Whoa!!! Now wait a minute! Whilst you suggest its adoption as a development of gtkdialog itself, that statement above is like accusation of theft. This is not avconv versus ffmpeg (!!!) and there were justified reasons for the decision (for now) to fork: basically the very many legacy gtkdialog applications, which would all need to be modified if, as rcsrn52 basically pointed out to me, would need to be modified if my simple but fundamental core change to gtkdialog was adopted in gtkdialog itself.disciple wrote:forking seems rather gratuitous.
Others could have contributed to the discussion, or involved themselves in the discussion that already took place if they had wished. Fact is, as I earlier states, I was just trying this for my own interests and possible needs and with no intention of forcing something into gtkdialog itself, since that would have serious consequencies for the existing shell/gtkdialog legacy app developers (in terms of having to modify their code, when they understandably wouldn't want forced to do so). The following is just an extract:
rcrsn51 wrote:I understand the issue. I was involved in the discussion when it arose several years ago.
But gtkdialog is a mission-critical piece of software. Before people start tinkering with it, I would like to understand the rationale.
So patching the gtkdialog binary is now off the table?wiak wrote:I didn't manage to get my desired change to work anyway.
...wiak wrote: I doubt if it was ever on the table for most people. I was just wanting a patched version for my own interest really and still do, if only to understand why gtkdialog doesn't accept bash -c "commandstring" (in <action> tags) as way of 'seeing' the export -f bash functions, when /bin/sh is dash Of course, were an improvement possible to gtkdialog without upsetting the apple cart more generally that would surely be good. I remember BarryK didn't seem to trust Thunor's first attempts to mod gtkdialog, but that is all water under the bridge now.
wiak wrote:Yes, that will be possible.rcrsn51 wrote:So "gtkwialog" could be added to Fred's repo as a separate item?
Then people who want to build new apps using it can do so, without worrying about side-effects on older gtkdialog apps.
Or people could gradually upgrade their gtkdialog apps to gtkwialog after proper testing.
Frankly, if there is to be that kind of comment/suggestion of being 'gratuitous' then simply best I stop doing this development altogether; no harm done, let others develop how they will. You may not have meant what your comment implies, but comments like that at the very top take away from actual development work time and just throw spanners in works, sorry.wiak wrote:Gtkwialog operates somewhat differently in one core/fundamental way to legacy gtkdialog, which can result in some if not all existing bash/gtkdialog scripts needing several modifications to use it
...
Being a fork with diverged functionality gtkwialog has been given a different code name hence users can choose to use either legacy gtkdialog or this new gtkwialog as they see fit. People who have legacy shell/gtkdialog apps , who do not wish to modify them, will want to continue using legacy gtkdialog. In time it may or may not diverge further.
EDIT: By the way, there is also an issue, that gtkwialog may evolve further, which means its operation may end up differing even more from gtkdialog; I have previously stated that I my intention was to keep changes to the minimum because I want the interface to be as familiar to developers as possible (to make it easier for them to use it if they wish), but with the changes made there are some little extras that follow on naturally in my own wishes. Actually, one of my considerations was to try and better accommodate legacy applications, not to protect legacy gtkdialog, nor to threaten its core position in Puppy Linux, but simply to make app porting to the likes of Pristine Debian simpler for developers (and, particularly, simpler for myself). Gratuitous? No.
Anyway, for anyone at all who does disapprove of what has taken place, not much has taken place at all - consider it an experiment then that can be forgotten.