The GUI bugs are a different category, for sure. I feel despair about the GUI, too, not because of automatically generated and submitted reports but because I find that GTK+ in C is really difficult to understand and use properly and interface to our engine.
My current concept for the rewrite, regarding the GUI, is to separate the GUI and the engine. The engine can run on its own, in its own process, unaffected by any kind of problem in the GUI. If the GUI crashes, then it can just restart and reconnect to the running engine. Separating into two processes also raises the possibility of writing the engine and the GUI in different languages (for example, the GUI could be implemented in Python, or in Javascript/Typescript) or domains (e.g. in a web browser). I have a of code written already (it looks like it's about 37000 lines of Rust) but it does not add up to a useful program yet. Probably, an equivalent of pspp-convert or pspp-output will be the first useful program. You can clone what I have from git://benpfaff.org/pspp. It's in the `rust` branch. The Git history is a mess, don't bother reading it, just look at the files. I do have a rewrite of the manual available for browsing: https://benpfaff.org/~blp/book/html/ On Fri, Jun 13, 2025 at 11:00 AM Friedrich Beckmann < friedrich.beckm...@posteo.de> wrote: > you are a brave man! I just made some first steps with embedded rust and > found it a new learning experience. I thought that this must be the feeling > of new students when you read compiler messages and they just do not mean a > lot to you. Then I copied the messages to google hoping to find something > reasonable. But I liked all concepts I found and were new to me. > > Many recent bugs are gui bugs quite often resulting in a crash. I have the > feeling to work on some pretty old historic stuff with the gtk3 concepts > where we are now at gtk 3.24.49. Maybe graphs could be rendered on a > webbrowser via plotly. It seems that the gui handling with all the signals > at various places are kind of very difficult to keep under control. > > > Am 13.06.2025 um 19:43 schrieb Ben Pfaff <b...@cs.stanford.edu>: > > > > It's impossible to keep up with the flood of "security" bugs that people > generate automatically and submit. I am not going to try. > > > > I'm working on rewriting PSPP in Rust, which will by itself eliminate > all of these and produce a better piece of software anyhow. > > > > On Fri, Jun 13, 2025 at 10:22 AM opensuse.lietuviu.kalba < > opensuse.lietuviu.ka...@gmail.com> wrote: > > Hi, > > Please look at these security bugs: > > • https://savannah.gnu.org/bugs/?67049 > https://bugzilla.suse.com/show_bug.cgi?id=1242838 > > • https://savannah.gnu.org/bugs/index.php?67069 > https://bugzilla.suse.com/show_bug.cgi?id=1243731 > > • https://savannah.gnu.org/bugs/index.php?67071 > https://bugzilla.suse.com/show_bug.cgi?id=1244376 > > • https://savannah.gnu.org/bugs/index.php?67072 > https://bugzilla.suse.com/show_bug.cgi?id=1244378 > > • https://savannah.gnu.org/bugs/?67073 > https://bugzilla.suse.com/show_bug.cgi?id=1243054 > > • https://savannah.gnu.org/bugs/?67074 > https://bugzilla.suse.com/show_bug.cgi?id=1243052 > > • https://savannah.gnu.org/bugs/?67075 > https://bugzilla.suse.com/show_bug.cgi?id=1243053 > > • https://savannah.gnu.org/bugs/?67079 > https://bugzilla.suse.com/show_bug.cgi?id=1243322 > > Any news? It would be nice if some PSPP developer could put some > comments in https://bugzilla.suse.com too. > > -- > > Regards, > > Mindaugas > >
