Shell is not related to our codebase, and if you haven't seen many new features lately, it's not due to a lack of ideas. As we added features, maintaining the project over time became increasingly difficult. Despite the significant refactoring required for Gnome 45 compatibility, we have released the update. Due to our conviction in the innovation behind the Material Shell Workflow, our frustration with the limitations on our evolution, and my dissatisfaction with the current project status, I have decided to embark on a new journey. Firstly, I seek to unite Material Shell users and build a community of open-source enthusiasts eager to contribute to and fund innovative projects. I am reaching out to the Free Explorers, the Fearsome Pioneers exploring the unknown realms of the digital world, to join me in carving new paths for mankind. To support this initiative, I intend to create the best tool for us – Veshell, a powerful vessel for exploring and navigating the digital wilderness. Veshell, a standalone Wayland compositor powered by Flutter and Smithay, will unlock the full potential of the Material Shell workflow while offering a more stable approach. Godspeed, everyone ⛵. |
Is your feature request related to a problem? Please describe.
Lack of updates and work in progress (beta) software combined with few pulls and small inactive devs.
Describe the solution you'd like
Find out how we can keep this project alive or abandon before bugs get too absurd. (Update & fix ratio to slow)
Suggestions:
-> Make Some kind slack / matrix room , find out who is capable of what
I could make one if @PapyElGringo is okay with that.
Organizing the work
Make a feature and priority list according to the votes from the matrix room.
Don't reinvent the wheel
The code for resizing windows etc seems buggy / unclean we should cross reference with competing projects in the gnome tiling extension market. Use synergies and help these projects , take code from them but provide (try) fixes for new both effecting issues due to upgrades etc.
https://github.com/paperwm/PaperWM
No one can work for free for ever, maybe setup a ViewKey Published BTC/LTC/XMR wallet
and make than either a bounty for feature requests or the lead dev decides which pulls is worth how much and the money is fairly split.
I have donated in the past but i dont know what or how much it takes to keep this project alive and make is great as its vision of beauty full designed gnome enhancement without makes trouble, logout kicks etc....
I would be willing to put up 500$ for further work on the project and would try to aid the devs.
If anyone could give anything (not yet) and we could get some more i guess its more worth the effort.
We have a lot of 'noob' issues, missing config stuff. I guess if we have someone experienced who can show
a noob like me a easy way on what to watch and shows the do's and don't does than i guess we can easily bump the pulls.
Otherwise this project will die without a miracle and it could take ages before material shell is enterprise ready or at least ready for stable usage.
Describe alternatives you've considered
Tell us the project is dead, we move to alternatives without the hassle.
The text was updated successfully, but these errors were encountered: