Truly deleting ‘removed’ files from Lightroom

When you tell Lightroom to deleted rejected photos, it pops up a dangerous dialog box:

Screen shot of Lightroom dialog asking if you want to actually delete rejected photos, or merely lose track of them

Though it does explain itself well – i.e. if you want to actually delete the photos, you need to click “Delete from Disk” – the default option is that misleading “Remove” button, which doesn’t really remove the files at all – it merely makes Lightroom lose track of them. They’ll still be there on disk, wasting space forever.

And, you can’t directly undo this operation, so if you hit return a little too quickly, or misread the dialog at any point, you’re seemingly pretty screwed (if you have a Lightroom catalog of any significant size).

Luckily, there is a way to find these undead files – that doesn’t require you walking through every single file on disk one by one & comparing against Lightroom’s view of the world.

1In the left-side panel, under the “Folders” section, select all the folders and right-click on them (if you have multiple volumes listed under “Folders”, you’ll have to do this one volume at a time as Lightroom won’t let you select folders across multiple volumes simultaneously). You’ll get a contextual menu:

Screen shot of the contextual menu from right-clicking on an entry in the 'Folders' section of the Lightroom left-side panel

2Click “Synchronize Folder…”. A dialog will appear:

Screenshot of the "Synchronize Folder" dialog

You probably want to uncheck “Remove missing photos from catalog” (if it’s not already disabled) and “Scan for metadata updates”, as those are unrelated to the purpose here and have their own ramifications. Instead, just select “Import new photos” and “Show import dialog before importing”. Then, click “Synchronize”.

3Lightroom’s standard import dialog will now appear, and will slowly sort through all the files in the folder(s) you selected, filtering them down to just those that exist on disk yet are not tracked in Lightroom – e.g. all those rejects you accidentally “Removed” but didn’t really remove previously. You can now review those and see what you’ve got – it’s possible you’ll find in there media you didn’t intend to delete, but rather were somehow misplaced by Lightroom at some point.

You might want to, in the import dialog, change your preview generation setting to ‘Minimal’ in order to minimise import time & wasted preview generation. You could also choose to add some keywords to the imports, e.g. “to be deleted” or “recovered” or “undead”, if you’re not going to just immediately delete them anyway.

In any case, you can now import some or all the undead files. Importing them might seem counter-productive, since the goal here is to delete them – but it’s necessary for the final step…

4Once they’re imported, you can now immediately mark them as rejects and delete all rejects again – this time correctly choosing “Removing from Disk”.

So while it’s a bit roundabout, it does get the job done pretty quickly and easily. Now if only Lightroom would fix that stupid dialog to make the default option the one that actually does what you told Lightroom to do to begin with. 🙄

iCloud ‘Optimize Mac Storage’ breaks the Mojave installer

Yet another example of a really bizarre macOS bug that’s pretty inexcusable as a test escape, given it occurs with the default installation settings on a completely clean OS install.

In short, the Mojave update installer does not work (on High Sierra at least) if you have ‘Optimize Mac Storage’ enabled for iCloud Drive (System Preferences > iCloud pane > iCloud Drive Options… button > Documents tab > Optimize Mac Storage checkbox).

Specifically, the installer reports:

Installation requires downloading important content. That content can’t be downloaded at this time. Try again later.

…and indeed fails to download the actual Mojave update files (the installer app as ‘installed’ via the App Store is merely a 22 MB bootstrapping app, that downloads the actual image only after you run it & start the installation).

Even more obnoxiously, if you use the dosdude1 Mojave Patcher Tool to force-download the entire installer, as soon as it completes the 6.5 GB download and produces the ‘Install macOS Mojave’ app in /Applications, the system deletes the downloaded installation files out from under that app, rendering it just as broken as the official App Store version. Infuriating.

Aside: to be clear, turning off ‘Optimize Mac Storage’ enabled me to produce – and keep – a working installer as downloaded by dosdude1’s tool. I did not verify that it also fixes the regular installer as downloaded via the App Store.

I also ran into the “The recovery server could not be contacted” error message even before all the above, but thankfully that was fixable via the means normally prescribed online – running “sudo ntpdate -u”.

The Myth of The Web

The recent kerfuffle with Microsoft Edge vs YouTube was particularly interesting since while I have no specific knowledge of that instance, I certainly do have some cultural insight from nearly eight years working inside the so-called Chocolate Factory – though not on any web stuff, to be clear, so my experience is in the broadest internal sense.

Like everyone else last week, I was trying to determine how much intent or malice was behind Google’s actions, but with a marginally more informed perspective – or at least a relatively unusual one.

Permit me to first provide some larger context, though.

When I worked at Apple, back in ~2006-2010, I insisted on using Camino, because it was superior to Safari at the time (which, among other flaws, was particularly crashtastic in its early years – an attribute which thankfully is long gone, but has burned itself permanently into my emotional memory).

That choice to not use Safari caused periodic issues because some Apple-operated websites wouldn’t work properly with anything but Safari. When I reported those issues internally, the typical response was “we only support Safari”. From the perspective of Apple, once they had their own web browser, that was all that mattered. Thankfully it wasn’t a huge issue since the web wasn’t that important to day-to-day work at Apple, as they used native applications for most things (sidenote: I still miss Radar… I didn’t miss Xcode for a long time, until I was forced more recently to use IntelliJ). And certainly the world outside Apple didn’t care about this cute little ‘Safari’ thing, at the time.

My experience at Google was essentially the same.

The vast majority of Google’s internal websites do not work properly in any browser except Chrome. This is a very real problem since it’s practically impossible to perform any job function at Google without using their internal websites heavily, since Google is so dogmatically opposed to native applications. Google has worked extremely hard to [try to] make it possible to do almost anything through Chrome (often to the point of absurdity).

Ironically even Microsoft – whom I currently work for, via LinkedIn – are on the Chrome bandwagon, as some of their websites – that I am required to use for work – require Chrome.

Most interestingly – and distinct from Apple’s behaviour, where dysfunction in browsers outside their own was predominately based on actual functional differences between them – this ‘requirement’ to use Chrome is often not because of any actual, functional dependency on Chrome, but because Google’s (and Microsoft’s) web developers will specifically require a Chrome user-agent, and explicitly block any other browser. While this is easily worked around by spoofing the user-agent field – and is how I know that the purported Chrome requirement is usually a fallacy – it emphases the mentality at Google:

There is no web, there is only Chrome.

This is, I believe, the crux of the matter in not just this Edge vs YouTube issue, but with web development broadly in space & time. The vast majority of web developers don’t create content for The Web, they create content for a browser. One browser, usually.

I saw it unashamedly unfiltered inside Google, but it inevitably leaks out in time – through things like carelessly & needlessly crippling other browsers’ performance on YouTube.

While today that browser happens to be Chrome, before Chrome existed there was still always that browser – e.g. the 90s and much of the 00s was defined by Microsoft Internet Explorer’s dominance and the refusal by the majority of so-called web developers to create content for The Web rather than just Internet Explorer. (Of course, back then The Web really was almost synonymous with Internet Explorer, with ≥90% marketshare for many years, so at least it was more pragmatic back then than Chrome obsession is now.)

So, I’m actually sceptical that the YouTube team explicitly sabotaged Edge – rather, I think it’s just one of endless cases of web developers not really caring about The Web – ignorance & indifference, in other words, rather than [outright] malice. But just as caustic & dangerous.

What particularly concerns me today is that it’s not quite the same as the terrible 90s and 00s. Then, when Internet Explorer dominated, the vast majority of important websites were not operated by Microsoft.

Today, Google’s web properties are dominant in mindshare if not marketshare to the point of essentially being monopolies (certainly in the case of YouTube, at least) – way more than anything Microsoft ever did was.

More to the point, 90s web developers chose to develop for Internet Explorer exclusively – they were not coerced into it for the most part, nor firmly bound to that choice, because their corporate masters did not have a horse in the browser race and were pragmatically & unemotionally going for audience reach. A meritocracy was possible, and existed to a degree, and was essential to the rise of Firefox, my nostalgic Camino, and yes, even Chrome.

Google very much does have a horse in that race, and I know – from many years of experience inside Google seeing their unfiltered opinions – that they absolutely do want Chrome to become the only horse in that race. Not because of some comically-evil secret council scheming at the heart of Mountain View, but because they culturally & corporately just don’t care about anyone else. Modern Google is just as paranoid, fearful, power-hungry, and ruthless as 90s-era Microsoft ever appeared to be – Google want control, and the browser today is as fundamental to control as operating systems ever were.

Given all that, my fear is that there’s no longer a practical way for another browser to compete on merit with Chrome – anymore so than a third party app store can compete on iOS, for example.

Chrome is open source in the literal sense, but not in the more important governance & existential senses. The only way to give The Web a chance is to remove any corporate browser bias from the minds of the top websites’ developers – Google’s web developers. (Or, technically, to just supplant Google’s numerous dominant web properties. Good luck with that.)

This assuredly won’t happen anytime soon by way of government intervention, given the current U.S. political circumstance, but it is conceivable that Google themselves would perform this surgical separation voluntarily, for the good of The Web.

Sadly, I fear that’s unlikely in an era post-“Do no evil”.