Optimized For the Fast-Fading PastPublished 10 years, 7 months past
I have a theory, one that I’m sure has been formulated by someone else much earlier than me, that all power users eventually get left behind. They get stuck in a highly-optimized box canyon of their own making, one that is perfectly tuned to their way of working and interacting with data and is of interest to precisely nobody else in the world.
Let me use myself as an example. I’m currently running OS X Snow Leopard, 10.6.8, with no intention of upgrading. This is because after Snow Leopard, there is no more Rosetta. That means that my preferred personal mail client, Eudora, will not work. Neither will Word 2004. Both are, in effect, upgrade deal-breakers for me.
But why would I hang on to such relics?
Well, Eudora has been my mail client for quite literally two decades, and thus it has two decades of archived mail that I can search very quickly and easily. I have tried out migrations to other clients; they crash trying to suck in 3GB worth of mail text in Eudora’s special format. I could simply declare a break and move on to a new client with no stored mail, but as soon as I upgrade my OS, even the archives will be inaccessible. This is a major barrier. There are possible solutions, but trying them is incredibly time-intensive with no actual guarantee of success.
As for Word 2004, I have it customized so that ⇧⌘S shifts keyboard focus to the Styles combo box. There I can type the name of the style I want and hit return. This is really important when I’m writing a book whose files eventually have to be passed off to a publisher’s production staff, whose toolchains depend on proper use of styles. O’Reilly in particular went to a lot of effort, back in the day, to create style who had vi-style shortcut names, so I can highlight a few words and type ⇧⌘S fc [return] to set the highlighted text in the “literal” style (used for property names and the like). Versions of Word after 2004 do not possess this feature. I own Word 2011, and often use it to view documents sent to me by others, but I can’t use it as an efficient book-authoring tool because it amputated a feature I use a lot.
So the objection isn’t a simple “I like what I know, dadgumit!”, though of course I do like what I know (we all do). The real problem is “I have built my workflow around these things, and breaking them is unacceptable”.
I hear similar complaints from my designer friends. They’ve gotten so expert at using a particular piece of software that they bemoan even the hint that it will get a significant ‘upgrade’ — which often sounds like “break everything I do while likely adding a metric ton of crap I don’t need” to the power user — or even be discontinued. Although for the power user, discontinuing is often preferable; at least when software is discontinued, it works exactly as you expect for as long as you can keep it running.
The web doesn’t inherently fix this problem, either. When Twitter finally retired the API access points that Twitterrific 3 depended upon, my desktop Twitter client irretrievably broke. Why not upgrade to the latest Twitterrific? Because version 3 allowed me to display my timeline with all tweets collapsed, except for the currently-active tweet. It was an incredibly compact, high-density, useful interface. Version 4 does not permit it. no other Twitter client I’ve tried permits it. In fact, every other Twitter client I’ve tried has come off as cartoonishly clumsy and sprawlingly obtrusive when compared to the sleekness of Twitterrific 3 — including, as I say, the newer version of the very same Twitter client.
Granted, that’s more of a UI preference than a functionality problem, but UI preferences are often what drive us to use things, or not use them. I’m much now less present on Twitter than I was before the break, and when I do go on Twitter, it’s either via the official Twitter client on my iPhone or via twitter.com itself on the desktop.
Getting back to my increasingly-aging OS version, it helps that, to echo one my long-time personal heroes Tim Bray, I have no particular interest in what’s come after Snow Leopard. Dragging window edges might be nice, but I’ve lived without it for a very long time and rarely ever missed it. (Not never, but rarely.)
Yes, the newer OS X versions have a whole bunch of hip new cloud features, but in my case that’s actually a bug, not a feature. I instinctively distrust cloud-based storage for a variety of reasons. The security concerns are pretty significant for me, and for that matter having everything stored remotely is a good idea only if I have 100% reliable network access everywhere I go. Well, I don’t (and neither do you).
But of course the rest of the world is moving in a different direction, leaving people like me behind. That doesn’t mean that the rest of the world has gone mad, or is wrong to move in the direction it does. This isn’t a querulous demand that everything be frozen in the spot I like because if it was good enough ten years ago, it’s good enough now. That’s not how the world works. What I’m doing here, if I’m doing anything worthwhile at all, is documenting the point at which I came to the end of my box canyon, pulled out a guitar, and strummed a quiet ballad to the memory of my own forward progress.
As I say, I think all this happens to every power user at some point or another. We become enmeshed in a web of interlocking dependencies, and sooner or later lock ourselves into a particular place. The odds of it happening increase with age, but that’s less a function of biological age than it is elapsed time. The younger you start, the younger you’re likely to reach this point.
I will have to exit my canyon eventually, of course — but when, how, and why all remain very open questions, and I do not look forward to the turbulent transition periods that are likely to follow.
Special thanks to Tim Bray, Grant Hutchinson, and Jon Tan for their insights and feedback on this post.
While I am certainly one to leave old tech behind (sometimes gleefully) I totally get where you’re coming from.
I’m still using the official Twitter app for OS X, despite it not being updated for about 18 months, and despite that it forgets my account information every time I quit it (but only for my main account — not any of the others I use) because it has the quietest design, and it does exactly what I want it to do.
Eric, I totally hear you, but if you acknowledge that you’ll eventually “exit the canyon” — wouldn’t it be better to bite the bullet and do it now, while the methods of migration are still reasonably current?
As an example, I used to love an email client that was discontinued around Mac OS 8. I waited too long to export my data, and eventually ended up losing it. Even if it hadn’t been lost, it would be fully unreadable now, stuck in its bizzarro ancient proprietary format.
Suck as it might, I like to get while the gettin’s good.
I’m an aging power user. The only thing working in my favor is that I kill my macs in about 3 years from purchase on average. Same thing happens. Logic board dies and fixing is the same price as a new computer – upgrade time.
End result? A graveyard of old macs and forced data loss/upgrade occasionally.
I’ve had to let go of concepts like keeping archived email. I can live with that.
This is a very interesting post, though I find it most interesting for what it only hints at.
I generally consider myself a “power user” in that I use a computer all day long and seem to know more about every aspect of how it works than almost anyone I encounter on a daily basis. But I’ve never been a “power user” to the extent you describe here, of tweaking configurations and keyboard shortcuts and having everything exactly the way I want it. I’ve always eagerly updated to the newest OS versions and (when it’s not prohibitively expensive, as with Adobe Creative Suite) other software as well. I roll with the changes, embracing their inevitability.
Which leads me to the interesting bit only hinted at here. One of your primary reasons for hanging on to Eudora is so you don’t lose access to a two-decade archive of email messages.
Why do you need all of those email messages?
Maybe you (and I’m not really trying to pick on you here; I’m speaking in more general terms) write emails of an enduring quality far surpassing my own, and there was a time when I obsessively collected and archived all of this stuff too. But at some point about 5 or 6 years ago, I realized there was no possible reason I would ever want to look at those old emails again. In fact, especially since most of my email has long since become work-related, if I need to look back at an email that’s more than a year old, I’m probably doing something wrong.
So… the question then isn’t really why you’re holding on to 20-year-old email. It’s why are you holding on to anything? Perhaps I’ve gotten to a point where I’ve too fully embraced the ephemeral nature of being, but after a couple of decades of adulthood, moving around the country and switching jobs and having kids and everything else that goes with that, there’s simply too much to hold onto, and too little reason to do it.
That’s not to say I don’t have my lingering obsessions. The room directly behind me is filled with 30-year-old video game systems and cartridges. But I’ve let go of as much or more than I’ve held onto.
I’m not really sure how all of this should apply to you (or, “one”) confronting the inevitability of software obsolescence. But I guess one possible answer is that it shows that a shift in perspective is possible. I did it.
Interesting that three of you (so far) put forward the idea of abandoning data as a way to avoid getting trapped. I do actually use my mail archives, sometimes searching for the email addresses of people I haven’t written in a long time, sometimes searching listserv archives that are not available via the web. And sometimes it’s worth it just to be able to go back and see what I said about something years ago. I do have compelling reasons to search my own archives, in other words. I probably hit ⌘F a few times a week.
(That said, I acknowledge that I am a data hoarder, and it has an effect on my decisions.)
Data is only half the story, of course, as my anecdote of Word 2004 shows. Sometimes it’s much more about capability, and honestly Eudora is more capable than most mail clients available today. (I have other legacy dependencies in that area that I didn’t get into, because I knew they’d hijack my overall point.) Word 2011 is objectively (and subjectively) less useful for me than Word 2004. To drop Word 2004 would be to seriously retard my efficiency at writing. Again, I may have to suffer that hit one day, but that day is not today.
Because while I could suffer the pain now rather than later, why would I do so? What I have works and works very well, so until there is literally no other option than to suffer the pain, there’s no compelling reason to do so. When that day does finally come, then I won’t be any worse off. To go back to the mail example, Eudora’s files, even if they are a touch quirky, are all plain text. So I can grep them once running Eudora no longer becomes an option, or possibly even write my own customized search engine.
(But then what happens when the language I in which wrote the engine gets an upgrade and obsoletes or drops language features I used to write the search? See also: Twitter API.)
Your mail question with Eudora is interesting. I guess you were on a pop server initially, then never moved it. I have been using email since 1991 and I do keep my mails. I was using Eudora too, but I moved often my mails without losing any bits of it from software to software.
Two very important steps of my ~20 years mail life.
1. IMAP: Switching from POP to IMAP. One day I decided to make the move. I created an imap server and manually moved my mail some bits by bits from the POP account to the IMAP account in the Mail UI. Once it is on the imap server… it saves you a lot of troubles for the future in terms of changing client.
2. DATED SPACE: Everyone think I’m a weirdo. But here it is. All the mail I receive in a specific month is going to one unique folder: friends, mailing lists, notifications, etc. EVERYTHING. The question but how do you find your mails. I do not use the folder structure at all. I use dynamic views based on search criteria. In Mail.app currently harnessing the power of Spotlight. The good thing is that a mail can be in more than one context such as the mails talking about webstandards.org or all mail from eric meyer. A mail will appear in two circumstances. Thunderbird has also dynamic views. Second and very important about your scalibility issue. A dated space folder structure means not that many emails in one folder. So filesystems cope easily with it. The imap server and the local copy on your computer.
And now… for the icing on the cake… My mail folder is… 15,05 Go.
5 times more than yours. ;)
So what I would do if I was you.
1. Create an imap account
2. Put all mails which are only local on the imap server.
3. Close the POP account
4. Eventually change your mail client when you want.
I think everyone eventually comes to the point you have. I know I did with e-mail some time ago – and with the same client, Eudora. I eventually spent a day researching solutions and after backing up about a half gig of mail I managed to convert the Eudora mail structure into that of Thunderbird using a separate tool and then imported it rather than just did a direct import.
I’ve also converted it to mutt as well. Yes I guess I’m a masochist but doing the conversions taught me a bit more about e-mail formats. The main site I used for reference was qwerky.50webs.com/eudorarescue/readme.htm
It’s a bit older now but Eudora’s file formats have never changed much over time and so should be at least partially relevant. Hope it may help.
The canyon – a very interesting analogy. I too find myself getting “canyoned” so I have developed a couple of techniques to, if not stop the process, then at least slow it down.
With every major upgrade of OSX I start with a clean install on a new hard drive. The I only install or transfer what I really need. I must admit to not yet having done this for Mountain Lion. I am still on 10.7 – canyon alert? Physically I start out with the new system as an external drive, work it up, then swap the internal for the external once I am ready to use the new system day to day. If I really need to go back I can boot from the old system. It is amazing how well this technique can separate “need” from “want”.
Ok, ok, I know it doesn’t solve Eric’s disappearing features/software problem but it is a way of moving on with a minimum of pain. And there’s more. As well as an historical stack of backup disk drives in the cupboard I also have on the shelf “hardware backup” – three working Mac laptops of various ages. If I really want to go back – for example all the way to Hypercard on the OS9 emulator under OSX 3 – “Panther”, then I can.
Unfortunately when it comes to programming languages I am thoroughly “canyoned”. I use Awk for everything and literally won’t use any electronic device that that doesn’t have it available.
I share your thoughts and pain.
For me it eventually come to point where I no longer have fun with computers. I did when I was using ZX 81, ZX Spectrum, Atari 65 XE, Amiga, but not modern computer — because they made me lose my habits. Like having big archive of emails. Instead I have Lotus Notes archive, Sylpheed archive, elm archive, and so on until current email client — Kmail from ancient KDE3. Because of this slicing, I barely even touch all those archives, or memories of people I was in touch some years ago.
And even all almighty open source software, or open standards data are not come to rescue, because they also are not fully compatible, or do not support given format… in short, more hype, than practical solution.
Another story is with the software — even if you somehow convert the data, you soon will find out, the app you picked, does not have keyboard configuration, it does not let you add filters, or custom buttons to toolbar, or anything else you are used to.
Cruel as it sounds, better be prepared to live with credo from Heat: “Don’t let yourself get attached to anything you are not willing to walk out on in 30 seconds flat if you feel the heat around the corner.” Same here, make an archive, change your habits, and move on.
Mail, for me, is not really a problem. I’ve gotten used to using Opera for my mail, and going to a dedicated client would be a step back in terms of immediacy, but the mail itself, going back almost ten years, is all on an IMAP server.
I do, however, have quite a bit of software which I could not live without. Opera itself, for instance, is customized just how I like it, and over the years I have had to change the way I do things as Opera has made The Old Way untenable from time to time. The benefits of upgrading have always outweighed the drawbacks, but there may very well come a time when this is no longer the case. That would be a dark day indeed.
In your anecdotes about Eudora and Word 2004 you say that power user equals to ability to customise your configs and that power users do it much more often than laymen.
I believe that I am a power user, but I avoid customisations as much as I can and in that I certainly increase my power.
But let’s get back to your story. Say, you decided to write your own search engine for your mail archives just to retrieve email addresses. In the end, like you said, mail archives are just text files. So
grepthem! Grep cannot be more stable and is unlikely to become a deprecated “feature” of OS X. It’s not as convenient as striking cmd+f, but why build customised solutions just to have a Spotlight function? You can wrap your grep into a script, use it in a Automator workflow and assign this a systemwide shortcut _if you wish_ (I wouldn’t do that). But in the essence it will stay the good old grep.
About Word 2004. I don’t know how exactly you customised the styles, but I am sure that other editors have shortcuts for styling as well. Pages does. Maybe Pages is not the best option, but there are other editors out there that will definitely get you out of the canyon. The questions is just: do you really want to get out?
I recently upgraded to Mountain Lion from a very old MacBook, and the number of issues the upgrade has caused is still growing. Entourage is out, Mail is in. I had to PAY to upgrade BBEdit and SnapZPro – ugh. Quicken won’t work, gotta figure that out. Every Office app I open has a complaint about the Office database. Dreamweaver hangs when I try to add a new property to a CSS rule. I can print but not scan and the Apple update that was supposed to fix it did not. I could go on, but you get the idea. On the plus side, zowie is this new baby fast.
Consider yourself to be a client. Would you consider hanging on to technology that old to be a plus or a minus?
You could setup a Virtual Machine that you use mostly for Eudora and Word. It would have to be Windows or Linux based, since you really can’t make a VM of a Mac (it’s possible but is more trouble than its worth). Then you could use your old programs in that VM while moving forward on your new machine. Just be sure to backup the VM, since it has your email client and emails in it. But if its too much trouble to use a VM, just face you’ll have to move on to new tools.
Great post, I’ve been wanting to write something like it for a while.
In my case, the canyon was created by a particular software development workflow, and someone else in the same position managed to find a workaround that provided forward compatibility. So don’t give up on progress, because it’s actually REALLY nice to be able to re-size windows from any side.
There should be a tradition in the industry (if not a law) that releasing the source code is the final step before cancelling support for an application. I’m not sure how difficult it would be to create an Intel build of Eudora (or Word 2004, ugh) using contemporary APIs, but nothing in software is impossible if someone can be motivated to do it. The limiting factor here is that there isn’t enough profit in the project for the corporate owner to invest in it, and so a dwindling number of users are stuck.
For that matter, who thinks it is impossible for Rosetta to run on Mountain Lion? Did Apple discontinue it because it is truly incompatible or insecure, or because they would rather devote their unlimited resources to making iCloud suck less? If the latter, then Rosetta should have been thrown to the developer community: fork this and make it work, good luck.
I suggest you take a look at an email database archiving program I recently started using to archive old email:
Mail Archiver X by Moth Software Mainz
I have all my email going back to 2000 archived and available in this slick program. If I’d been more concerned about keeping older email, I probably could have figured out how to keep it as well, It’s built in Real Basic using a Valentina database. I keep a couple of years ‘hot’ and store the rest in Mail Archiver X.
Reasonably priced at $35.95 USD too.
(Disclaimer: I’m a customer, not an agent or part of Moth Software.)
For my company, the reason we don’t upgrade that often is cost, time and training. Cost: we are a small media buyer with 15 computers. As soon as we start updating operating systems etc it gets expensive not just in software but in time. Also, not all of our staff will be able to learn new software themselves so again another time cost from another member of staff teaching them.
RE: Missing the bloody point | Bram.us
[…] the new tools that are coming. Don’t be afraid of the Big Bad Command Line™. Don’t become optimized for the fading past. Don’t get […]