The GIMP wants out of my photo workflow

I’ve been a big fan of the GIMP for as long as I’ve owned a computer. It’s an easy-to-use, free and open-source image editor with loads of functions, and there are countless user-contributed plugins, brushes and filters. I’ve written a few posts on ways to use it.

However….(clouds obscure the sun and thunder crackles in the distance)…the latest and greatest GIMP, version 2.8.0, is a lot less nice to use for my most common edits.

Saving a JPEG is now considered an export, not a save. This sounds like a trivial change in terminology, but it has workflow implications. “Save,” “Save as…” or “Save a Copy…” will only save a native GIMP XCF file. Typing another file extension in one of those dialogs invokes an order to go use “Export” if I want that filetype. When I’m saving a JPEG photo with a couple of tweaks, for which it’s not worth saving intermediate steps, this throws me for two wobbles:

First, the “Export” menu item is not grouped with the “Save”-type items; instead it’s a long mouse sweep into another section, JUST PAST “Overwrite foo.jpg.” Clicking a tiny bit prematurely has seriously unwanted results.* I can overcome this by getting used to using the keyboard shortcut, although I would argue that since the menu is there, there ought at least to be an option to lay it out a little more safely.

What I can’t see a way around is that, each time I export a file, I face a popup asking me to confirm that I want to close the original file “without saving,” because I have not saved an XCF.

This is a huge sticking point, because apparently, this is very important behaviour. It protects the file from being inadvertently closed after saving in a format that doesn’t preserve all the info present in the native format.

Here’s a 2009 quote from GIMP developer Martin Nordholts:

“A lot of people over at GIMPUSERS.com want the old save and “export” mechanism back. The conclusion to draw from that is that they are not part of the user group we are targeting. We are not trying to make GIMP into an excellent JPEG touchup application, we are making GIMP into a high-end photo manipulation application where most of the work is done in XCF.”

So this has been in the works for some time. Also: Ouch. I’ve been blissfully unaware that, for at least the past three years, I haven’t been supposed to be using the GIMP unless my edits merit saving layers and masks, etc.

Not that anyone cares, but in my assessment, allowing me to select a file format in the “Save As…” dialog doesn’t have to bother those who want a failsafe against losing the extra data saved in an XCF file. A warning dialog can pop up when users select a non-native format, in case that wasn’t really what they wanted to do, with a “don’t show this warning again” checkbox to restore the old workflow for those who want it. Or the behaviour could be set in preferences.

I suspect such suggestions for compromise have been made numerous times. They are not being entertained: here’s a May 2012 forum comment on gimpusers.com by developer Michael Natterer:

“You can stop this crap, the save and export stuff is going to
stay as specified here:

http://gui.gimp.org/index.php/Save_%2B_export_specification

Period. This is the final word in this discussion.”

Hm.

So for now, what are my options? Perhaps I can write a plugin that lets me close a file after export without nagging me to make an XCF copy. Maybe there’s some other software I can switch to for my routine photo processing.

As it happens, I already regularly use two programs explicitly intended for photo “development:” Darktable and Digikam.

I started using Darktable for RAW processing but it will take JPEGs as input too. I find the user interface a bit tedious at the moment: the plugins in the plugin palette are represented by highly-stylized and sometimes non-obvious icons, so I have to run my cursor over them all and read their identifying tooltips. I also find the plugin group tabs a bit awkward to navigate; they get crowded and require a lot of vertical scrolling. It allows presets to be saved, and will do batch export. There isn’t (yet) any sort of tool for drawing selections and masks, or painting in adjustments (like dodging and burning), but there is a spot-removal tool. It’s already a sweet application, and it’s under active development.

I use Digikam to tag and view my photos, so using its built-in editor could have workflow advantages. It’s now pretty fully-featured (and handles RAWs). Files can be kept track of as versions of the originals. Again, to my knowledge, no area selections or masking. Periodically, I check out Digikam’s editing and batch-processing capabilities. Thus far, I’ve always returned to my comfort zone (the GIMP) to just get the job done. One thing that could help tip the balance back to Digikam in more cases is to be able to save sets of operations and their presets in the batch queue manager for later recall (this is bug 251126 on the KDE bug tracker).

A major advantage I see to the GIMP is its curves tool. I haven’t got the knack for manipulating curves quickly in Darktable, where the positions of the markers on the horizontal axis determine the resulting shape when the curve is dragged. More importantly, neither Darktable nor Digikam seems to have an equivalent function to the dropper tool, built into the GIMP “Curves” dialog, which allows me to sample a point in the image, and marks its position on each channel of the histogram superimposed on the curve. This allows me to really quickly zero in on the parts of the curve that correspond to the bits of the image I want to adjust, and it seems like an obvious feature for a curve tool to have.

For the sake of (relative) completeness, I had a look at the curve adjustment tool in Krita, even though as I understand it, it’s really meant as a painting program. As far as I can see, its curve tool has no dropper and operates only on single red, green, or blue channels, with no option to cover luminosity or value. Pinta, too, has the dealbreaking lack of a dropper.

Curve adjustment is one of my most common tweaks, so the GIMP is still the most comfortable tool to use. Darktable and Digikam are both close and improving all the time. Importantly, what I mean by improving and what the developers mean by improving are pretty well-aligned. Maybe there’ll come a day when I can stop using the GIMP for most of my edits, which, it seems, would make its developers happy.

What a strange thing I just typed. I still can’t get my head around this, even though I understand that driving away the users whose primary use-case they disdain is not necessarily the problem it would be for commercial software developers. We’re not customers, in the sense that I’d be Adobe’s customer if I bought Photoshop (which could only happen if I ran Windows or Mac OS), so the weight of my opinion has to be measured in units of something that the devs value. It looks as though maybe that would be some flavour of prestige. Will the GIMP be more prestigious when people stop recommending it to their friends for photo adjustments? I honestly don’t know.

* I think, after reading this, that the devs think someone using GIMP for editing JPEGs will want to overwrite their original. Does anybody really want to do that? For the record, I sure don’t.

 

Two things from my life today

When I was a kid I pushed the graphical capabilities of the available computers to their limits. I coveted the Commodore Amiga for its powerful graphics performance. I did black and white drawings pixel by pixel on an Apple Macintosh and printed them on a dot-matrix printer.

At some point we had a PC and a Mustek hand scanner (you rolled it across the page you were scanning and if you rolled too fast the image was squished). When we had enough RAM to scan an ink drawing and colour it, I was in there, rolling scans, stitching them together, and waiting while bucket tool did its job. The capability to do a gradient fill felt like real progress.

My three-year-old made this (aside from me writing my name in it at her request) tonight with the GIMP.

3 year old scribbles in GIMP

And now for something completely different. F made dinner tonight:

Steak on a plate

The presentation struck me as a bit Magritte-ish. I made him go get a pack of sugar snap peas from the fridge.

Coming soon to YouTube: Nym Wars 3D

Disclaimer: This is a post in which I really am thinking aloud. OK, not really aloud, unless you’re using a screen reader, and of course at that point, when you’re hearing it, I’m not thinking it anymore…but I digress. I’m not an expert on this topic so I may be missing the most salient arguments. Just be glad I’m not posting the entirety of process that got me to this point. Also: This post may be boring. Hmm, I probably shouldn’t include that disclaimer here, or I will start to feel like I have to put it at the top of everything I post. OK, feel free to start reading. Or not, if you think it’s going to be boring. As you wish.

When I read Wil Wheaton’s Google is making a huge and annoying mistake blog post, about Google rolling out a requirement that YouTube users “upgrade” to Google Plus accounts to continue rating videos, I was a bit bemused.

My initial reaction was to ask how Google could possibly avoid linking its services and the user data that go with them. If Facebook and Apple already link all their data, then can Google afford not to? Wouldn’t it be bowing out of the competition to serve the most lucratively-targeted ads? Wouldn’t that be tantamount to suicide? Google’s thing is ads.

Presumably they can’t just “upgrade” everyone’s “YouTube” accounts to “G+” ones that they can use for posting comments and +1-ing stuff all over the web if they want, because G+ requires you to give them more information.

And this is where I realize the obvious: linking YouTube to G+ brings the nymwars to YouTube. (This is, in my opinion, a great post on the nymwars).

As long as G+ is just a sort of “completely optional” social network, terms like requiring a “real name” that uniquely identifies the user simply make it less attractive to those who can’t afford to be identified (this, for instance, via Making Light).

Requiring the G+ identity’s “real name” to upload to YouTube surely conflicts with Larry Page’s boast (in his 2012 Update from the CEO) that YouTube “enables an activist in Syria to broadcast globally.” Unless he’s deliberately chosen as an example an activist for whom it’s safe for everybody to know who she/he is.

But from his same statement, I get the idea that not only does Page want a “beautifully simple experience across Google,” but that he’s now deeply invested in the idea of clearing out all our web personae from Google services (or worse, sorting and collating them) and making Google the definitive source of info on people. Applying behavioural data, so valuable in targeting ads, to searches for people.  Annnnd … becoming an authoritative verifier of identity (via)?

I don’t have any new insight on the nymwars, but I’ll put in two cents on one aspect: the fact that Bradley Horowitz so trivialized the concerns of users over the G+ “real names” policy, pretending it’s an issue with the “signup completion” process and happily proclaiming that the majority of the problem is sorted because a nickname can now be added to a G+ profile, sends a chill down my spine. If Google is up to something it doesn’t want its users to notice, well, the best course is probably to continue to try to deflect the questions just like that.

At any rate there’s a pretty fundamental clash arising between the way people use the web and the terms on which Google plans to allow us to use its services.

The price of Google’s services is going up.* The changeover is going to be pretty disruptive for content producers on YouTube who, like Wil Wheaton, have a stake in likes or subscriptions.

As for whether the brand can afford the move to consolidate all its users into G+ entities, I don’t know. Can annoyed users, including influential ones like Wheaton and Neil Gaiman, affect the course Google has plotted? Will Google services just become so uncomfortable that they start to shed users? Or will it be a storm in a teacup until enough of us are acclimatized to the warmer water in our fishbowls (the two metaphors don’t – intentionally, at least – refer to each other in some clever way – unfortunately).

As for whether Google can afford not to make the move, I wonder that too. What revenue model can Google turn to in the near term if it chooses to ease off on a core strength (acquisition and analysis of user data) and let Apple and Facebook outdo it in advertising?

The landscape can change fast through the jockeying of competitors for position (heh, more mixing of metaphors). If Apple decides to compete with YouTube, users of this service will have to create an account with Apple, if they don’t already have one, and there they’ll be, inside Apple’s garden, and Apple will know who they are.

Are there more creative, more disruptive (in that new, positive way) paths Google should be plotting? Are they plotting them, while I’m distracted by the old-fashioned negative disruption they seem to be effecting on their old properties? I’ll be interested to see how it unfolds.


*It’s been said many times that if we’re not paying for a service, then we are not the customer; we are the product. But that’s an incomplete picture, since we do pay for these services with our data, which is in our possession before we hand it to the providers. I’m sure that’s been said many times too.