Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    Jack Ostroff
    @ostroffjh
    Minor oddity - something in the build or run changes line 103 of data/addin-xml-strings.cs sticking "Tag" in the middle of the line. It's just in a Catalog.GetString, so it doesn't hurt anything. It's happened several times recently.
    Stephen Shaw
    @decriptor
    @ostroffjh you said that you had a VM you were testing this in, right?
    Could you create a small VM with the extension node problem and then share the VM with me?
    Jack Ostroff
    @ostroffjh
    Let me see what it takes to get it into a transferrable state. I don't think I've ever done it, but I think there is a way to export. Give me a minute.
    Jack Ostroff
    @ostroffjh
    Updating ubuntu first, then will change root pass before sending. Not sure how I'll transfer it to you - depends on exported size.
    Jack Ostroff
    @ostroffjh
    Looks like the export is just about 10G - not sure how realistic that is. It's Ubuntu 18.04. I'll see if I can come up with a list of packages I have added myself.
    Not until tomorrow though, or maybe after the weekend.
    Jack Ostroff
    @ostroffjh
    Sorry - ubuntu has gotten too confused about what was manually installed and what is just a dependency. I'll try to work through it, but if you create a new VirtualBox VM, and install Ubuntu 18.0.4 (I'm running the Mint desktop in this one) the install git. Clone f-spot. You already know most of the things it needs you to add - but if you forget, then a run of autogen.sh or subsequent configure will let you know what you forgot. I wish it was easier, and I'll continue to think about anything better.
    Jack
    @genterminl
    @decriptor per f-spot/f-spot#164 after removing gtk-sharp-3 on my Artic Linux box, I now get the same Extension node not found in path: /FSpot/Sidebar error as in my Ubuntu VM. My guess now (and it's only that) is that there is something about the version of mono-addins being used.
    Jack
    @genterminl
    New insight - I found the mautil program (part of mono-addins) but running it gives me an array bounds error, but I haven't been able to tell if the error is in the mautil program itself, or some bit of local configuration it is tripping over. I'll continue to dig.
    Stephen Shaw
    @decriptor
    if you have a stack trace send it my way. I might be able to see something
    Christopher Hoskin
    @mans0954
    image.png
    I've managed to get my f-spot package working on Ubuntu 20.04!
    To stop it crashing on startup, I had to comment out the call to CreateAlternateRgb in BuildProfiles https://github.com/mans0954/f-spot/blob/ubuntu/focal/debian/patches/lcms.patch
    I think it was mono that was SIGSEGVing rather than f-spot, so a try..catch block didn't help. Not sure when an alternate RGB profile is required, or what the impact of not having it is, but superficial testing suggests everything is working fine.
    Christopher Hoskin
    @mans0954
    (Anyone wanting to use this package should be sure to backup their f-spot database ~/.config/f-spot first. Also be aware that it pulls in forward ports of a number of old packages that are no longer maintained.)
    Stephen Shaw
    @decriptor
    @mans0954 that's awesome. What version of mono are you using?
    Christopher Hoskin
    @mans0954
    @decriptor 6.8.0.105 - the version that's packaged with Ubuntu 20.04.
    Jack
    @genterminl
    @decriptor running "mautil info" in the bin directory gives me:
    System.IndexOutOfRangeException: Index was outside the bounds of the array.
    at Mono.Addins.Setup.SetupTool.PrintAddinInfo (System.String[] args) [0x00014] in <cda31b7a418d44299cf1cbced0cd1512>:0
    at Mono.Addins.Setup.SetupTool.RunCommand (System.String cmd, System.String[] parms) [0x00010] in <cda31b7a418d44299cf1cbced0cd1512>:0
    at Mono.Addins.Setup.SetupTool.Run (System.String[] args) [0x0005c] in <cda31b7a418d44299cf1cbced0cd1512>:0
    at Mono.Addins.Setup.SetupTool.Run (System.String[] args, System.Int32 firstArgumentIndex) [0x00018] in <cda31b7a418d44299cf1cbced0cd1512>:0
    at mautil.MainClass.Main (System.String[] args) [0x0036b] in <ad9065f288ac45e9a7c5537c7bdea7e8>:0
    Stephen Shaw
    @decriptor
    I'm going to try and get a decent ubuntu VM running. Virtualbox has been a bit challenging it seems, but I'm thinking about pulling some of my database conversion changes into master
    that aren't necessarily database specific
    Christopher Hoskin
    @mans0954
    Hello. Has anyone seen an issue where the 's', 'v' and 'delete' keys have no effect when typing in the 'Description' box beneath a photo in editing mode? It only seems to affect lowercase 's' and 'v', uppercase are fine, as is Backspace. Very strange.
    Jack
    @genterminl
    Can you check if they are being intercepted as shortcuts for commands which simply happen to be no-ops? I can't check as I still can't get a working version.
    Christopher Hoskin
    @mans0954
    @genterminl I see messages like (f-spot:1155498): Gtk-WARNING **: 16:44:12.483: gtk_widget_event(): unhandled event type: 907212224 when I press 's', 'v' or 'delete'. Pressing other letters doesn't bring up these messages.
    The 'Description' box is a GTK Entry https://github.com/mans0954/f-spot/blob/master/src/Clients/FSpot/FSpot/PhotoView.cs#L353 If I Place an Entry on a dialog e.g. ImportDialog then I am able to type these characters fine. Placing a second Entry box on the PhotoView also has these problems. So it seems to be something about PhotoView ?
    Christopher Hoskin
    @mans0954
    image.png
    I can set a breakpoint and get the above call stack when pressing keys other than 's', 'v' and 'delete'.
    Setting the breakpoint at the top of the stack i.e. in FSpot.Utils.DelayedOperation.HandleOperation, execution doesn't stop when pressing 's' 'v' or 'delete', but does when pressing other letters.
    Jack
    @genterminl
    Understand I'm really just guessing here, but it smells like something is intercepting those keypresses before they get to that Entry. I'm curious if line 341 might be relevant.
    Christopher Hoskin
    @mans0954
    @genterminl I'm out of my depth with this, so your guess is as good as mine. Do you mean https://github.com/mans0954/f-spot/blob/master/src/Clients/FSpot/FSpot/PhotoView.cs#L341 ? Tried commenting that out, but still no joy.
    Christopher Hoskin
    @mans0954
    Making progress. I've commented out just about everything with Handle in it from MainWindow.cs, and 's', 'v' and 'delete' are restored! Now to selectively re-enable until I figure out what's causing it.
    Christopher Hoskin
    @mans0954
    Christopher Hoskin
    @mans0954
    My current hypothesis is that win.ViewMode isn't getting set correctly, so that the handlers think we are in IconView when we are in fact in PhotoView. Pressing v in IconView brings up a larger thumbnail - 's' and 'delete' don't obviously do anything though. I don't understand what Changed?.Invoke (this); does though https://github.com/mans0954/f-spot/blob/master/src/Clients/FSpot/FSpot/MainWindow.cs#L816
    Christopher Hoskin
    @mans0954
    My hypothesis was incorrect. What I now know is that 's' is mapped to MainWindow.HandleSharpen, 'v' is mapped to MainWindow.HandleLoupe and 'delete' is mapped to MainWindow.HandleRemoveCommand. These mappings are specified in src/Clients/FSpot/ui/main_window.ui. When these two lines are commented out https://github.com/f-spot/f-spot/blob/master/src/Clients/FSpot.Gtk/FSpot/MainWindow.cs#L496-L497 then these key mappings do not affect text typed into the Description box. When https://github.com/f-spot/f-spot/blob/master/src/Clients/FSpot.Gtk/FSpot/MainWindow.cs#L496-L497 are not commented out, then these key mappings seem to be triggered.
    Jack
    @genterminl
    So now my question is whether all of this is simply a temporary state due to all the recent work, and will get fixed with ongoing code cleanup, or if it has always been like this and nobody noticed until now. I would think such shortcut interception should not occur when the focus is within a text entry widget, but I have no idea what the mechanism for that would/should look like.
    Christopher Hoskin
    @mans0954
    @genterminl I'm building off of where f-spot was on the 21st Jan 2019 https://github.com/mans0954/f-spot as I know that will give me something that builds into a working state and is reasonably stable and feature complete. There's a warning in the README about "the code base is in heavy flux right now" which has scared me off trying to pull in the more recent commits en masse. The problems been around for a while, but then disappeared for a bit, and then came back again. This is the first time I've had the time to properly dig into what's going on.
    Christopher Hoskin
    @mans0954
    So, it seems what's supposed to happen is that MainWindow.HandleSharpen checks to see if the focus is currently in a Gtk.Entry box, and if the current event is an EventKey. If both of these conditions are true, then it calls Window.Focus.ProcessEvent to process the current event, otherwise it shows the Sharpener tool. However, rather than processing the 's' keypress, the call to Window.Focus.ProcessEvent results in a message of the form (f-spot:285489): Gtk-WARNING **: 15:13:59.515: gtk_widget_event(): unhandled event type: 1502284000 and so the 's' keypress is lost.
    Christopher Hoskin
    @mans0954
    My current fudge-around is just to add modifiers to the keys which didn't have them: https://github.com/mans0954/f-spot/blob/ubuntu/focal/debian/patches/Fudge-key-conflicts.patch
    Stephen Shaw
    @decriptor
    @mans0954 @genterminl if you want to file an issue on f-spot/f-spot I'll try and see if I can figure it out and fix it
    @mans0954 Is there something I can do to help reduce the number of patch you have?
    Christopher Hoskin
    @mans0954
    @decriptor - I wonder if we could have a stable branch and a development branch? The stable branch would compile to something stable, feature complete and usable (assuming the dependencies were available) and would get PRs for minor bug fixes. Perhaps we might even make some 0.9.x releases from it. Refactorings, replacement of old libraries etc would happen on the development branch, with no expectation that it would necessarily build to something usable at a given point in time? But I appreciate that might just spread you even more thinly.
    Stephen Shaw
    @decriptor
    @mans0954 makes sense. I'm trying to lean towards that with branchs for dev work in general, but isn't entirely what you asking for. I was thinking about it recently and thought it might be nice to just get master back to a good solid stable state.
    I have it sort of working on Windows even and am starting to try and stablize it.
    If we want to focus on stabilizing master I'm all for that
    (sorry, I'm tired. I'm hoping that makes sense? :) )
    NilsGN
    @NilsGN

    Hi, I'm new here.
    I really like f-spot!
    I'm for now in Ubuntu 18.04.5 LTS.
    And f-spot version: 0.9.025
    There are two problems for me:

    1. I can't drag and drop file[s] from filemanager to f-spot window. Something happens but with no result.
    2. I can create tags - but I can't add an icon. If I do, try, f-spot crash.

    So, is there any solution?
    Can I add more information?

    Regards

    PS
    Is this the right place to state this question?
    DS

    Stephen Shaw
    @decriptor
    @NilsGN thanks for the feedback. Could I have you create a couple issues on github.com/f-spot/f-spot for me? I'm trying to update/refresh f-spot. It is going fairly slow unfortunately and work has been extremely unfriendly to my personal life lately so most things outside of work have been put on hold :(