Dick Pountain/Idealog 310/13:53 9 May 2020
When a couple of issues ago Editor Tim asked for tips for newly-working-at-home readers, mine was 'buy a Chromebook', which forced me to face up to how far I've drifted from the original Personal Computer Revolution. That was about everyone having their own CPU and their own data, but I've sold my soul to Google and I can't say I miss it. When first turned on, my Asus C301 Chromebook sucked all my personal data down automatically within five minutes, because it was all in Google Keep or on Google Drive. I do still have a Lenovo laptop but rarely use it, except via those same Google apps, and I don't miss the excitement of Windows updates one bit.
My love for Google Keep isn't a secret to readers of this column, and it only grows stronger as new features like flawless voice dictation and pen annotations get added. Remember I'm someone who spent 30+ years looking for a viable free-form database to hold all the research data - magazine articles, pictures, diagrams, books, papers, web pages, links - that my work makes me accumulate. The task proved beyond any of the database products I tried, with Idealist, AskSam and the Firefox add-on Scrapbook lasting longer than most. Those with long memories might remember how Microsoft promised to put the retrieval abilities I need right into Windows itself, via an object-oriented file-system that they eventually chickened-out from.
Keep's combination of categories, labels, colour coding and free text search gives me the flexible retrieval system I've been seeking, though it still isn't quite enough on its own: while it can hold pictures and clickable links they're not so convenient as actual web pages. For a couple of decades I religiously bookmarked web pages, until my bookmarks tree structure became just a unwieldy as my on-disk folders. Nowadays I just save pages to Pocket, which is by far the most useful gadget I have after Keep. A single click on Pocket's icon on the Chrome toolbar grabs a page, fully formatted complete with pictures and a button to go to the original if needed, so making bookmarks redundant. I use the free version which supports tags similar to Keep's labels, but there's a paid-for Premium version with a raft of extra archival features for professional use. And like Keep, Pocket is cross platform so I can see my page library from Windows or a phone.
Does the cloud make stuff easier to find? Within reason, yes. Save too many pages to Pocket and and, as with bookmarks, you've merely shifted the complexity rather than removing it. Sometimes I fail to save something that didn't feel important at the time, then discover months later that it was, and Chrome's history function comes in handy then. I use it most to re-open recent tabs closed by mistake (I have an itchy trigger-finger) but by going to https://myactivity.google.com/ I can review searches years into the past, if I can remember at least one key word. Failing that, it's plain Google Search or the Internet Archive's Wayback Machine, recently released as a Chrome extension.
My music nowadays comes entirely from Spotify, end of. My own photographs remain the main problem. I take thousands and store them both on cloud and local hard disk, organised by camera (eg. Sony A58, Minolta, Lumix), then location (eg. Park, Italy, Scotland). I've tried those dedicated photo databases that organise by date, but find them of very little help: place reminds me far more effectively than time. My best pictures still go onto Flickr, tagged very thoroughly to exploit its rather superior search functions (it can even search by dominant colour!) Pictures I rate less Flickr-worthy I sometimes put on Facebook in themed Albums which also helps to find them The technology does now exist to search by image-matching, but that's mostly used by pros who need to spot theft or plagiarism. I can only express what I'm looking for in words, like 'Pip fixing the Gardner diesel engine'.
What's required is a deep-AI analysis tool that can facially identify humans from their mugshots in my Contacts, recognise objects like tables, chairs or engines, can OCR any text in a picture (like 'Gardner' embossed on a cylinder block) and then output its findings as searchable text tags. It wouldn't surprise me if some Google lab is working on it. I do realise that were Google to go bust, or the internet close down then I'd be stuck with local data again, but if things get that bad then foraging for rats and cats to eat will probably be a higher priority. Right now my tip would still be, keep your feet on the ground and your data in the cloud(s)...
