Anjuta 3.2.0 Fails to execute program when Glade UI file has unsaved changes.

FYI: I am able to build and run Anjuta 3.2.0 from the source download. I am pretty busy "using it" at the moment, so I don't have sufficient time to analyze it completely to find this problem for myself, but if someone points me in the right direction at least to identify the likely files involved, I'd be happy to see if I can fix this for you. I am an embedded systems programmer with 30+ years programming experience if that makes any difference. I can't seem to find any documentation which describes the code organization and with everything running in plug-ins, it is difficult to find my way around.

... it is especially difficult to find my way around now that Find-in-files is gone!

I could reproduce this and thus open an upstream bug - see the link above. I will add some hints where you can find the appropriate code there.

Another FYI: I have been using Anjuta exclusively on a fairly large project for the past 12 months. I am the only programmer on the project and it includes 85 source and header files with over 50,000 lines of code I have written myself plus 35 Glade UI files most of which incorporate custom active graphics (i.e. the custom graphical buttons are image widgets inside event boxes which "depress" when touched ) and the screens update accordingly. (The client wanted an "iPod/iPhone look".) There are literally hundreds of widgets on some of the screens. I also use just about every graphical widget known to GTK somewhere in this system. The product is designed to be used in a custom embedded system with a large resistive touch screen so the controls need to be large enough for use with gloves on. It uses sqlite for all data logging and configuration and hundreds of custom graphics for everything on the screen except the text (including custom graphical scroll bars). All user visible text is language independent and the chosen language can be changed by the user on the fly. However, I do not use any pango or similar gettext or system supported language system. I perform the language updates via a language table from sqlite and all writes to the screen are done by my translate functions via phrase indexes. I use Anjuta's Glade plug-in exclusively for all UI manipulation so I am becoming quite the expert GTK screen builder.

All this to say that if you would like me to test anything in Anjuta, I probably have an environment in which I could do that for you. Frequently IDE developers don't actually 'USE' the system they work on like I do. Making it is typically significantly different from using it. If you'd like some screen shots, I'd be happy to provide them.

That sounds impressive ;)

Our usual test object if anjuta scales is the anjuta project itself which is currently about 150k lines of code but it is pleasant to know that it works for others, too and we might want to come back to offer.

In general if you have any suggestions etc. it is easier if you directly file them in GNOME Bugzilla (shortcut in anjuta: Use Help->File bug) because more developers are reading the bugs there and I don't have to upstream them manually. Launchpad is nice to filter out some crap from, well, not so experience Ubuntu users but still upstream works better for us. You can also use the mailing list of course for general comments.