The first two X-Apps are ready
by clem 111

In preparation for Linux Mint 18, a new project called “X-Apps” was started, which goal is to provide default and generic applications for traditional GTK desktop environments (Cinnamon, MATE, Xfce…).

Since the long-term plan for GNOME applications is to better integrate with GNOME Shell and to embrace new UI guidelines and techniques which are specific to this environment, it became more and more urgent for traditional desktops to find suitable alternatives. GNOME applications once looked native in other desktops and provided users with a consistent experience. This is no longer the case, yet this is still part of our expectation and so this is something we had to address.

A few years ago, developers used to write “Linux apps”. Most of them (at least all the GTK ones) looked consistent, with similar widgets, similar look and feel and you could use them in your favorite environment without worrying about them looking out of place. Nowadays, you can write “GNOME apps” or “Ubuntu apps” using specific techniques or following specific concepts which make them look awesome in their specific environments. And in a way, this is great. But as apps become desktop-specific or distro-specific they need to be replaced in environments they no longer properly support.

The idea of maintaining Cinnamon applications, or investing time in developing MATE applications or Xfce applications was rejected. For the most part, these desktops present similar needs so it made sense to maintain one generic set of apps which can be used in all of them, and to do so outside of the scope of any of these desktops.

The core ideas for X-Apps are:

  • To use GTK3
  • To use a traditional UI (titlebars, menubars)
  • To be generic, desktop-agnostic and distro-agnostic
  • To provide the functionality users already enjoy (or enjoyed in the past for distributions which already lost some functionality)
  • To be backward-compatible (hopefully all the way to GTK 3.10)

The first two X-Apps to be ready are the text editor and the media player.

The text editor is based on the GTK3 version of Pluma and features the same functionality and UI as gedit 2.30:

Capture du 2016-02-03 13:26:54

The media player is based on totem 3.10:

Capture du 2016-02-03 13:26:40

It provides the same UI and Mozilla plugin (for Quicktime and WMP support) Totem 3.10 did.

111 thoughts on “The first two X-Apps are ready

  1. Reply Peter Feb 3,2016 14:45

    Hello,
    I like your idea about X-Apps. It is important to bring the whole linux world forward, not just its own ship :-)

    Congrats for this design decision! Hopefully, sometimes I will have some time to participate to one project or another.

    Just some ideas/questions…

    Are there any plans to provide a PDF reader X-App? Maybe ATRIL based? I mention this, because it is a big deal for public administrations and offices. Specially for PDF forms, annotations, and printing (previews).

    Will you merge filemanagers from different desktop environments too? That would be a big advantage, especially for plugin developers…

    Keep up your great work!

    Best regards,
    Peter

    • Reply clem Feb 3,2016 14:58

      Yes, very likely, and atril is a good candidate (compared to evince it features epub support for instance).

      Regarding nemo, caja and possibly thunar… it hasn’t been talked about. I know it would solve some issues and there are many people who would like to see nemo become an independent application. It’s not something we “need” to do so it probably will take longer (assuming it ever happens). There’s also something in the file managers which is tied to their environment.. it’s their management of the desktop itself (i.e. ~/Desktop). I can see how it could suit multiple environments… but there’s just a little more to think about here.

      • Reply Elad Hen Feb 3,2016 18:10

        Hi Clem,

        This is great news. My question is – all of the free (as in speach) pdf readers in Linux today, notwithstanding Atril and other progeny of Evince are broken regarding searching and copying in Right-to-Left languages (it finds words only if you type them backwards, and if you copy it pastes backwards [cba instead of abc] in the destination). This is an ancient bug that no one seems inclined to solve, and is one of my biggest ongoing bugbears as a linux users.

        Is there a way (please please pretty please) that you guys take a look at this problem?

        Thanks in advance. I love your work and I love the Idea of X-Apps.

        • Reply Elad Hen Feb 3,2016 18:13

          Bah! Notwithstanding should be *including8 Atril. It also has this bug…

        • Reply clem Feb 3,2016 19:34

          Ok, it’s funny you mention this because that’s precisely what I worked on today (eog and evince). I usually don’t talk much about what’s not ready yet.. but this is coming soon so.. I obviously scrutinized the git history of evince (it’s important to understand what is gained, lost, or made DE-specific, and when) and I did notice a couple of RTL fixes. “If” we decide to maintain a reader as an X-APP, it goes without saying that fixes and features will be ported (whether it’s evince bug fixes or atril’s faster rendering and epub support) and that you will be able to open issues for them.

  2. Reply A. L. Feb 3,2016 15:31

    Congratulations! This is a [b]great[/b] initiative that is desperately needed nowadays.

  3. Reply kaizer Feb 3,2016 16:42

    Hi Clem,
    Can you please explain me the benefit of developing x-app compare to using directly existing generic app like for example geany and vlc?
    Thank you in advance

    • Reply plata Feb 3,2016 17:58

      If you look at geany and vlc, you can study the result of a truly cross platform software: as always, making it work everywhere implies that the app does not really fit anywhere. You can run vlc on KDE, Gnome, Windows… therefore it cannot be consistent with the design guidelines of any of them. This makes it look out of place everywhere.

      Basically, there are two ways to walk around this issue:
      1. Have different applications for every desktop environment (e.g. kwrite vs. gedit)
      2. Use one core and attach different user interfaces for KDE, Gnome…

      Obviously, x-apps belong more to category 1 (although it’s not exactly one DE but rather a group of DEs following similar design guidelines).

      The second category can quite easily become a mess programming wise (as e.g. Qt imposes a tight coupling between data and GUI by automatic MWC structures etc). Therefore you usually will not do this for rather simple applications like a text editor. Complex applications often follow their own design principles based on the specific workflow they are designed for and are not adjusted to fit particularly well into a certain DE because of this. (Some developers might adjust their apps a little. E.g. Mozilla will make sure that Firefox fits into Windows 7/10)

      In my opinion, this is the reason why you find many DE specialized small apps and big apps which are not specialized.

      As for geany and vlc in particular: They belong somewhat between those categories. Geany is more than just a text editor but not quite a real IDE yet. VLC can do a lot more than only play videos.

      For any DE you want a set of basic apps to cover the basic functionality such that they integrate really well into the whole look and feel (e.g. text editor, basic video player etc.). These are exactly the ones which are planned to become x-apps (@clem correct me if I’m wrong).

    • Reply clem Feb 3,2016 19:25

      It’s continuity. We’re already using gedit, we used it for years and for as long as we can remember. It works, people like it, it fits its purpose perfectly, and it has done so all this time and represented the best tool for us while geany was there and available as well. If you look at gedit 3 as a new and different application (and this is more and more true) and you consider that gedit 2.30 is was somehow discontinued, then the question is simply to accept its loss (and therefore to switch to an alternative) or to maintain it. You’re seeing a new project here, but to many users this will be the same text editor as before. If you look at totem and totem-mozilla (that’s also still important), it’s the same story… we actually do ship VLC with it but it never replaced it. Now, with all that said, the goal is to define what we need (and what we’re not ready to lose), it’s a case by case story, so in some cases it will be preferable to consider the app discontinued and to simply ship with an alternative instead. Development/maintenance costs and expectations are both very important here. We can’t meet our expectations by shipping with apps designed for another desktop, or by switching them for alternatives which didn’t fit as well or long term by freezing them and having nobody develop them further… and in terms of costs, there’s an important cost related to patching and relying on components which don’t actively support the environments you committed to. Forking a new app costs a lot at the beginning but you also get a lot back from it long-term. (Note: It’s funny you mention geany, I really like that editor myself. There were discussions about a year ago with MATE and Geany devs on the IRC about the possibility for Geany to be shipped in a “user mode” and act as the default application in MATE and Cinnamon. In the end it didn’t result in anything, it was interesting but it wasn’t what the Geany devs had in mind and not everybody within the MATE team was keen on retiring Pluma anyway).

    • Reply JosephM Feb 3,2016 21:05

      Part of the issue with choosing an existing app is hidpi support. In many cases you want the sort of look and feel of gtk2 but need gtk3 to compatible. Not many existing apps fit the bill for the combination of needed features and it allows the flexibility of making the changes when/where needed.

  4. Reply Sam Burgos Feb 3,2016 19:48

    So far you have x-editor, x-player and blueberry (as you mention in other post), also you might be planning a pdf viewer (with might meaning not planned but a possibility), any other app you may like to port/create/fork/(put action here) for the project? Also any plans for documenting or give some help to install to other distros out there like Arch or Fedora, whose GTK version goes beyond than the one on Ubuntu/Mint?

    • Reply clem Feb 4,2016 01:41

      These two apps already work and were tested with GTK 3.10, 3.14 and 3.18. We’ll support all distributions, just like we do with Cinnamon and other cross-distro projects. If there are issues with GTK 3.20, we’ll probably get PRs from bleeding-edge distros like Arch and Fedora. You can compile and install these apps the traditional way, but they’ll likely be packaged by most distributions pretty soon. Debian, Arch, Fedora and OpenSUSE in particular are usually quite efficient at doing that.

  5. Reply Lando Feb 3,2016 22:52

    I just want to say that I think the Mint team is doing a fantastic job! Linux Mint 17.x was the best Linux distro ever released IMHO. The focus on LTS really paid off. Well done!

    It’s a shame you are not able to fix Windows too:)

    I think the best FOSS apps are Firefox/Pale Moon, VLC, Libre Office, Qbittorrent. These apps use different toolkits but integrate perfectly in both MATE and Windows.

    I think X-Apps make a lot of sense. It’s GTK3 apps for traditional desktops. Only thing that worries me about GTK3 is that MATE will lose its ability to recolor GTK themes. If you change colors in Appearance then only GTK2 apps are affected.

    • Reply Monsta Feb 4,2016 08:25

      We (MATE devs) suspect it’s due to deprecated stuff like GdkColor. If we guessed right, porting to GdkRGBA and other modern GTK+ stuff might just solve this problem.

      • Reply Lando Feb 5,2016 08:51

        OK, let’s hope so. Funny thing is that both QT apps and Libre Office follow GTK2 color adjustments nicely. It’s GTK3 apps that look out of place. Somehow I feel there is less integration between GTK2 and GTK3 than there is between GTK2 and QT.
        I guess that when Gnome 2 was the leading desktop then other toolkits wanted to blend in seamlessly.
        I understand GTK3 is the future. Hopefully MATE can keep all its unique features.

        • Reply clem Feb 6,2016 11:01

          GTK3 is a work in progress. GTK+ is stable and finished.

          When it comes to theming and look and feel, there’s no such thing as GTK3. There’s GTK 3.10, GTK 3.12, GTK 3.14, GTK 3.16, GTK 3.18 and recently one version which changed everything again.. GTK 3.20. The GNOME devs are doing a very good job but there’s a huge gap between what they do and what people think they do. This isn’t a toolkit which was ready to use for app developers and artists alike when it hit version 3.0, not at all. This is a brick needed for the development of GNOME Shell, which is itself a bit of an experiment. GNOME isn’t following a waterfall model where everything was decided and functional specs were finalized before being implemented in 3.0… you’re not looking at bug fixes with every new version of GNOME, you’re looking at trunk development.

          Also, GNOME 2.x was dominant as a DE, and GTK+ as a toolkit and the project’s components integrated very well outside of GNOME. It was natural for everybody to try and fit with it as well as possible. Compare that with GTK3 where there’s uncertainty, DE-specificity, distro patches, different versions (which you need to target.. for instance eog 3.14 didn’t work with GTK 3.10, due to deprecation fixes and the use of 3.14 specific features)…

          GTK3 works great and it’s very stable by itself. In fact it’s a brilliant toolkit. It changes a lot though and it’s a rocky ride. If you’re an artist you need to commit to modifying your theme every 6 months for every new version. If you’re a developer you need to accept that your app might start looking or behaving slightly differently every 6 months. If you’re a distribution you need to patch stuff or make compromises every 6 months. It’s a great adventure and it produces great results but it’s a lot of work for anyone relying on it.

          If you were to work on QT integration with GTK3, wouldn’t you hold off a little until GTK3 reaches the point where it stops changing and only receives bug fixes?

          • Reply Lando Feb 6,2016 17:43

            Of course you wouldn’t want to integrate with a moving target. That’s the problem with GTK3. Will GTK3 ever hit “stable” and only receive bug fixes? I doubt it.

            What this means in the long run is that a GTK3 version of MATE will not be MATE. The biggest strength of Gnome2/MATE is that it is/was the universal platform where everything (except GTK3) blends nicely.

            To be perfectly honest I don’t see the point of porting MATE to GTK3 unless there is a stable/final version of GTK3.

            I do understand GTK3 is the future and it is required for Wayland etc. However you do have a fine GTK3 desktop called Cinnamon. I am afraid that everything that is unique about MATE will be lost if it is ported to an unfinished GTK3. What we really need is a stable GTK3. It was released in 2010 and many apps like Firefox still use GTK2. It’s because it is not “stable” and noone knows if it ever will be.

            Obviously I don’t have the answers, just a feeling that porting MATE to an unfinished GTK3 will take away the biggest strength of MATE.

  6. Reply Mike Feb 4,2016 09:52

    X-Apps is a brilliant idea, I’ve always asked my self why every DE has it’s own text editor, calculator, image viewer, etc.

    The only thing that would beg for is, do not create an X-Office suite.

  7. Reply Caner Feb 4,2016 10:11

    Is the naming X-Apps somehow related to X Display Manager?

    • Reply clem Feb 4,2016 11:08

      X connotes a few things:

      – The idea of an Xorg common base
      – “ex”, how it was before (before it went away)
      – Genericity

    • Reply Monsta Feb 4,2016 11:51

      …and by pure coincidence, Ubuntu 16.04 (which will be the base for Mint 18) is codenamed “Xenial Xerus”. :)

  8. Reply Mohammad Saad Feb 4,2016 10:35

    Will there be an x-app web browser too? Would love to use though.

      • Reply Mohammad Saad Feb 4,2016 13:40

        Can you then make it after the release of Linux Mint 18? I don’t trust Firefox, Vivaldi or Chrome for browsing web and would love to see something different, even if it takes time. Please do look into it.

        • Reply peter e Feb 4,2016 14:20

          @ Mohammad Saad
          Please explain why you distrust Firefox. I am a long time user, I need something to go on.
          I suppose Thunderbird would also be on your list of things to no trust.

        • Reply clem Feb 4,2016 16:43

          No, sorry Mohammad, the reason we’re starting to maintain these apps is because we need them in the scope of Linux Mint. Firefox on the other hand works very well for us as a default browser and there are plenty of alternatives for people to choose.

          • Reply Mohammad Saad note: Saad is my first name Feb 5,2016 07:25

            Can you then ship pale moon or or some other browser by Linux Mint 18 then? I do use Firefox but they are putting more focus on firefox hello, pocket integration and becoming chrome like. As long as Mozilla is good in making firefox secure, it’s fine then, but if things should go haywire anytime, you should be ready to ship another one in there.
            You have also seen Brave browser right? Just downloaded LMDE 2, used Linux Mint Maya, Nadia, Olivia, Petra, Qiana, Rebecca, Rafaela and Rosa. Thank you clem. Linux user since 2012 (Windows 8 disaster). Just keep on improving guys, I really love Linux. :D

          • Reply ZenTiac Apr 7,2016 12:50

            I agree Firefox default settings make it look like chrome. However you can right click at the top and left click Menu Bar to add it again. Hope this helps. This may not be what you are having issues with. However I thought I would mention it.

    • Reply Salamandar Feb 23,2016 02:37

      I completely agree. The existing X-apps are already late seeing the original apps… Let’s also see Evince and its gtk2 fork Atril.

      Why wouldn’t you make forks that just change the headbar/toolbars ?

  9. Reply Brian Read Feb 4,2016 11:54

    Usual question – I run Cinnamon on Fedora 23 – how can I pick these up? I already use puma as a simple text editor for exactly the reasons you mention, so am keen to try these.

    • Reply clem Feb 4,2016 12:02

      You need to wait a little bit for new apps to reach distributions. I know OpenSUSE and Fedora maintainers already started looking at them, but there’s a process specific to each distro, they might require changes from us (the text editor had to change name slightly for instance, this was done today) and it might take a few days before you see Fedora RPMs.

  10. Reply Brian Read Feb 4,2016 12:44

    ok, thanks, I’ll need to know what the names are in order to load them?

    • Reply clem Feb 4,2016 16:42

      The editor is called “xed”, the media player is “xplayer” and the image viewer which is almost ready is called “xviewer”.

      • Reply Bob Long Feb 5,2016 01:17

        So xed will be an extension of CP/M’s “ed” editor? Just kidding!

        Seriously, will xviewer have the ability to detect hyperlinks and thus be clickable? Document Viewer 1.12.0 (evince) can’t do that if the link is simply “plain” text.

        • Reply Monsta Feb 6,2016 10:10

          xviewer is an image viewer, not a document viewer :)

        • Reply clem Feb 6,2016 10:35

          The first goal is to get these apps ready and fully functional. This is now done for xed, xplayer and xviewer. After that we can start improving them. Just like any other projects, we’ll take “issues”, and pull requests, so if there’s something you think is missing, head towards github.com/linuxmint and make it happen :)

  11. Reply Jacques Feb 4,2016 17:28

    Great work Mint team! The future looks very exciting thanks to all your hard work and vision.

    One app that nobody has mentioned so far… A Twitter app? Corebird development has been picked up again by an Ubuntu dev and it is already available for Ubuntu 15.10.

    elementary has Birdie, which is now exclusively developed for them.

    Corebird should work on Mint 18, right? Are there any apps I don’t know about that works great on Mint?

    Thank you.

  12. Reply jmmourinho Feb 4,2016 21:25

    Good good!

    You guys got it in the start of the new Gnome/KDE/Unity/metro DE’s fallout with a shiny new yet traditional Cinnamon and a legacy supporting Mate. You rose to the top in Distrowatch, many users use it and I sure do always come back to Mint (presently LMDE Cinnamon btw).

    Glad to see you taking one more step in that path, both toward the future but securely grounded in the present and supporting of the past.

    Since some of these new apps are based on Mate ones, mean the X-Apps will end substituting those in the long run?

    As for the long run too, is there any plans for an “app framework” or “app prototype” like Elementary’s Granite, for developers to base on to build their app in a consistent theme to the traditional Cinnamon/Mate/Xfce desktop layouts?

    • Reply clem Feb 6,2016 10:28

      About a framework: It’s likely we’ll see the need for common libraries as we start maintaining more and more apps, at the very least for utility functions, DE detection, configuration etc. Regarding toolkits, it’s likely we’ll see widgets appear as well, and for two reasons. The first reason is that GTK3 is a GNOME component. It is released as such, maintained as such and we see more and more signs of it being designed as such. As time goes by GTK3 makes applications look more and more like GNOME Shell. You probably noticed the differences in dialog boxes, sidebars, button boxes etc.. between Mint and LMDE (i.e. between GTK 3.10 and GTK 3.14) already. I’m firmly opposed to forking GTK3 (there’s been some heated debates about this, especially within the MATE team) and if we did I don’t think it would solve all our problems (and it would introduce new ones because we couldn’t just “replace” GTK, we’d need to support both toolkits, it would be a real mess). So our current solution is to patch GTK3, which means you get a consistent experience in Linux Mint, but things look out of place in GNOME, and in other distributions. It works for us short-term but it’s not ideal. A more long-term solution is to start developing GTK widgets… for instance we’d make our own filechooser dialog, and our apps would call that rather than the dialog provided by GTK itself. So that’s one reason why we’d want to start maintaining widgets… Another reason is related to the trend. You can see it in Windows/Mac/GNOME/Unity, style is being reinvented. A lot of it is experimental, it shocks people, but once they get “used to it” (we often hear that) they either love it or hate it. I think some of it might become popular quite fast with a vast majority of people, including people who reject that new look. To take an example, have a look at the Cinnamon System Settings.. it’s using traditional GTK layouts with a combination of newer-looking widgets, revealers, stack-switchers, editable-entries, cinnamon-boxes (sorry, I don’t know what to call these since they’re specific to cinnamon still). We need to be careful with this, it might fit one environment and not another just yet, but going forward it might become something that fits and elegantly marries traditional with modern elements. We already have a need for a lib delivering these widgets (they’re duplicated across some mint apps and cinnamon at the moment), so it’s likely we’ll have something like that in the future, yes.

      About MATE apps being discontinued: That’s a decision for the MATE team. We wear different hats when working on these projects. If I wear the Mint app I can tell you Mint 18 MATE Edition is likely to use X-Apps. If I wear the MATE hat, the scope is different and I need to take other distributions into account, pluma supports GTK+, xed doesn’t… pluma is compatible with python plugins, xed isn’t… etc etc.. It’s also worth considering that the idea behind the X-Apps has been discussed for a while now, not only within the Mint team, but also within the MATE team, and in some occasions with Xfce devs, the fact that Mint decided to own leadership and maintain this project is because it failed to reach consensus in the first place. Stefano for instance, was opposed to the idea of dropping MATE apps. We’ve a history and success behind us and we’re confident we’ll gather momentum behind these apps, but we need to prove that once again before MATE and XFCE are in a position to consider reducing their scope and relying on X-Apps, if they ever do.

  13. Reply Artem Feb 4,2016 23:38

    Clem, you may have trademark problems with both xplayer and xviewer, google the names.

  14. Reply drbogi Feb 5,2016 08:38

    I hope you made the right decision, because it really has become pointless to have specific application for each DE. If we add the Ubuntu and Debian base, 32 and 64 bit distribution, the number of combinations becomes worrisome.

    I’d have three questions:
    1. How will fit the existing plugins for X* applications?
    2. Will there be both 32 and 64 bit versions?
    3. Will it be used with the LMDE Mate DE, ask for another? :)

    Greetings.

    • Reply clem Feb 6,2016 10:44

      1. They won’t. We can make X-Apps compatible with GNOME-Apps plugins but I don’t think that’s the best solution, especially long-term. xplayer ships with a collection of plugins (xplayer-plugins and xplayer-plugins-extra), xviewer-plugins was ported also (https://github.com/linuxmint/xviewer-plugins). I see a lot of plugins which shouldn’t be plugins though and which functionality should be present by default in the core application itself. For instance, the ability to show/hide tabs in xed should be something that is there, not something that you need to activate via a plugin. We’ll take this case by case, and we’ll review ideas and missing plugins reported to us. We might port a specific plugin or integrate its functionality depending on what it provides. In any case, if there are plugins or features you’re missing, please create a github issue and/or pull request.

      2. Yes.

      3. Yes, that’s part of the reason to develop generic applications. We want efforts made into improving them to benefit as many editions as possible. Each innovation performed say on xed, will improve a very wide landscape, Mint Cinnamon, Mint MATE, Mint Xfce, LMDE Cinnamon, LMDE MATE, other distributions using xed, ppa users in Ubuntu etc etc.

  15. Reply acrophoenix Feb 5,2016 10:34

    Hi Clem, I really like your idea of X-Apps. Thanks for the fantastic job your team is doing!

    Since you will be working on an X-pdfviewer, may I ask if you are considering to add the functionality of true annotation that is long missing in all linux pdfviewers? I mean, people should be able to highlight, comment, add note etc. to the pdf file itself, just like in Adobe reader.

    This is the one thing I’m still relying on from Windows. With it there, I will be able to ripe off my Windows partition. Thanks a lot!

    • Reply clem Feb 6,2016 11:04

      Why not. I don’t think we’ll start implementing new features right away (in the scope of Mint we’ll probably start to do so with Mint 18.1) but yes, we’ll consider things like that.

  16. Reply Mahen Feb 5,2016 12:31

    Hi Clem !

    Nice ! But isn’t it possible to benefit from the work done for instance in XFCE instead of reinventing the wheel ?

    For instance, Ristretto is a very fast picture viewer (faster than eog). Parole is a very nice media player…

    As a Linux user, I’m frequently frustrated to see this kind of situation :) Even if you’re clearly not responsible for it. When one considers KDE, for instance, the basic apps are soooo powerful and fast even if binaries and dependencies may seem heavier (gwenview, dolphin, okular… both features-full and fast… once loaded !) but when it comes to a GTK desktop, there are many not-that-satisfying choices.

    Please share work with XFCE !! :-) I’m pretty sure they share most objectives and views with you.
    Thanks for your work and pragmatism.

    • Reply clem Feb 6,2016 11:10

      We’re working on continuing to provide the apps many of you are already using and which no longer exist upstream. So we’re not reinventing anything, quite the opposite :)

      In each case, Xfce apps (and alternatives also) are considered as a base or even something to switch to. They are, also, DE-specific though and not generic.

    • Reply session Feb 8,2016 09:47

      Really? Huh. One of the reasons I don’t use Xfce is that, on my hardware from the year 2000, native apps are slower than their Mate counterparts. Ristretto is terribly laggy when dragging large pictures, and that’s with image quality limited in settings. Since the beginning, Metacity/Marco has rate limited the updates to a window’s size, reducing CPU use during opaque move/resize… I wish Xfwm did that.

      • Reply Mahen Feb 10,2016 16:08

        @clem : thanks for the pragmatic reply. Let’s see what happens :) Thanks again for your work.

        @session : Well this display issue is a distinct problem. I meant many GNOME apps are particularly slow even if other parts of GNOME are pretty fast. eog is terribly slow for instance. It doesn’t even preload the next picture ! On slow PCs it takes several seconds to display a picture when it’s quasi-instant with other viewers. It seems like a stupid example but it’s been this way for YEARS. I remember filing a bug about this, 5 years ago (even more), people replied that it was indeed necessary but not possible because of the app architecture…

        I have similar observations with Nautilus VS Dolphin in huge drawers etc… When they contain thousands of files, if Nautilus (and its derivatives) takes half a minute, it takes 1 or 2 seconds in Dolphin.

        Some apps are not only “uncluttered” but are also extremely slow, which, along with stability, is a HUGE issue IMHO.

        Of course I don’t blame the Mint / Cinnamon team at all. I just hope they are aware of it when forking existing apps..

  17. Reply mrman Feb 5,2016 14:39

    Awesome Idea.

    Again, Mint team being sensible and pragmatic whilst the others play in the corner with their shiny new toys forgetting to share with the rest of the group.

    At this rate Mint will be the leading Distribution on Linux Debian base if it isn’t already.

  18. Reply Peter Feb 5,2016 15:13

    @Mahen I really like this!

    It is important to reduce work and profit from already existing projects. Maybe also to combine efforts from other projects. For example, ask people from ATRIL and XFCE and others to join the effort of building a DE independent set of core X-applications, or at least backends.

    I think, that there should also be a restrictive definition of how GUIs should look like and act. For instance, menues or keyboard shortcuts etc.

    If we fork projects over and over again I fear that we have not enough man-power to bring these projects to a real powerful state. However, X-Apps as idea points already in the right direction…

    Beside X-Apps also configuration tools (display settings for example) could be shared and brought forward together with other DE developers.

    What do you think?

    Cheers,
    Peter

    • Reply clem Feb 6,2016 11:37

      Yes, indeed, and as you’re talking about this it’s possible Cinnamon might have to lose its Users & Groups configuration tool, to switch to a generic one.

      This might sound new because it’s the first time we talk about it here, but we’ve been talking within the different development teams about common apps for more than 2 years now and before that we’ve been talking with GNOME devs (like Jasper for instance) about the need for Alan Day UIs (which look awesome in GNOME by the way, credits where credit is due, but out of place in traditional DEs..) to be configurable, about the need for GTK to be told by the DE how apps should look like, about the need for app developers to be able to implement both menubars and headerbars (for instance) and have their app look native both in GNOME and traditional desktop environments… and it wasn’t a waste of time, far from it, but by the time GTK 3.18 came by and it was the time for us to work on Mint 18, not only was nothing ready to solve all these issues, but nobody was working on them. And that’s ok.. that’s a scenario we had in mind as well. It doesn’t mean we won’t work with MATE/Xfce (or even upstream) in the future, but it’s quite hard to gather people on an idea and it’s much easier (and much needed at this stage) to lead by example, to implement that idea and show people the results it brings us. At the very least it solves our own problem, and I expect it to become much wider than Mint and also much wider than the particular DEs we talked about.

      Regarding manpower I think we’ve a responsibility here. The Xfce team is pretty small, and so is the MATE team once you remove the Mint devs from it. Mint is among the best funded open-source projects in the world, with a budget allocated specifically to development. We’ve seen this with Cinnamon, once you create momentum, once you satisfy a need many people have and you create the conditions for people to contribute to it, many people help. A lot of that man-power isn’t currently working on MATE or Xfce apps, it might not even be working on GNOME apps, a lot of it is spread or inactive. After a while you see people stay and form a core team, but new people come from pretty much everywhere.

  19. Reply Tonatiuh Feb 5,2016 16:27

    Hi Clem and Linux Mint team.

    When the X-apps are ready.
    Are will avaible to users on Linux Mint 17.x? Or will be need to upgrade to Linux Mint 18?

    • Reply clem Feb 6,2016 11:40

      I want GTK 3.10 to be supported as much as possible so Mint 17.x and Ubuntu trusty users can backport these apps or use PPAs, but in the scope of Mint itself these apps are prepared for Betsy and Mint 18 (i.e. they would work fine in Mint 17.x but we’re not planning to backport them to the Mint 17.x repositories).

  20. Reply KDB Feb 5,2016 21:32

    Hi Clem,

    Since xed is based on Pluma, and xviewer on eog/eom, does it mean that Pluma and eom will be retired and their dev will now develop xed and xviewer?
    Efficient collaboration? :)

    KDB

    • Reply Monsta Feb 6,2016 10:13

      MATE is still GTK+2 by default, GTK+3 build is available but is still considered experimental. We’ll have to keep compatibility with both toolkits for a while. So Pluma will still be Pluma.

    • Reply clem Feb 6,2016 11:44

      That’s a question for the MATE team. I think it’s too early to consider retiring Pluma from the MATE project personally. Before talking about that, X-Apps need to be successful of course, but MATE also needs to consider whether it’s ok or not to stop supporting GTK+ (it’s not necessarily a hard decision.. we can decide to ship MATE-core.. panel/libs.. in GTK+ and switch the apps to GTK3, but it’s a decision which needs to be taken unanimously and with all distributions in mind).

  21. Reply Giuse Feb 6,2016 06:29

    You said chat x la che integration with both but what about Wayland? Will you look at that?

    • Reply clem Feb 6,2016 09:41

      Nobody in our team is working on Wayland. Will we look at it in the future? Maybe, it depends how viable it becomes (I don’t mean whether it just works, but whether it becomes something we can actually use… all apps would need to work, without regressions or loss of features, all toolkits and environments would obviously need to work as well, all currently supported GPUs would need to work as well as before, including those using proprietary drivers… etc etc.). There might be a part of that which is our own responsibility, we could for instance allocate some resources to make Cinnamon a Wayland compositor, but until we know more about the likes of GTK+, MATE and other components we support, there’s no real reason for us to invest in it just yet.

  22. Reply MisterE Feb 6,2016 21:52

    Maybe another X-Apps candidate is a terminal like MATE- or GNOME-Terminal?

  23. Reply mta Feb 7,2016 11:02

    maybe a better way to name X-apps, xed, xvid, xpic, xpdf..etc…etc ???

    • Reply Monsta Feb 8,2016 09:19

      Well… a PDF reader called xpdf, despite being very old, is still provided in Debian and Ubuntu repos. Besides, Atril and Evince aren’t just PDF viewers, they support other document formats as well.

  24. Reply Dominik Feb 7,2016 13:13

    I like Cinnamon because its stable and easy to operate
    also for people that just wanna use there PCs
    for this group its irrelevant which operating system they have
    as long as it works

    and cinnamon fits this nearly perfectly

    Hopefully you will find enough developers for so many new components …
    the stability should not suffer

    during the last few years i have ported about 50 PCs and notebooks to LinuxMint
    mostly as i already wrote,
    these users have just little PC skills (like 90% of the worldwide computer users)
    and there is one really big and important feature still missing in LinuxMint

    a good working auto update function
    like windows has ;-)

    i am really always explaining them, everytime i see these people, how to update there mint
    but at the end 50% of them, still don’t uses the manual mint update

    but i cant blame them for that
    they just wanna surf a little bit on the net
    buy something online
    and maybe watch a youtube video
    most of them are not installing new programs
    or understand what i mean if i try to explain them security issues

    So please integrate a default
    auto update function

    that would be so important to spread it to a wider group of PC users

    • Reply Bob Long Feb 9,2016 01:15

      @ Dominik

      “a good working auto update function”

      There already is “unattended-upgrades” in the repositories; I’ve never used it. Does anyone know if it works well in Mint, and what might have to be done to not interfere with the current Update Manager?

      • Reply peter e Feb 9,2016 16:19

        I found that the unattended-upgrades package is already installed on my LMDE Mate… I see no option to turn it on or off in Update Manager.

        auto update is Not Desired! that is a spookie thing for me, I want to control it.

        That said, I have installed Mint Mate and LMDE Mate on many machines, and most users don’t bother with updating, ever.
        6 months or more with no updates, no problems! this is not like Windows that is constantly broken and constantly needs attention. “latest and greatest is not necessarily better”

        • Reply Dominik Feb 9,2016 17:49

          I can just hope you are new at linux
          I know that many people think that there are no viruses malware
          whatever out there for linux

          but the truth is, security problems, are not windows exclusive ;-)

          for example
          i don’t wanna use online banking with an outdated browser

          its good to have the control about all updates
          and of course experienced users should have the ability to update there system manually

          but most people are not experienced
          i am trying to spread mint between normal users, not nerds and geeks
          and you already wrote it
          “6 months without updates”
          its definitely better to get updates automatically than getting no updates
          i am pretty sure that most of the no problems users will never notice that they are already part of a bot net

          at least
          unattendend-upgrades
          you wouldn’t propose that if you would have ever tried to use this on mint

    • Reply gunvolt Feb 9,2016 20:53

      I think that that is a good idea. There is unattended upgrades available in the repos and installed by default and can be enabled by command line.
      However, I think Mint should fork it to make it work with the security levels(like in update manager) and with a gui configuration tool(in mintupdate)

  25. Reply gunvolt Feb 7,2016 23:12

    Now we also have Xviewer(image viewer) in development(github).
    How about a PDF viewer(XPDF?) and Archive manager (xzip?) Wherever this project goes is up to you.

  26. Reply jors Feb 8,2016 04:09

    I congratulate you on your clem work now for the future also you have to think also in creating new applications where GNU / Linux in general there are few or no applications ie those categories where there are weaknesses in comparison to windows

    I would like an application of Automator radio as ZaraRadio Dinesat or Jazler future

  27. Reply Anand Feb 8,2016 07:14

    An aside – I just love Segfault blog, it is so wonderful to pick what’s there on your (dev) minds and understand the direction the project si taking!

    P.S. I couldn’t do any official work today morning due to this though ! :P

  28. Reply mark Feb 8,2016 12:58

    Hopefully you can lower the burden of maintenance too.

    That is, 3 years down the road how simple it is to maintain a lot of apps.

    Gnome, while we can all agree that they have become essentially mad even if we all find nicer words for that, at the least pushes out new releases very regularly and at a fairly fast pace (even if nobody wants the crazy changes they are focusing on).

  29. Reply Andrea Feb 8,2016 21:11

    hi Clem

    This idea with x-apps ist really good. But one question:
    how can I install these new x-apps?? I would like to use them already in LMDE2-Betsy. To be honest, I also enjoy LMDE2-Betsy Cinnamon.

    And one more question: will there also be a complete office-suit in x-apps consisting of writer, presentation and calculator as well as database?? This is most what I use. And I would also like to see something for graphics based on Digikam oder XnView. This is also a thing what I use much and I also appreciate the creation of an pdf-Viewer. This is also what I use.

    • Reply clem Feb 9,2016 13:52

      Hi,

      These apps will make their way to LMDE in the coming months.

      There’s no need for office apps. LibreOffice is doing a great job and we’re already using that suite in all editions.

      So far, we can confirm the apps will include:

      – a text editor
      – a PDF reader
      – an image viewer
      – a media player
      – an archive manager

      When it comes to picture management, gthumb is no longer usable, but we might simply choose an existing alternative here and there might not be a need for an X-App for this. It’s too soon to say, it’s not decided yet.

      We’re also looking at small utility programs (calculator, scanning tool, dictionary etc etc..) and administration tools.

      • Reply mikef90000 Feb 10,2016 00:13

        Clem, that’s a good short list.

        One feature glaringly omitted from too many Linux apps is a Multiple Tab option! Opening multiple images or PDFs from the FM results in multiple app instances and windows – ARGGH! Please consider adding this feature early ….

        WRT to an X-App image viewer, I’ll wait to see how far you all want to take it. IMO the basic image viewers on Linux are too limited, and the ‘photo’ editors usually have some annoying feature or limitation. Perhaps the xfce devs can be persuaded to upgrade ristretto ….

  30. Reply gunvolt Feb 9,2016 20:55

    What you should do is actually fork every GNOME tool that Mint uses by default. At some point they’ll all be conforming to GNOME’s insanity.

  31. Reply Sanders Feb 15,2016 18:27

    Sadly, at some point what will be needed is is a fork of GTK3 itself, eventually it will be necessary even if it is not convenient now.

    At some time in the future the Linux community needs to evaluate if it is the right time, to once and for all, have a GUI toolkit that developers other than those involved in the Gnome project can depend on.

    In my case I want to write Linux applications, not Gnome-only

    The reality is that The Gnome project has set the Linux application ecosystem back 5-6 solid years, and at times it feels like what they really want is to make their own operating system.

    • Reply clem Feb 16,2016 09:56

      I’m pessimistic because as you said, they’re not really developing a toolkit, they’re developing a product, and so are we. That said, there are many reasons why I’m not comfortable with the idea of a GTK fork. We talked about it a lot, especially within the MATE team where some of the GTK and Glib changes across the years really irritated some of the developers. Say somebody forked GTK. We’d end up with two separate toolkits each DE has to support. 6 months later, their themeing engines would be incompatible (GTK is incompatible with its past selves, almost every release)… so we’d need themes to support both, and cause many of them wouldn’t.. we’d need DEs to let you configure each toolkit separately. On top of that, even if MATE, Cinnamon, the X-Apps, the Mint tools, MDM and all the projects we’re associated with switched overnight to that new toolkit, you’d still have many apps using GTK.. if the main goal was user experience and consistency, it doesn’t solve the issue, not globally.

      It’s sad, but the most efficient way a product distribution (such as Mint or Ubuntu) has to work around the shellification of GTK is by freezing GTK/GNOME and/or by patching it. That only solves the issue for that particular distro, and only short-term, but it does it well. It’s not great long-term though, and it also means GNOME isn’t GNOME in these particular distributions.

      Another solution, which doesn’t solve it all, but which I like much better, is to develop widgets and libraries to supplement GTK. Say lib-blah provides dialogs which look traditional and GTK makes all your dialogs look like they’re out of GNOME Shell. We can ship such a lib across the Linux landscape, for distributions to ship and developers to depend on. We make do with the GTK changes and constant regressions (that’s already the case anyhow)… so that lib would have to constantly evolve and adapt to newer GTK versions. This is being done more or less in isolated manners. There’s a growing need for this. As a developer you’re using both GTK and lib-blah, and whenever a GTK component looks too shellish you can switch to a lib-blah widget which does the same thing but doesn’t change the way your application looks.

  32. Reply Andrew Feb 18,2016 22:42

    Google Translate:
    Hi, after the release of Linux Mint 18 will be updated release LMDE 2.1?

    • Reply clem Feb 18,2016 23:12

      LMDE 2 is continuously updated so it doesn’t have point releases. You’ll see the X-Apps, Cinnamon 3.0 etc etc.. all land into the Betsy repositories. At some stage we might decide to update the ISO images themselves, but that isn’t related to any particular release.

  33. Reply R2D2 Feb 20,2016 16:16

    Hello,

    Linux Mint 18 Sarah “Alpha/Beta release” is available for test ?

    • Reply Reorx Feb 22,2016 20:45

      LM18 Sarah will be based on Ubuntu 16.04 which will be released in April (2016). It (LM18) will not be available until some time after Ubu 16.04 is released…

  34. Reply Rob Feb 23,2016 22:25

    Off track here, new to Mint, really enjoying it, struggling with networking, is there any way to make this more straight forward? Particularly when you have Win users that won’t shift over on your sharing network….

  35. Reply linuxero Feb 25,2016 08:02

    Hi Clem,
    Will be something like a banshee player or rhythmbox player (with support for images in .ogg? It could be called “mintPlayer.” :-D I like me your new text editor. WoW :-)

  36. Reply Yan Pas Feb 27,2016 19:07

    High Clem! Really good news. I’ve tested latest ubuntu 16.04 and found that a lot of cool apps were gnomified. For example gnome-logs has become awful, it lost half of functionality… and there is no gui app for reading logs now.

    Do you plan to carry ALL gnome apps to X-apps project? Freezing old code base would be enough. It’s painful to get updates from gnome devs, I only obtain less features each new release. All gnome gtk2 apps work fine for today.

    I think community would only benefit from it. Plus you may collaborate with xfce, mate devs.

  37. Reply simo Feb 28,2016 23:31

    Great work Mint team! Linux Mint 18 looks very exciting thank you Mint team to your idea and vision.

    – I think, the x-pdf reader should have a pretty full and customizable toolbar, ie: open, print, search, first or last page, pagen number, previous or next page, page history, zoom in and out, fit page, fit width, display mode (single page, double), continuous, snapshot, select text, hand.
    – Regarding nemo, the zoom level of 150% for the icon view is blurred, this zoom does not display the size 64 * 64 with sharpness, it is a little larger. In addition, the space between the columns (of fils) is very high. These two items are original in nautilus. But gnome has resolved them in 3.18 and 3.20. I hope that cinnamon 3.0 will do the same thing for nemo.
    – In system settings of cinnamon 2.8, you added the “input method” parameter which is a repetition of the parameter “language” ???? personally, I can not find a reason.
    – I think, Linux Mint also needs its own black theme.

  38. Reply Aubrey Mar 3,2016 14:32

    Clem,

    One area that I see that Linux in general is lacking is in OCR tools. Maybe this would be an area for XAPPS to fill a need.

  39. Reply Dennis Mar 7,2016 22:35

    Hi Clem,
    Nice idea but totem is an awful player, as most Linux D.E. media players are. They either don’t play disks, or they don’t play from folders, only single files and most don’t rip disks or record and non of the D.E. ones have a stop button, this drives me nuts ! ! VLC player is the only one that is free of these flaws but the only thing is it doesn’t auto play.
    If your going to do this and I applaud the idea but lets have
    something Fresh and New that’s fully functional, has a stop button and looks really great to boot (if you’ll excuse the unintentional pun)!

    • Reply ZenTiac Apr 7,2016 13:29

      Hi Dennis,
      I always go open with then choose VLC and tick make it default. Do this for all formats you play just once and it will then auto open with VLC. Unfortunately different people have different preferences. Hope this helps. Sorry if it is not a reply you needed.

  40. Reply Alex Mar 17,2016 18:06

    Just ship it with VLC and don’t create an app for videos ( if it’s possible to ship it with vlc legal reason I don’t know).

  41. Reply TheSwedishLitkeMan Mar 17,2016 22:45

    Will the media player support playing music and videos from a upnp/dlna server?

  42. Reply ZenTiac Apr 9,2016 00:56

    I personally love the idea of X-Apps. Well done.

  43. Reply Xis Apr 16,2016 15:35

    I really love the idea of X-Apps, but i can’t avoid thinking on this:
    https://xkcd.com/927/
    I hope it helps at long term to decrease splits and increase merges, but time will tell…

  44. Reply Sam Apr 25,2016 00:38

    Hi Clem,

    Have you seen clementine music player?

    I personally find it to be one of the best, because of all the great integrated features. This maybe something for the far future, but I think it’d be great if this integrated into the mint environment. Some features I think that set it apart:
    1. Automatic metadata tags : fetches them online
    2. Integration with Online Services : Spotify, Soundcloud, you name it.
    3. Online search for tracks.
    4. Option to play tracks from library vs folder (it has a view where you can see tracks in album/artist/track fashion, or simply folders from your disk, which is pretty great for some people)
    5. Artist bio : One click fetches the artist bio, and pics from wikipedia I believe.
    6. Integrated lyrics fetching : one click is all it takes to have the lyrics of your track.
    7. Focus on playlist over library : double click tracks in your library to create the list you wanna listen to now.
    8. Android Remote : to control Clementine with your phone!!!!!

    I’m a big proponent of theirs, much like I am for linux mint.
    https://www.clementine-player.org/

    Their only drawback I feel is a bit of a tacky interface.

  45. Reply Stringent Apr 25,2016 17:03

    I was actually wondering if this set of x-apps is going to include an audio player?
    Is it maybe the xplayer (based on Totem) going to take the role of an audio player the way it does (has the basic functionality, ie. can play files, but without that look&feel of audio player), or maybe another player could serve as basis: Clementine (full blown), Audacious (simplicity), Rhythmbox (traditional)…?

    Getting rid of Banshee would be an opportunity to get rid of Mono altogether :)

    • Reply clem Apr 26,2016 10:41

      Hi,

      I think we need two applications for audio. One for opening media files (whether that’s audio or video), and for that we used totem (which is now xplayer in Mint 18). And one for listening to and organize music within a collection, and for that we have Banshee.

      • Reply Stringent Apr 26,2016 17:37

        hi,
        well, understand that of course, but the question was whether an (x-)app, covering the functionality that Banshee currently covers, is envisaged in future?
        Would developing such (potential) x-app be an incentive to get rid of Mono, or the Mint team is perfectly happy the way it is now (Mono+Banshee being distributed)

        thanks

        • Reply clem Apr 27,2016 10:12

          There’s no plan for such an app at the moment.

          Regarding Mono it’s not considered an issue.

  46. Reply Tim Apple Apr 27,2016 17:42

    Hey Clem,

    I was wondering if x-apps have their, or will have their own site/blog or whichever to follow progress more closely?

    Cheers,
    Tim

  47. Reply bluedxca93 Apr 29,2016 13:52

    Hi, will there be an engrampa-nemo extension and an calculator app without titlebar header bars?. Or can an older file-roller (like 3.6.3) be forked and integrated into nemo?

    Epiphany Webbrowser has these damn bars so long already.
    Have a compiling and mostly working fork of older 3.3.3 Version on my harddisk.
    But it isn`t using the new webkit like it schould etc. maybe i could post it up on github.

    In fact seeing the revisions most changes are only on gui.And guess its like this for all applications.

  48. Reply Bard May 9,2016 16:33

    I have tried out xreader and xviewer on Arch. Great job! Will you add the browse tool (hand) into xreader like in okular etc. in the future? In my opinion it’s a killer feature that i missed in most pdf-reader.

  49. Reply Daniel Eriksson May 14,2016 13:34

    I really appreciate this project and the fact that you will provide good apps with a more classic design, which I consider superior to many “modern” designs. Both these apps as well as Cinnamon and Mate looks really great and I’m considering switching to one of them soon. Thanks for all your good work!

  50. Reply dave May 18,2016 08:19

    Hi Clem,
    Have you considered this project from Korora team?
    It’s a lightweight environment – agnostic for constructing graphical user interfaces.
    Here some info: https://github.com/kororaproject/kp-lens
    https://kororaproject.org/about/news/lens-an-alernative-to-desktop-agnostic-uis

    • Reply clem May 18,2016 09:00

      Thanks Dave,

      That’s pretty cool. We use something similar in mintinstall (and back in the days also in mintwelcome), with python talking to webkit and using it to render the UI. This is at play also in MDM. It’s never been abstracted like this before though. We need to work on an upgrade mechanism for Linux Mint 18, this might come in handy.

Leave a Reply