Afterthought. Contemplating the above post a day later and thinking "yeah, right, that's typical me in hyperbolic mode when annoyed" but on refection would I change it?
No, I wouldn't; because I'm still annoyed and my frustration goes deeper than just GIMP and its developers, my issue is that the problems that plague GIMP also plague much open source software and we need to find solutions.
First, I'm a vehement supporter of FOSS and the need for more choice in the marketplace even if that means having more commercial software to choose from. As I see it, much of the key software (Windows, Android, MS Office, Photoshop, etc.) that monopolizes market now does so because of Big Tech's market domination—not necessarily because it's the best available.
It's the sheer domination of the software ecosystem by Big Tech that's the problem, thus good open source products such as Linux, GIMP, LibreOffice, etc. all struggle to get a mainstream airing irrespective of quality.
It's clear to me there are several important reasons why open source has not achieved a wider penetration amongst more general/less technical users. From that perspective FOSS software suffers from a fundamental and systemic problem which is it's principally developed by people who like developing code for its own sake who then open-source it for free.
Without the monetary incentive one doesn't have the right to demand of them the form software should take. Dictating how software should work and or what features it must have would serve little or no purpose, if they'd disagree they'd either ignore you or just pack up and leave the project.
The problem is obvious, FOSS developers are skilled and it's that skill that directs the way they'll tackle a project—thus the software ends up working their way and with features they reckon it should have—and more often than not this is in direct conflict with the way ordinary nontechnical users want it to work.
Despite my narky comments about GIMP's developers, I understand that in their technical judgment they had good reason why they omitted the Fade feature and I accept they are under no obligation to include a feature that from their perspective is out of place and does not logically dovetail with other features. Given also they're offering their time for free then there's really not much one can object to in the circumstances.
Nevertheless, that FOSS developers first and foremost develop software as they see fit and from their perspective and that that view often clashes with the requirements of nontechnical users poses a serious conundrum for open source software. The upshot clear: if ordinary users see no real advantage and or find it too difficult to convert to FOSS then it won't attract the number of users it deserves.
With commercial software user requirements are often determined by marketing based on user demand and programmers are obliged to program accordingly. When Microsoft says "we're about to start on yet another GUI for the next version of Windows" the choice is stark—knuckle down or leave. Without the monetary incentive, FOSS is at a significant disadvantage.
Let me offer my own perspective here. I am an IT professional and have been for decades. Moreover, I've headed an IT department where help-desk was a key service and from experience I know how difficult it is to get users to adopt new software packages, especially so when the replacement package offers essentially the same features, for example, replacing MSO with LibreOffice often meets with stubborn resistance.
Even I get annoyed when I have to change software and I find the UI and or features are noticeably different to what I'm used to. I'm an Arch user so I'm used to tweaking Linux et al but I deal with hundreds of different packages on Linux, Android, Windows and other setups and it's dead easy to forget the idiosyncrasies of a package that one hasn't worked with for several months. Having to relearn its features and quirks because its so different to what one was working on only moments before is not only time consuming but also a damn nuisance.
So from my perspective standardization is just about everything. Trouble is programmers all to often consider themselves 'artists' and unnecessarily keep reinventing the wheel. Look at the outrageous plethora of Windows GUIs. MS started out well by adopting IBM's well thought out CUA (Common User Access)† then essentially abandoned it for others that were often grossly inferior—think Metro for instance. And what about MSO's 'Ribbon'? The heretic(s) who devised that should have been condemned to time in the stocks. The unmitigated fucking hide of them to force millions of users to waste millions of manhours relearning functions that didn't need to be relearned. What is so outrageous was that after mandating the change to the 'Ribbon' those authoritarian shits at Microsoft could easily have left the old GUI there as a fallback option—but they didn't!
Who paid for those millions of wasted hours? Right, it certainly wasn't Microsoft.
This kind of outrageous and unacceptable behavior is why it's so important that FOSS be easy to convert to and be as compatible with the commercial product as is possible. With its Windows monopoly MS can afford to experiment with GUIs and weather the changes when they fail and still survive unscathed but I'd contend that in this regard FOSS is much more vulnerable.
In this regard we've witnessed a long lineage of failures in converting Windows users to Linux due to compatibility issues, despite the Linux brotherhood praising its simplicity the fact is that Linux is just too different to Windows for millions to make the leap. The problem is further compounded with the myriad of Linux distros available—and I'd venture that most Linux diehards have no idea the fear that that number of distros generates in newbies. FOSS developers should heed the message.
I am firmly convinced that the lack of consistency across FOSS development and its failure to adhere to uniform standards in the many instances where it's possible and logical to be consistent is a major impediment to its widespread adoption.
The other impediment slowing down FOSS adoption harks back to the fact that open source is free and without monetary incentives is that there are no guarantees that a FOSS project won't end up as abandonedware. I've long advocated a halfway revenue-neutral measure where certain FOSS projects only offer the source for free and charge a small nominal amount for binaries with licenses stipulating that one's only allowed to compile the source for one's own use—distribution thereof is prohibited. It seems to me most users would prefer to pay a small nominal amount of say $10 or $20 for the binaries. That way, developers could receive some recompense for their work, then they would have both the incentive to remain with the project and also likely to be more responsive to users' requirements.
I know I'm not alone in coming to these ideas, after I posted the above comment I came across this HN story that addresses very similar issues:
"Free software scares normal people"
https://news.ycombinator.com/item?id=45760878
I'd urge you to read it.
___
† https://en.wikipedia.org/wiki/IBM_Common_User_Access