I think it's less about aesthetics and more about cognitive load.

Nonstandard UIs aren't end of the world, but if possible I don't want the extra cognitive load of interacting with 5, 6, 7, whatever slightly different GUI paradigms.

Like most developers I'm using multiple applications at once. Typically I'm using a terminal, text editor, browser, Slack, email, and maybe a database GUI.

A developer's cognition and cognitive load are limiting factors. Increase my cognitive load by X% and there is going to be a Y% increase in how long it takes me to do something and/or a Z% decrease in the quality of the work. Even if X% is small, it adds up pretty quickly. After all if we are employed full time that is 2,000+ hours per year of work.

Again, not just aesthetics. It's little things like, "does this software support the standard keyboard shortcuts for my system?" If CMD+W closes a window in every app except one that's kind of a pain. Etc.

To be honest i find dbeaver well designed. Even though the interface is cluttered with a ton of functionalities, it has everything I need where I expect it to be! I don't recall googling "who to do X with dbeaver ?".

Counterpoint, I love DBeaver and have used it frequently. Did go through a scare with the schema viewer though where I thought that I had somehow deleted the entire table schema, after doing a backup of course. But it took some googling to find out that I had not committed the changes to the database, and also that there was an option to "lock" interactions with the database to prevent the kind of scare I had. So as an novice/intermediate SQL user, the clunky interface can be a detriment.

I'd generally agree with you. DBeaver interface is quite discoverable. I reach out for it whenever I need something more visual than pgcli - esp. things like ERD, inline editing, easy cross table navigation etc.

However, I'd also agree with the GP comment. Its use of SWT ui results in some rather weird oddities. The biggest one is that if we configure gnome to use dark mode in linux, we get this absolutely unusable UI with barely visible text in many places. It becomes very obvious in that case which of the widgets are coming from gtk and which are custom rendered - and the distinction between them is stark.

And there doesn't appear to be a clear way to fix it without actually disabling the dark mode at OS level. I'd be perfectly fine if it used some framework that rendered its own widgets that looked a bit different or even if it ignored the dark mode completely and stayed usable.

While your point is understandable there are various types of people. An important aspect about user experience is aesthetic and ease of use. Some people care purely about functionality and others have mixed opinions on this. It's not fair to call it petty when you guys are just two different customers and users with different needs.

People sometimes forget the importance of user experience and it's why some amazing software barely gets used.

Personally I care about aesthetic and consistency but willing to sacrifice depending on what I'm doing.

Some applications are designed so well, it makes me look for excuses to use them. Other times the good UI makes it faster and easier to get the job done. An ugly UI makes me less likely to want to use an application, and a bad/confusing one makes it harder for me to get the job done, especially if it’s used infrequently.

UI and UX does matter, even when talking about productivity software.

There can also be an element of inspiration and influence. If all my tools are well designed, I’m more likely to set a high bar for whatever I’m building. If I’m using ugly and poorly designed tool. The bar for whatever I’m making will likely drop to the environment I’m around.

Imo it's not just about being functional so I can get work done. It's also about enjoying what I'm doing and having aesthetic tools is important to me. It's still functionality the most important? Yes. Do I prefer non-ugly tools to ugly tools even if I had to trade a bit of functionality for a lot of prettiness? Yes, every single time (as long as I can still get done what I need to get done, ofc)

Just to add insight: to me, inspiration is important, and beauty inspires me. When I coded a lot back in the day, I changed the coding font from time to time, for example.

A fun representation of the need for such beauty is in the game of Rimworld. In this world, the characters you manage react to the beauty of their surroundings, they need it to be beautiful to a degree, similar to how they need comfort, and recreation. Mood is central to the game, and an easy way to improve it generally is to create a beautiful environment, even if the characters are going to do mundane work like butchering.

At the same time, I find it more pleasant to use other software. DBeaver does everything I need, but it’s hardly ergonomic.

I'd argue it's not really about prettiness but about fitting in, so that the UI isn't jarring or distracting. I think the situation would actually be similar if the rest of the system was ugly and one particular app's UI was pretty, rather than the other way around.

I have a PowerGrep license. It's UI is so cluttered that I almost always find ways to make due with less capable alternatives.

It's not about pretty so much as simple things like whitespace or other details. I would never release a UI that acted weird in any way. For instance, DB Browser doesn't scale when undocking. I have to close the app and open it up and then load the database again.

Another example is Total Commander. I would buy it if it worked they way I think and had proper whitespace instead of being infinitely configurable and using non-conventional UX defaults (hit R I should jump to the first file staring with R, FFS).

If I have to suffer through this shit then so be it. If I don't have to suffer through it...delete.

Underscoring the "HN comment" part. This is a userbase who are presently interfacing with a wall of text, some tasteful greytext and a coloured bar. This appears to be a crowd with low visual needs.