Craig Anderton explains why the dove of software‑reviewer happiness has flown out of his life, to be replaced with the turkey of user beta‑testing hell...
I used to look forward to having some new piece of software to check out. But now, whenever I get assigned some new program to review, I get a sense of impending doom and foreboding. Here's why.
To begin with, Anderton's new Rule Of RAM is that no matter how much you have, the next program you need to review will require twice as much. Ditto processors: my 166MHz Pentium PC was actually considered cool as little as a year ago. Now, it's "sub‑minimum requirements." And my Power Computing PowerMac was eliciting comments like "too bad you have such an old computer" until Power Computing took pity on me and let me trade it in for a newer model, so I could run the software I'm being asked to review.
More and more software demands that you essentially tie up the entire machine to run the program. Furthermore (and especially on Windows PCs) programs require certain system tweaks to work at their best (graphics acceleration on, graphics acceleration off, or whatever). Admittedly, one of the things I like about Windows is the way you can customise it for the task at hand to get that extra ounce of performance... but the problem now is that no two programs seem to want the same customisation. I finally gave up and installed a special hard drive just for reviewing Windows programs. I wipe it clean periodically, reinstall the system, make it the boot drive, and install the program from there and do whatever it wants me to do. At least that lets me maintain my normal boot drive, so I can get real work done.
The Mac side is definitely better, but not by as much as Mac loyalists would have you believe. Invariably, something that works under System 7.1 won't work under 7.5, and as for System 8... well, call me a coward, but I haven't installed it yet, and won't until everyone else finishes beta‑testing it. Then you need the latest version of some extension or driver for the hardware that works with the software you're reviewing (with which some other program on your computer is not compatible, so you have to keep the old one around anyway). At last, after you have all your patches and "nominal charge" upgrades sorted out (they are usually nominal charges — until you have to pay 17 at once), it's time to discover which extensions are fighting which, and why your computer crashes at seemingly random intervals.
Eventually, after a few fun hours (or days) learning about the emotional problems of computers, triumph — it's time to actually run the program! That is, if you can figure out the user guide, which is usually an eclectic mix of the printed manual for version 1.0, several loose‑leaf sheets of paper with updates and corrections for later versions, the special 'read me' file on floppy disk (which, if you don't read it before installation, causes the end of the world as we know it) and, of course, the on‑line documentation, which contains additional information not mentioned anywhere else.
But you get going, and you try the program out with your interface of choice, only to find some weird problem. The interface company says it's the software's fault, the software company blames the interface manufacturer. Or they tell you to go to someone's web site and download some patch, which typically has instructions like "to install this patch, first install the patch, then run it." Then, all of a sudden, you find you can't fax documents any more, or your system clock has been reset to the year 1834.
Nonetheless, being a Professional Software Reviewer™, you make however many calls to tech support you need to make, then move on and start working with the program. You make careful notes of its wonders and its flaws, and spend some quality time making it run through hoops. Finally, you send a copy of the draft off to the manufacturer for fact‑checking to make sure the price hasn't changed, and that you didn't miss some workaround or cool tip.
Invariably, the response is "Well, gee, we're up to version 1.75ab3.45gtx.0002 now, which fixes most of the bugs you mentioned and adds a few new features. Just go to our web site and download the update..." Then it's back to screwing up your system again, with yet more extensions, DLL files, and other sanity‑robbing goodies. If the computer successfully boots after re‑starting, you're fortunate indeed.
Dutifully, you rewrite the review in light of the new changes, and find out that now you have some other questions. It's back to the phone... "Yes, there is a conflict with (fill in the blank), but our next revision will fix that." Of course, the next revision is always slated for release two days before the deadline for the article, but actually arrives two days after the issue has gone to the printer.
Meanwhile, it takes another 30‑60 days before the magazine hits the shops, at which point you look like an idiot because the review is of something that no one can buy anymore, since they're now up to version 1.77cq.vwbug66.jb.007.
Of course, anyone who uses software — not just reviewers — has to run the same gauntlet. But what ticks me off is that you people out in Readerland install the program, and after eventually getting it to work, you're home and dry (until your animal brain overtakes your rational brain, and decides to upgrade, that is). In Reviewerworld, however, as soon as one program is wiped from the hard drive, you have to go through the same thing all over again with another review. Of course, there is an upside to this; in Reviewerworld, the magazine eventually sends you a small piece of paper known as a cheque, and you obtain the satisfaction of knowing that, thanks to your specialised knowledge, years of experience, and hard work, you have indeed netted something more than the national minimum wage. (Interestingly, one magazine now pays me extra for soundcard reviews... sort of "hazardous duty pay.") Of course, you have a tech support phone bill that will make phone company shareholders smile, and your system is still screwed up; but you'll get around to fixing all that soon... in fact, just as soon as you download driver patch update 2.3tu34iu5, which is going to solve all your conflict problems, once and for all...