I currently have the best problem ever:

Our client applications and APIs are so fast, users are complaining they don't notice that the UI changed in between the time they clicked a button and their eye drifts to the adjacent component. So now we're having to build visual cues to account for that.

#KillingIt

Worst code review experience?

Hard to pick just one, but most were in a big meeting room with 4+ other developers not related to the project and with some playing Monday-Morning-Quarterback instead of offering productive feedback.

In one code review, the department mgr reviewed the code from a third party component library.

<brings up the code on the big screen>
Mgr: "I can't read any of this, its a mix of English and something else."
Me: "Its German."
Mgr: "Then why is 'Button' in English? This code is a mess."
Me: "I'm not exactly sure how I should respond, I mean, I didn't write any of this code."
Mgr: "Yes, but you are using it, so it's fair game for a code review."
Me: "Its not really open source, but we can make requests if you found something that needs to be addressed."
Mgr: "Oh yes, all this...whatever this is..<pointing again to the German>"
Me: "I don't think they will change their code to English just so you can read it."
Mgr: "We paid good money, you bet your ass they'll change it!"
Me: "I think the components were like $30 for the unlimited license. They'll tell us to go to hell first. Is there something about my code you want to talk about?"
Mgr: "<Ugggh>...I guess not, I couldn't get past all that German. Why didn't we go with an American company? Hell, why didn't we just write these components ourselves!?"
Me: "Because you gave a directive that if we found components that saved us time, to put in a request, and you approved the request. The company is American, they probably outsourced or hired German developers. I don't know and not sure why we care."
Mgr: "Security! What if they are sending keystrokes back to their servers!"
Me: "Did you see any http or any network access?"
Mgr: "How could I? The code is in German!"
Monday-Morning-Quarterback1: "If it were me, I would have written the components myself and moved on"
Me: "No, I don't think you could for less than $30"
Monday-Morning-Quarterback2: "Meh...we get paid anyway. Just add the time to the estimate."
Mgr: "Exactly! Why do we even have developers who can't read this mess."
Me: "Oh good Lord! Did anyone review or even look at my code for this review!?"
<silence>
Mgr: "Oh...ok...I guess we're done here. Thanks everyone."
<everyone starts to leave>
Me: "Whoa!...wait a sec..am I supposed to do something?"
Mgr: "Get that company to write their code in English so we can read it. You have their number, call em'...no...wait...give me their number. You keep working, I'll take care of this personally"

In they nicest way possible, the company did tell him to go to hell.

I do not like the direction laptop vendors are taking.

New laptops tend to feature fewer ports, making the user more dependent on adapters. Similarly to smartphones, this is a detrimental trend initiated by Apple and replicated by the rest of the pack.

As of 2022, many mid-range laptops feature just one USB-A port and one USB-C port, resembling Apple's toxic minimalism. In 2010, mid-class laptops commonly had three or four USB ports. I have even seen an MSi gaming laptop with six USB ports. Now, much of the edges is wasted "clean" space.

Sure, there are USB hubs, but those only work well with low-power devices. When attaching two external hard drives to transfer data between them, they might not be able to spin up due to insufficient power from the USB port or undervoltage caused by the impedance (resistance) of the USB cable between the laptop's USB port and hub. There are USB hubs which can be externally powered, but that means yet another wall adapter one has to carry.

Non-replaceable [shortest-lived component] mean difficult repairs and no more reserve batteries, as well as no extra-sized battery packs. When the battery expires, one might have to waste four hours on a repair shop for a replacement that would have taken a minute on a 2010 laptop.

The SD card slot is being replaced with inferior MicroSD or removed entirely. This is especially bad for photographers and videographers who would frequently plug memory cards into their laptop. SD cards are far more comfortable than MicroSD cards, and no, bulky external adapters that reserve the device's only USB port and protrude can not replace an integrated SD card slot.

Most mid-range laptops in the early 2010s also had a LAN port for immediate interference-free connection. That is now reserved for gaming-class / desknote laptops.

Obviously, components like RAM and storage are far more difficult to upgrade in more modern laptops, or not possible at all if soldered in.

Touch pads increasingly have the buttons underneath the touch surface rather than separate, meaning one has to be careful not to move the mouse while clicking. Otherwise, it could cause an unwanted drag-and-drop gesture. Some touch pads are smart enough to detect when a user intends to click, and lock the movement, but not all. A right-click drag-and-drop gesture might not be possible due to the finger on the button being registered as touch. Clicking with short tapping could be unreliable and sluggish. While one should have external peripherals anyway, one might not always have brought them with. The fallback input device is now even less comfortable.

Some laptop vendors include a sponge sheet that they want users to put between the keyboard and the screen before folding it, "to avoid damaging the screen", even though making it two millimetres thicker could do the same without relying on a sponge sheet. So they want me to carry that bulky thing everywhere around? How about no?

That's the irony. They wanted to make laptops lighter and slimmer, but that made them adapter- and sponge sheet-dependent, defeating the portability purpose.

Sure, the CPU performance has improved. Vendors proudly show off in their advertisements which generation of Intel Core they have this time. As if that is something users especially care about. Hoo-ray, generation 14 is now yet another 5% faster than the previous generation! But what is the benefit of that if I have to rely on annoying adapters to get the same work done that I could formerly do without those adapters?

Microsoft has also copied Apple in demanding internet connection before Windows 11 will set up. The setup screen says "You will need an Internet connection…" - no, technically I would not. What does technically stand in the way of Windows 11 setting up offline? After all, previous Windows versions like Windows 95 could do so 25 years earlier. But also far more recent versions. Thankfully, Linux distributions do not do that.

If "new" and "modern" mean more locked-in and less practical and difficult to repair, I would rather have "old" than "new".

Rant. (I love and respect all people! Especially developers.)

You frontend imbecils! I just can’t deal with you any more. I’ve had it.

Stop-inventing-new-components-where-there-are-fully-developed-and-working-concepts!

I mean. Just fucking stop! If I see another worthless datetime picker with an ”innovative” design I am going to hunt you down and freaking scream in your face.

And make fucking buttons look like tappable/clickable. It’s not fucking hard! Imbecils.

Oh, ooo, look at me, I am a frontend developer and I am in UX la-la land and what I am doing is sooo hard. Fuck off with your fucking moving gradients and n:th-child childish playground.

”Yeah, I exchanged the spinner…”

Fuck you. Your not contributing. Nobody cares! We’re not doing anything for the business by having a web which can be seen on a fucking telephone. EVERYBODY IS SITTING WITH SEVERAL GIANT MONITORS AND A FUCKING WORKSTATION FOR THIS. NOBODY ASKED FOR IT. AND YOU SPEND COUNTLESS HOURS ON IT.

”Yeah, I made the site work on ipad”

Please. Why? It’s not worth anything. Zero value.

”Yeah, the toggle component is now changed since we started to use the biddle-flipflup lib and it works almost the same”

No! NO! It does not work ”almost” the same. The psychology of the toggle is now wastly different. What was On before now looks like Off and it is fucking worse!!!

Imbecils. I hate you.

And no, I can’t do your fucking work! And I know that you do other non-ui stuff as well sometimes… but anyway… I have no interest to be in that clusterfuck that modern frontend is today. It was really fucking bad twenty years ago and it is just as bad today and you are not helping.

”I’ve improved the button so now it aaaaalmost does not look like a button. But I am getting there!”

Fuck you.

This is why I hate Windows:

For about a month now, I've been learning/working on salesforce, so my Macbook was enough for me.
Today there was a bug in something I built in iur Python backend, and since it has a dependency on windows, I booted the old guy up.

And this is what I see. For about 1.5 hours this went on.

Then it started, but system consuming 100% disk and 80+% CPU. Can't do a thing.

And when zoom finally opened (for a quick meet), the camera turned on halfway down the meeting, and then the system restarted on it's own.

Old man showing that same screen again for more than 30 minutes.

Since I have dual boot on this one, I hard-shutdown it using power button, and now boot into ubuntu 20. This works so beautifully (although it froze for about 5 seconds before popping up the updates panel, something I CHOSE to keep enabled). I try going back to Windows, and it's hell again.

Here I am now trying to set up a ec2 instance and setup the app source there so that I can debug with RDP.

And yeah, Component Object Model is a motherfucking bitch. Person who invented it should die. People who build apps leveraging this should die. Business leaders who say "Hey this app (built with COM) can solve the problem easily, so use this" should die choking on their own phallus. And developers like me who keep using this because "the last guy did this" should die too.

Microsoft and it's products are the death of sane people.

Fucking Gates. Its the same damn hardware.

I am currently looking for a DAW (Digital Audio Workstation), because my music projects are starting to get a little too complex for Audacity.

So I started looking for a good, easy-to-learn, ideally free program, and quickly learned that Avid now has a free version of Pro Tools called First.

So I go to their site and fill out the registration form to get the download. In addition to creating an account with Avid, you also need to create one with iLok, which apparently has something to do with how they manage their licenses. Kinda overkill for a free program, but okay...

I download the program (about 3gigs...), install it and try to start it. It gives me an error message about missing some service. Okay? I'm confused because I notice that an 'Application Manager' service has appeared in my tray, and when I open that I can log into my new account just fine. But it still doesn't work.

There's a link in the error message to the iLok website, and it looks like ai need to dowload and install another component. Why didn't that get installed with the program if it's required?

Hmm...

So I go to the iLok site, download it and install it. Pro Tools First still won't start. I realize that the PTF installer asked me to reboot, which I didn't do because: a) I always have a lot of windows open, and b) How often is a reboot ACTUALLY required? Why would you need to reboot?

So I (begrudgingly) reboot, and now the program seems to start initializing... but then it throws an error message about some plugin that it can't load because it doesn't work for the 64 bit version. Then... why are you even looking for it?
And then it says something like: 'I can't handle that, I'm just gonna shut down'.

What?

I try starting it again. Same error appears, but then it gets past it this time... Only to throw another error message about something else it can't load, and therefore it must shut down.

Deep breath.

Third time is the charm, the program actually made it to the project create/load screen! Huzzah!
So I look around a bit, but don't do much. It doesn't seem too intuitive to me, so I start watching some tutorials on YouTube from Avid themselves. It's a little late by now, so I don't get my hands dirty that day.

Next time I want to try out the program I start it up, still get error messages, but it does seem to initialize okay. But then the 'Create project' button doesn't react when I press it.

It turns out that the program takes a looong time to log in to the avid account, even though the manager service is running and logged in...

When it finally logs on I create a new blank project, but it doesn't ask me where to save it to. I see there is a counter saying 1/3 and looking around I find some info about 'cloud based projects'.

It would seem that this program only supports saving projects to the cloud, and you get only 3 projects total. Three. THREE?

Ahem...

I add an instrument track to my new project and select the one and only plugin, which is a synth. I don't see the plugin window, like in the tutorials I watched. I fiddle around with the windows, but I only manage to get the layout fucked up. There's a handy 'Window' menu, but none of the options resets the view. The main window is now sporting a WINDOWS FUCKING 7 BORDER! And partially blocking the view of the top menu.

Blaaargh!

Frustrated, I shut the program down and restart it. I now select one of the project templates (after waiting for it to LOG IN AGAIN!) in the hope that I might have a bit more luck with that starting point.

But when the template has loaded, out of nowhere, the program goes from maximized to windowed mode! And the fucking Win7 border is back again, still messing with the main menu!

FFS!

I get the sucker maximized again and select one of the synth tracks, and Lo and Behold! The synth plugin window actually shows up! But of course there is no sound produced when I play, neither with the keyboard or my midi keyboard.

Oh no, that would have been too easy.

I see some the meters moving when I play, but no sound is produced. I check the options menu, but find out nothing useful except for the fact that the program only support 48kHz sample rate. That's pretty disappointing when you have a 192kHz/24bit soundcard.

I'm done. This piece of shit software is NOT for me. It's bloated, complicated to sign up for and install, extremely limited and buggy as hell!

The final insult is that it takes 5 minutes to uninstall because there is no uninstall option in the so-called 'Application Manager' (of course fucking not!), and doing it through Programs & Features there are 5 (FIVE!!) different apps and services to uninstall, one by one.

0/10, would not recommend.

In flutter , there’s something called TextButton.icon. Which render a button look like this :

(👍🏼 Like Button)

But there’s this tiny twat decided to use countless of nested column in a nested row and containers just to create a fucking button! This particular class contains 1438 lines of code! Most of the code are redundant and nested fucking shit.

I want to punch this guy so hard but I do not intend to start a ww3 with china.

That means I have no choice but refactor it as I implement a feature requested by the product team, every components break. It is like a land mine field here. One changes , the entire application crash.

So there are useless mother fucking Sherlock fucking holmes kept telling me that “don’t worry about refactoring now , just complete the task.” , like seriously “how in the name of mother fucking god of all arseholes can I complete my task when I can’t change even one component?”

These people are fucking genius. Their intelligence resurrected Einstein and made him die the second time.

I see a lot of talk about complex numbers, and yet
for all they are worth, I have not once been
able to find an explaination on how to calculate
them by hand, namely the real component.
For example
(-5)**0.5
(1.3691967456605067e-16+2.23606797749979j)

2.23606 is obviously just the square root of 5, but where the hell did 1.369 come from?

Apparently no one fucking knows, and no site I've found gives a simple explanation for someone new to math in general.

"use a calculator", "hit a button",

How about no.

Company started automated testing recently, and the devs need to review the test scripts.

The tester assigned to my component writes script to trigger button click and nothing else to verify the result.

I couldn't​ even. I just left work for the day.

So... I take over this one ticket to test... the ticket mentions some visual component popping up when a button is clicked. It says there is a success and a failure message. The title of the story also mentions another functionality.

I start testing and some fellow QA asks me why I'm testing in this environment. Turns out, three people are sharing one environment and three different things are deployed...

I ask the dev whats going on because I heard there are multiple people deploying stuff...
He just tells me "oh, my changes are deployed I just checked".
I tell him that it's not about that but about communication and testing one thing at the time. Then I tell him, that I wouldn't test until his stuff is the only stuff there.

Some time later he hits me up again, now with the env to himself.

I test and quickly I see, that there is only the positive message even when I make sure that the backend is not reachable. I tell the dev what I found and he tells me "oh no, it's just the implementation of the popup thing, it's just frontend for now"...
I tell him, that the ticket should say so.
No answer for like 1-2 hours. Then I get an "ok".
End of the day.

Next day I come in and the fellow QA tells me, that the dev asked him to test the ticket.
I ask him if he changed anything about the scope of the ticket, he says no...
I'm like "ok... know what... begin testing and then tell him what I already told him".
So he's testing and then tells him again to update the scope.

Later in the daily the the dev's update is besically "they won't test my ticket..."

It would have taken him like 1 fucking minute to update the ticket...

The whole QA team was always trying to being helpful and even when the tickets where sometimes not 100% clear we always made it work... but now we are more and more going towards "MR does not meet ticketdescription, fix it" and "I don't care if its just a small thing... fix it and then come back to me"...

Seriously frustrating some times...

The beginning of my small app. Just a small question to the devRant community. As I am relatively new in Android app development I was wondering which layout is okay to work with?
The reason behind this question is that I have found that sometimes I'd move a component(e.g., button, text field) by a small fraction and boom! everything else is messed up.

The datepicker saga

Part one

So I begin work on a page where user add their details, project is late, taking ages on this page

Nearly done, just need a component to allow users to put in some date of births. Look for react components.

Avoiding that one because fuck Bootstrap.

Ah-ha, that looks good, let's give it a go.

CSS doesn't exist, oh need copy it over from npm dist. Great it applied but...

... WTF it's tiny. Thought it was a problem with my zoom. Nope found the issue in github.com and it's something to do with using REM rather than EM or something, okay someone provided a solution, rather I saw a couple of solutions, after some hacking around I got it working and pasted it in the right location and yes, it's a reasonable size now.

Only it's a bit crap because it only allows scrolling 1 month at a time. No good. Hunting through the docs reveals several options to add year and month drop downs and allow them to be scrolled. Still a bit shit as it only shows certain years, figure I'd set the start date position somewhere at the average.

Wait. The up button on the scroll doesn't even show, it's just a blank 5px button. Mouse scroll doesn't work

Fucking...

... Bailing on that.

Part 2

Okay sod it I'll just make my own three drop down select boxes, day, month and year. Easy.

At this point I take full responsibility and cannot blame any third party. And kids, take this as a lesson to plan out your code fully and make no assumptions on the simplicity of the problem.

For some reason (of which I regretted much) I decided to abstract things so much I made an array of three objects for each drop down. Containing the information to pretty much abstract away the field it was dealing with. This sort of meta programming really screwed with my head, I have lines like the following:

[...].map(optionGroup =>
optionGroup.options[
parseInt(
newState[optionGroup.momentId]
, 10)
]
)...

But I was in too deep and had to weave my way through this kind of abstract process like an intrepid explorer chopping through a rain forest with a butter knife.

So I am using React and Redux, decided it was overkill to use Redux to control each field. Only trouble is of course when the user clicks one of the fields, it doesn't make sense in redux to have one of the three fields selected. And I wanted to show the field title as the first option. So I went against good practice and used state to keep track of the fields before they are handed off to the parent/redux. What a nightmare that was.

Possibly the most challenging part was matching my indices with moment.js to get the UI working right, it was such a meta mess when it just shouldn't have taken so stupidly long.

But, I begin to see the light at the end of this tunnel, it's slowly coming together. And when it all clicks into place I sit back and actually quite enjoy my abysmal attempt at clean and easy to read code.

Part 3

Ran the generated timestamp through a converter and I get the day before, oh yeah that's great

Seems like it's dependant on the timezone??!

Nope. Deploying. Bye. I no longer care if daylight savings makes you a day younger.

v0.0005a (alpha)
- class support added to lua thanks to yonaba.
- rkUIs class created
- new panel class
- added drawing code for panel
- fixed bug where some sides of the UI's border were failing to drawing (line rendering quark)

v0.0014a (alpha) 11.30.2023 (~2 hours)
- successfully retrieving basic data from save folder, load text into lua from files
- added 'props' property to Entity class
- added a props table to control what gets serialized and what doesn't
- added a save() base method for instances (has to be overridden to be useful beyond the basics)
- moved the lume.serialize() call into the :save() method on the base entity class itself
- serialized and successfully saved an entities property table.
- fixed deserializion bugs involving wrong indexes (savedata[1] not savedata[2])
- moved deserialization from temp code, into line loading loop itself (assuming each item is on one line)
- deser'd test data, and init()'d new player Entity using the freshly-loaded data, and displayed the entity sprite

All in all not a bad session. Understanding filing handling and how to interact with the directory system was the biggest hurdle I was worried about for building my tools.

Next steps will be defining some basic UI elements (with overridable draw code), and then loading and initializing the UI from lua or json.

New projects can be set as subfolders folders in appdata, using 'Setidentity("appname/projectname") to keep things clean.

I'm not even dreading writing basic syntax highlighting!

Idea is to dogfood the whole process. UI is in-engine rendered just like you might see with godot, unity, or gamemaker, that way I have maximum flexibility to style it the way I want. I'm familiar enough with constructing from polygons, on top of stenciling, on top of nine-slicing, on top of existing tweening and special effects, that I can achieve exactly what I want.

Idea is to build a really well managed asset pipeline. Stencyl, as 'crappy' as it appeared, and 'for education' was a master class in how to do things the correct way, it was just horribly bloated while doing it.

Logical tilesets that you import, can rearrange through drag-n-drop, assign custom tile shapes to, physics materials, collisions groups, name, add tag data to, all in one editor? Yes please.
Every other 2D editor is basic-bitch, has you importing images, and at most generates different scales and does the slicing for you.

Code editor? Everything behavior was in a component, with custom fields. All your code goes into a list of events, which you can toggle on and off with a proper toggle button, so you can explicitly experiment, instead of commenting shit out (yes git is better, but we're talking solo amateurs here, they're not gonna be using git out the gate unless they already know what they're doing).

Components all have an image assignable to identify them, along with a description field, and they're arranged in a 2d grid for easy browsing, copying, modifying.

The physics shape editor, the animation editor, the map editor, all of it was so bare bones and yet had things others didn't.

I want that, except without the historic ties to flash, without the overhead of java, and with sexier fucking in-engine rendering of the UI and support for modding and in-engine custom tools.

Not really doing it for anyone except myself, and doubt I'll get very far, but since I dropped looking for easy solutions, I've just been powering through all the areas I don't understand and doing the work.

I rediscovered my love of programming after 3-4 years of learning to hate it, and things are looking up.

I’m working on a react codebase and company decided to add a new module.
Now im writing markup and css to ensure UX is smooth as designers thought of it.
Imagine my horror when I start to code and find out no matter what HTML tag i use, it’s been FUCKING OVERRIDDEN in the global stylesheet. AND STYLES HAVE BEEN OVERRIDDEN WITH !important

They’re also using Ant design as a component library. Guess what, default ant design classes have been overridden too. So i try to use ant design button or card, and bam, MAGICALLY SOME DESIGN FROM SOME SHITHOLE MODULE DECIDES TO FUCK WITH MY STYLES

On top of that, styles of parts of application has been written in SASS, some part of application uses bootstrap components some use third party components like tables and responsive grids to suit to their preferences. Some parts use handwritten css. Some parts use CSS IN JS and styled components. THE FUCK IS THIS GARBAGE!!!! THE FUCKING CODEBASE HAS A MIND OF ITS OWN!!!!!! YOU NAME A WAY TO ADD STYLES TO A COMPONENT, ITS THERE!!!
And the company’s management thought a “fractal” approach to maintain each individual view is “best” for SCALABILITY!!! HOW THE FUCK DID IT NOT CROSS YOUR DUMB MIND THAT FRACTAL APPROACH ALSO GUIDES TO HAVE ALL COMMON STUFF AT ONE PLACE!!!! THIS CODEBASE HAS DUPLICATE STYLES AND DUPLICATE CODE IN ALMOST EVERY MODULE!!!!

Not to mention every developer choosing to freely decide the way they should write their code without any guidelines.

HOW THE FUCK PEOPLE WRITE THEIR CODE WITHOUT THINKING ABOUT OTHER DEVS!!! SO BASICALLY I AM NOT ONLY CLEANING SOMEONE ELSE’S SHIT BUT ALSO TRY NOT TO SHIT IN THE PROCESS!! FML

So, do any of your poor fuckers have the opportunity - nay, PRIVILEGE of using the absolute clusterfuck piece of shit known as SQL Server Integration Services?

Why do I keep seeing articles about how "powerful" and "fast" it is? Why do people recommend it? Why do some think it's easy to use - or even useful?

It can't report an error to save its life. It's logging is fucked. It's not just that it swallows all exceptions and gives unhelpful error messages with no debugging information attached, its logging API is also fucked. For example, depending on where you want to log a message - it's a totally different API, with a billion parameters most of which you need to supply "-1" or "null" to just to get it do FUCKING DO SOMETHING. Also - you'll only see those messages if you run the job within the context of SQL FUCKING SERVER - good luck developing on your ACTUAL FUCKING MACHINE.

So apart from shitty logging, it has inherited Microsoft's insane need to make everything STATICALLY GODDAMN TYPED. For EVERY FUCKING COMPONENT you need to define the output fields, types and lengths - like this is 1994. Are you consuming a dynamic data structure, perhaps some EAV thing from a sales system? FUCK YOU. Oh - and you can't use any of the advances in .NET in the last 10 years - mainly, NuGet and modern C# language features.

Using a modern C# language feature REMOVES THE ABILITY TO FUCKING DEBUG ANYTHING. THE FUCKER WILL NOT STOP ON YOUR BREAKPOINTS. In addition - need a JSON parsing library? Want to import a SDK specific to what you're doing? Want to use a 3rd party date library? WELL FUCK YOU. YOU HAVE TO INDEPENDENTLY INSTALL THE ASSEMBLIES INTO THE GAC AND MAKE IT CONSISTENT ACROSS ALL YOUR ENVIRONMENTS.

While i'm at it - need to connect to anything? FUCK YOU, WE ONLY INCLUDE THE MOST BASIC DATABASE CONNECTORS. Need to transform anything? FUCK YOU, WRITE A SCRIPT TASK. Ok, i'd like to write a script task please. FUCK YOU IM GOING TO PAUSE FOR THE NEXT 10 MINUTES WHILE I FIRE UP A WHOLE FUCKING NEW INSTANCE OF VISUAL STUDIO JUST TO EDIT THE FUCKING SCRIPT. Heaven forbid you forget to click the "stop" button after running the package and open the script. Those changes you just made? HAHA FUCK YOU I DISCARDED THEM.

I honestly cant understand why anyone uses this shit. I guess I shouldn't really expect anything less from Microsoft - all of their products are average as fuck.

Why do I use this shit? I work for a bunch of fucks that are so far entrenched in Microsoft technologies that they literally cannot see outside of them (and indeed don't want to - because even a cursory look would force them to conclude that they fucked up, and if you're a manager thats something you can never do).

Ok, rant over. Also fuck you SSIS

Just some figma improvements from the perspective of a new customer:

* Copy/paste is broken. If I want to make a change, I have to create a whole new
component. They recommend cmd+c/v for copypaste but as far as I can see it does nothing

* Needs to be an explicit component drawer button instead of hiding it under assets. Through me for a loop for a couple minutes.

* Empty textboxes shouldn't vanish because you happened to click in the wrong location
while setting your properties.

* Text should start big enough to actually see.

* "send to back/front", "hide item", "change transparency' all need to be prototype actions and more, give us access to object properties both by parent/sibling/child, and by
object id

* create a new frame based on a specified size is non-obvious and if you're creating
a lot of frames, what with copypaste being non-intuitive, it can become laborious.
This is especially so when you're copying frames in order to make minor changes and observe the differences side by side, instead of potentially destructive edits.

* I see no obvious way to manage transitions/animations between frames.

* The difference between frames and groups isn't sufficiently explained. The words
frame, groups, and layers all appear to new users to be used interchangeably, even
if they are distinct things.

Lads. I need your help.

I'm building a little web app project in class where a user can create their own micro-personal page similar to about.me, except a little more modular.

users can add cards to their page with a title, maybe an image, some text, and some buttons.

my question is: how can I represent such things in my database? I can think of how a json file representation might look, but not sure how that translates to SQL?

here's how I imagine it:

userProfileComponents ={
cards=[
{
title: sometitle
image: src
text: null
button: { icon: facebook, text: facebook}
},
{
title: another title
image: null
text: some stuff about me
button: null
}
]

I wouldn't want to create a table for each micro-page - cause that would scale like absolute shit. but I feel like cramming EVERYONES components in a component table would be hell on earth. any tips? thanks

It's so frustrating to explain rxjs pitfalls to the manager.

To avoid the diamond problems and glitches caused by combineLatest and debounceTime(0), I decided to use single stream with nested reactivity.

Then the manager asked me to use withLatestFrom instead of combineLatest.

Sure withLatestFrom makes sense to the original author, when the original author is able to remember the reactive graph and put proper dependencies to combineLasted/withLatestFrom accordingly, but anyone else who touches the component later needs to retrace the reactive graph to avoid the glitch. Sometimes it's just impossible when many dependencies are derived from combineLatest+debounceTime(0). When no one is trained to code reactively, I don't expect people to know where to put a dependency. After many trials and errors, the only way to avoid the diamond problem is to use nested reactivity where child streams are created within root stream each time root stream emits.

The mentioned manager put all sorts of side effects in observable chains. The manager keeps saying the stream is too large when their subscription functions (sometimes nested) are way worse with litered mutations everywhere. Anything in observable can be traced by go to definition but tracing side effects usually requires global searches.

Recently, he put startWith to the end of a synced stream to fix a bug where button would collapse when there is no content (initial null emission). Rather than fixing the default height, he thinks using startWith(defaultLabel) is a good idea. Of course, he doesn't know that a synced stream should only emit 1 value on new subscription and that extra emitted value will cause rxjs glich down the pipe.

I hate corporate jobs