(no subject)

[haptalaon]

The problem with programmers is that they have absolutely nothing in common with ordinary users of their programs, & it leads them to just...not understand how their tool is being used.

To me, I feel like programs actually should stay static. Anyone using your tool is likely to have made significant commitments based on the tool, and so delicacy is essential.

Updates are often keen to do something to the User Interface - like a landlord slapping a coat of magnolia paint on a death trap, to make it 'feel fresh' - not considering that really, the appearance is the last thing you should change. Because your users will like the current workflow (or they wouldnt be your users) and relearning it is wasted time and inconvenience; and at least some of them are going to have the big autism over a program which now looks wholly different. It's like a stranger coming in, taking my car away, replacing it with a new car and saying - "ta dah! this also drives!" - as if the aesthetic qualities of the original tool are interchangeable.

(We were in a car crash. My husband lost his big dopey bloke wagon and we've replaced it with a teeny tiny scarab jewel a third of the size, and I love it in a way I could never have imagined loving any car (ew).)

(Crary has a bit about this. When the internet is criticsed, some people respond 'but there have always been technological changes! this is no different, this is like the telephone and the train, you're just resistant to change'. Crary notes that digital technologies are different because they are continually in flux, and consciously being designed and redesigned so the modern person experiences permanent instability. There will never be a time when one is 'used to it', like one can get 'used to' a train or telephone, because digital technologies this week are in every way different than they were last month; and Crary notes this inability to 'keep up' is profoundly exhausting, as well as taxing politically as how can one ever organise to resist a fluid thing.)

& the same goes for removing functions. Because you can be sure if the function exists, someone will already be using it - and if it's no harm to keep it, it needs to be grandfathered in. Hey, let's quote someone smarter than me on this:

Stop Futzing With It

A corollary of that is 'just stop futzing with it'. Don't 'refresh' your interface. Gratuitous change is an easy way to break trust with your community, or break features you didn't even know you had.

On Livejournal, for example, every post and comment can have a little user icon next to it. Fans used to use these icons as a silent commentary on the comment they'd written. Sometimes the comment didn't even make sense without the accompanying icon for context. There were users who had amassed hundreds of user icons, and they would pick the one that best fit the comment. There was an art to it.

Until one day, Livejournal decided to limit people to 20 or 30 icons, and made the change retroactive. So they killed this feature of their site that they didn't even understand, and they destroyed a lot of data.

Shut Up And Listen

Which leads to the next point, 'shut up and listen'. People are using your sites in ways you might not understand. Why not have a look and see what they're up to? Or ask? People are amazingly forthcoming when you ask them, but it rarely occurs to site owners to do this.

Just because you wrote it doesn't mean you know what your site is for.

from Fan Is A Tool Using Animal, a wonderful essay.

Anyway, my linux laptop is still chugging away in its 15th year and does not have enough ram to play a YouTube video because no modern web designer cares about longevity either, so it's time to move JabRef over to my more functional laptop - JabRef, the database program I've been putting my Reading on for over seven years - but in the mean time, they've decided to do a little update and...

• It looks different now! It looks like a Duplo set instead of a lovely Win 97' Brutalist tool for grownups. It's all padding and pale text
• You can customise the look with CSS, but I shouldn't have to do that to keep it looking the way I'm familiar with
• In keeping with contemporary trends, it now has less features and less customisation
• Does my painstakingly smart automated make-this-into-a-website coding work any more? I've not checked yet.
• They've removed a handful of database fields, because why not. Was anyone using those fields? The developers don't seem to have considered that or care, but *I'm* using those fields and so that means data loss
• The developer involved opined that "If people don't like it they can use another program" - which, while true, is especially shit for a database program. Databases are longterm things - if you invest your data in them, they can't abruptly change what data is stored or how it is accessed - it's like a broken promise. Strangers don't owe me anything, and yet reliability is a bedrock quality for databasing software. Otherwise it's like storing your books in a soggy box. It fundamentally undermines trust in the tool (for example: which fields can I rely on now? i don't know which fields may be taken away in future.)
• Fun fact about that: today my husband brought out our old Magical Creatures encyclopedia and it has a massive, glorious monster mould all the way down the spine.

So yeah, I can't believe I specifically invested care into an offline program to stop this sort of bollocks happening, only for it to happen anyway. It's a serious inconvenience. & there is no other referencing program which works the way I need it to, so I'm going to try and keep my vintage lap in commission, but it's a really arse situation. Wasted time and more risk, and needlessly because isn't it simplest to just leave things as they are?

I know this seems like such a nitpick, but it's just unprofessional ultimately - for a designer to decide that nobody is using a particular field in a database program and therefore that field will be removed from all further versions, and to decide that in a top-down way, and decide what happens to the data that used to be stored there. That's not redesigning the program, that's redesigning my data.