When my mom asks about my problems :p

aslkfjasf. i've spent 12 hours today (and lots more over the past two days) trying to reproduce a bug that my [sort of] coworker insists is present. I haven't seen any proof of it anywhere, let alone steps to reproduce it.

I've poured through the code, following all of its tangled noodles of madness from start to fuck-this-shit. I've read and reread the pile of demon excrement so many times i can still read the code when i close my eyes. so. not. kidding.

anyway, the coworker person is getting mad because i haven't fixed the bug after days, and haven't even reproduced it yet. This feature is already taking way too fucking long so I totally don't blame him. but urghh it's like trying to unwind a string someone tied into a tight little ball of knots because they were bored.

but i just figured out why I haven't been able to reproduce it.

the stupid fucking unreliable dipshit ex-"i'm a rockstar and my code rocks"-CTO buffoon (aka API Guy, aka the `a=b if a!=b`loody pointless waste of mixed spaces and tabs) that wrote the original APIs ... 'kay, i need to stop for breath.

The dumbfuck wrote the APIs (which I based the new ones on mostly wholesale because wtf messy?), but he never implemented a very fucking important feature for a specific merchant type. It works for literally every type except the (soon-to-be) most common one. and it just so happens that i need that very specific feature to reproduce this bug.

Why is that one specific merchant type handled so differently? No fucking idea.

But exactly how they're handled differently is why I'm so fking pissed off. It's his error checking. (Some) of his functions return different object types (hash, database object, string, nullable bool, ...) depending on what happened. like, when creating a new gift, it (eventually...) either returns a new Gift object or a string error basically saying "ahhh everything's broken again!" -- which is never displayed, compared against, or recorded anywhere, ofc. Here, the API expects a Hash. That particular function call *always* returns a Hash, no matter what happens in the myriad, twisting, and interwoven branches the code could take. So the check is completely pointless.

EXCEPT. if an object associated with another object associated with the passed object (yep) has a type of 8. in which case, one of the methods in the chain returns a PrintQueue that gets passed back up the call stack. implicitly, and nested three levels in. ofc.

And if the API doesn't get its precious Hash, it exclaims that the merchant itself is broken, and tells the user to contact support. despite, you know, the PrintQueue showing that everything worked perfectly. In fact, that merchant's printer will be happily printing away in the background.

All because type checking is this guy's preferred method of detecting errors. (Raise? what's that? OOP? Nah, let's do diverging splintered-monolithic with some Ruby objects thrown in.)

just.
what the crap.

people should keep their mental diarrhea away from their keyboards.

Anyway. the summary of this long-winded, exhaustion-fueled tirade is that our second-most-loved feature doesn't work on our second-most-common merchant type.

and ofc that was the type of merchant i've been testing on. for days. while having both a [semi] coworker and my boss growing increasingly angry at me for my lack of progress.

It's also a huge feature, and the boss doesn't understand that. (can't or won't, idk)

So.
yep.
that's been my week.

...... WHAT A FUCKING BUFFOON!

I wrote a database migration to add a column to a table and populated that column upon record creation.
But the code is so freaking convoluted that it took me four days of clawing my eyes out to manage this.
BUT IT'S FINALLY DONE.
FREAKING YAY.

Why so long, you ask? Just how convoluted could this possibly be? Follow my lead ~

There's an API to create a gift. (Possibly more; I have no bloody clue.)
I needed the mobile dev contractor to tell me which APIs he uses because there are lots of unused ones, and no reasoning to their naming, nor comments telling me what they do.

This API takes the supplied gift params, cherry-picks a few bits of useful data out (by passing both hashes by reference to several methods), replaces a couple of them with lookups / class instances (more pass-by-reference nonsense). After all of this, it logs the resulting (and very different) mess, and happily declares it the original supplied params. Utterly useless for basically everything, and so very wrong.

It then uses this data to call GiftSale#create, which returns an instance of GiftSale (that's actually a Gift; more on that soon).

GiftSale inherits from Gift, and redefines three of its methods.

GiftSale#create performs a lot of validations / data massaging, some by reference, some not. It uses `super` to call Gift#create which actually maps to the constructor Gift#initialize.

Gift#initialize calls Gift#pre_init (passing the data by reference again), which does nothing and returns null. But remember: GiftSale inherits from Gift, meaning GiftSale#pre_init supersedes Gift#pre_init, so that one is called instead. GiftSale#pre_init returns a Stripe charge object upon success, or a Gift (and a log entry containing '500 Internal') upon failure. But this is irrelevant because the return value is never actually used. Pass by reference, remember? I didn't.

We're now back at Gift#initialize, Rails finally creates a Gift object using the args modified [mostly] in-place by all of the above.

Another step back and we're at GiftSale#create again. This method returns either the shiny new Gift object or an error string (???), and the API logic branches on its type. For further confusion: not all of the method's returns are explicit, and those implicit return values are nested three levels deep. (In Ruby, a method will return the last executed line's return value automatically, allowing e.g. `def add(a,b); a+b; end`)

So, to summarize: GiftSale#create jumps back and forth between Gift five times before finally creating a Gift instance, and each jump further modifies the supplied params in-place.

Also. There are no rescue/catch blocks, meaning any issue with any of the above results in a 500. (A real 500, not a fake 500 like last time. A real 500, with tragic consequences.)

If you're having trouble following the above... yep! That's why it took FOUR FREAKING DAYS! I had no tests, no documentation, no already-built way of testing the API, and no idea what data to send it. especially considering it requires data from Stripe. It also requires an active session token + user data, and I likewise had no login API tests, documentation, logging, no idea how to create a user ... fucking hell, it's a mess.)

Also, and quite confusingly:
There's a class for GiftSale, but there's no table for it.
Gift and GiftSale are completely interchangeable except for their #create methods.

So, why does GiftSale exist?
I have no bloody idea.
All it seems to do is make everything far more complicated than it needs to be.

Anyway. My total commit?
Six lines.
IN FOUR FUCKING DAYS!

AHSKJGHALSKHGLKAHDSGJKASGH.

tl;dr @Root refactors some spaghetti.

I'm refactoring an api that creates a support message. It's a post route.

When seeing a magic hardcoded message string, this route instead updates the user object, and does not create a support message.

It also returns different results if the user is muted (fine) or if saving the message succeeds or fails (fine).

But if the user is creating a duplicate message, it doesn't save the message (fine) and... redirects to listing their messages instead? Wat?

Also, when refactoring this (migrating to a new message backend), I discovered that not all routes return a response. If the message is a non-duplicate, from a non-muted player, from a non-redacted client, the route doesn't respond at all!

So, I'm having fun cleaning this up. I actually am. Except I'll need to support all of the legacy clients for the next lifetime or two. I mean, really. There are still people with Android v2 devices who are using this thing. not even kidding.

I'm freelancing and there was a guy who needed help with JavaFX. He gave me code only to his view class so I don't steal the rest... It was around 5k lines and it was full of compilation errors because of missing classes. While checking the code I realised something is really wrong with his model classes. So I asked maybe he could send me this one model class that was suspicious. So he did and it was around 10k lines long and had around 200 fields... ALL OF THEM FUCKING STRINGS except 3. You know what the rest of 3 were? 2 Lists of strings and a boolean... It was his "main" model class, he was using it for everything. It had setters for all fields and empty default constructor, so he would just instantiate the object and would set the fields that he wanted to use. Need new functionality? Just add 5more String fields and set them!

Besides the fact that there would be an error handler, wouldn't it store a phone number as an object other than a computable number, like a string, cuz phone number is like a handler, a reference, and not something you'd ever perform arithmetic on?

I swear I work with mentally deranged lunatics.
Dev is/was using TFS's web api to read some config stuff..

Ralph: "Ugh..this is driving me crazy. I've spent all day trying to read this string from TFS and it is not working"
Me: "Um, reading a string from an web api is pretty easy, what's the problem?"
Ralph: "I'm executing the call in a 'using' statement and cannot return the stream."
Me: "Why do you need to return a stream? Return the object you are looking for."
Ralph: "Its not that easy. You can return anything from TFS. All you get back is a stream. Could be XML, JSON, text file, image, anything."
Me: "What are you trying to return?"
Ralph: "XML config. If I use XDoc, the stream works fine, but when I step into each byte from the stream, I the first three bytes have weird characters. I shouldn't have to skip the first three bytes to get the data. I spent maybe 5 hours yesterday digging around the .Net stream readers used in XDoc trying to figure out how it skips the first few bytes."
Me: "Wow...I would have used XDoc and been done and not worried about that other junk."
Ralph: "But I don't know the stream is XML. That's what I need to figure out."
Me: "What is there to figure out? You do know. Its your request. You are requesting a XML config."
Ralph: "No, the request can be anything. What if Sam requests an image? XDoc isn't going to work."
Me: "Is that a use-case? Sam requesting an image?"
Ralph: "Uh..I don't know...he could"
Me: "Sounds like your spending a lot of time doing premature optimization. You know what your accessing TFS for, if it's XML, return XML. If it's an image, return an image. Something new comes along, modify the code to handle it. Eazy peezy."
<boss walks in from a meeting>
Boss: "Whats up guys?"
Ralph: "You know the problem with TFS and not being able to stream the data I had all day yesterday? I finally figured it out. I need to keep this TFS reader simple. I'll start with the XML configs and if we more readers later, we can add them."
Boss: "Oh yea, always start simple and add complexity only when you need it."

Frack...Frack..Frack...you played some victim complaining to anyone who would listen yesterday (which I mostly ignored) about reading data from TFS was this monumental problem no one could solve, then you start complaining to me, I don't fall for the BS, then tell the boss the solution was your idea?

Lunatic or genius? Wally would be proud.

A dev team has been spending the past couple of weeks working on a 'generic rule engine' to validate a marketing process. The “Buy 5, get 10% off” kind of promotions.

The UI has all the great bits, drop-downs, various data lookups, etc etc..

What the dev is storing the database is the actual string representation FieldA=“Buy 5, get 10% off” that is “built” from the UI.

Might be OK, but now they want to apply that string to an actual order. Extract ‘5’, the word ‘Buy’ to apply to the purchase quantity rule, ‘10%’ and the word ‘off’ to subtract from the total.

Dev asked me:

Dev: “How can I use reflection to parse the string and determine what are integers, decimals, and percents?”
Me: “That sounds complicated. Why would you do that?”
Dev: “It’s only a string. Parsing it was easy. First we need to know how to extract numbers and be able to compare them.”
Me: “I’ve seen the data structures, wouldn’t it be easier to serialize the objects to JSON and store the string in the database? When you deserialize, you won’t have to parse or do any kind of reflection. You should try to keep the rule behavior as simple as possible. Developing your own tokenizer that relies on reflection and hoping the UI doesn’t change isn’t going to be reliable.”
Dev: “Tokens!...yea…tokens…that’s what we want. I’ll come up with a tokenizing algorithm that can utilize recursion and reflection to extract all the comparable data structures.”
Me: “Wow…uh…no, don’t do that. The UI already has to map the data, just make it easy on yourself and serialize that object. It’s like one line of code to serialize and deserialize.”
Dev: “I don’t know…sounds like magic. Using tokens seems like the more straightforward O-O approach. Thanks anyway.”

I probably getting too old to keep up with these kids, I have no idea what the frack he was talking about. Not sure if they are too smart or I’m too stupid/lazy. Either way, I keeping my name as far away from that project as possible.

I am much too tired to go into details, probably because I left the office at 11:15pm, but I finally finished a feature. It doesn't even sound like a particularly large or complicated feature. It sounds like a simple, 1-2 day feature until you look at it closely.

It took me an entire fucking week. and all the while I was coaching a junior dev who had just picked up Rails and was building something very similar.

It's the model, controller, and UI for creating a parent object along with 0-n child objects, with default children suggestions, a fancy ui including the ability to dynamically add/remove children via buttons. and have the entire happy family save nicely and atomically on the backend. Plus a detailed-but-simple listing for non-technicals including some absolutely nontrivial css acrobatics.

After getting about 90% of everything built and working and beautiful, I learned that Rails does quite a bit of this for you, through `accepts_nested_params_for :collection`. But that requires very specific form input namespacing, and building that out correctly is flipping difficult. It's not like I could find good examples anywhere, either. I looked for hours. I finally found a rails tutorial vide linked from a comment on a SO answer from five years ago, and mashed its oversimplified and dated examples with the newer documentation, and worked around the issues that of course arose from that disasterous paring.

like.
I needed to store a template of the child object markup somewhere, yeah? The video had me trying to store all of the markup in a `data-fields=" "` attrib. wth? I tried storing it as a string and injecting it into javascript, but that didn't work either. parsing errors! yay! good job, you two.

So I ended up storing the markup (rendered from a rails partial) in an html comment of all things, and pulling the markup out of the comment and gsubbing its IDs on document load. This has the annoying effect of preventing me from using html comments in that partial (not that i really use them anyway, but.)

Just.
Every step of the way on building this was another mountain climb.
* singular vs plural naming and routing, and named routes. and dealing with issues arising from existing incorrect pluralization.
* reverse polymorphic relation (child -> x parent)
* The testing suite is incompatible with the new rails6. There is no fix. None. I checked. Nope. Not happening.
* Rails6 randomly and constantly crashes and/or caches random things (including arbitrary code changes) in development mode (and only development mode) when working with multiple databases.
* nested form builders
* styling a fucking checkbox
* Making that checkbox (rather, its label and container div) into a sexy animated slider
* passing data and locals to and between partials
* misleading documentation
* building the partials to be self-contained and reusable
* coercing form builders into namespacing nested html inputs the way Rails expects
* input namespacing redux, now with nested form builders too!
* Figuring out how to generate markup for an empty child when I'm no longer rendering the children myself
* Figuring out where the fuck to put the blank child template markup so it's accessible, has the right namespacing, and is not submitted with everything else
* Figuring out how the fuck to read an html comment with JS
* nested strong params
* nested strong params
* nested fucking strong params
* caching parsed children's data on parent when the whole thing is bloody atomic.
* Converting datetimes from/to milliseconds on save/load
* CSS and bootstrap collisions
* CSS and bootstrap stupidity
* Reinventing the entire multi-child / nested params / atomic creating/updating/deleting feature on my own before discovering Rails can do that for you.

Just.
I am so glad it's working.
I don't even feel relieved. I just feel exhausted.

But it's done.
finally.

and it's done well. It's all self-contained and reusable, it's easy to read, has separate styling and reusable partials, etc. It's a two line copy/paste drop-in for any other model that needs it. Two lines and it just works, and even tells you if you screwed up.

I'm incredibly proud of everything that went into this.
But mostly I'm just incredibly tired.

Time for some well-deserved sleep.

While working on my one of the first project in java i ended up using deprecated Calendar API for the date. Since deadline was near i thought it would be a good idea to use the JCalendar API for as date picker (which is a third party API).BAD IDEA. It was the night before the submission round about 11pm when i realized that there is no way to convert JCal object into Calendar and it turned out it is not working as expected you have to subtract a particular number from the year to get date right.

To convert JCal into Calendar i used the toString function to get the date in string sliced it using substring into year,month,day then had to assign date to Calendar object via constructor.
Had to write 70 lines of code just to convert JCal into Calendar...

And then there were other complications related to this problem. Had tu pull an all nighter just to solve date related problems
LESSONS LEARNED :
NEVER USE A DEPRECATED API
NEVER USE THIRD PARTY APIs WITHOUT RESEARCH

I inherited some Java code from a guy that thought you could only return strings from a method.

So how did he ever pass anything other than a string, I hear you ask? He had a (hand crafted, using no libraries, and not escaping anything properly) toXml() and fromXml() method on each object.

Yeah...

Imagine saving Integers and Floats in a MySQL table as strings containing locale based thousand sepatators...
man... fickt das hart!

Wait, there's more!
Imagine storing a field containing list of object data as a CSV in a single table column instead of using JSON format or a separate DB table.... and later parsing it by splitting the CSV string on ";"...

LPT: NEVER accept a freelance job without looking at the project's source first

Client: I have a project made by a company that is now abandoning it, I want you to fix some bugs

Me: Okay, can you:
1) Give me a build to test the current state of the game
2) Tell me what the bugs are
3) Show me the source
4) Tell me your budget

Client: *sends a list of 10 bugs* Here's the APK and to give you the project I'll need you to sign an NDA

Me: Sure...

*tests build*
*sees at least 20 bugs*
*still downloading source*
*bugs look quite easy to fix should be done under an hour*

Me: Okay, so, I can fix each bug for $10 and I can do 2 today

Client: Okay can you fix 8 bugs today for $40??

*sigh*

Me: No I cannot.

Client: okay then 2 today for $20 is fine, I want a refund if you can't fix them today

*sigh*

Me: Look dude, this isn't the first time I am doing this, aight? I'll fix the bugs today you can pay me after check they are done, savvy?

Client: okay

*source is downloaded*
*literal apes wrote the scripts, commented out code EVERYWHERE
Debug logs after every line printing every frame causing FPS drops, empty objects in the scene
multiple unused UI objects
everything is spaghetti*

*give up, after 2 hours of hell*
*tfw averted an order cancellation by not taking the order and telling client that they can pay me after I am done*

Attached is an image of a level object pool
It's an array with each element representing a level.
The numbers and "Final" are ids for objects in an object pool
The whole string is .Split(',') into an array (RIP MEMORY BTW) and then a loop goes through each element in the split array and instantiates the object from an object pool

PHP arrays.

The built-in array is also an hashmap. Actually, it's always a hashmap, but you can append to it without specifying indexes and PHP will use consecutive integers. Its performance characteristics? Who knows. Oh, and only strings, ints and null are valid keys.

What's the iteration order for arrays if you use them as hashmaps (string keys)? Well, they have their internal order. So it's actually an ordered hashmap that's being called an array. And you can produce an array which has only integer keys starting with 0, but with non-sequential internal (iteration) order.

This array weirdness has some non-trivial implications. `json_encode` (serializes argument to JSON) assumes an array corresponds to a JSON array if its keys are consecutive integers in increasing order starting with 0, otherwise the array becomes a JSON object. `array_filter` (filters arrays/hashmaps using callback predicate) preserves keys, so it will punch holes in the int key sequence if non-last items are removed, thus turning arrays into hashmaps and changing your JSON structure if you forget to discard keys before serialization.

You may wonder how JSON deserialization works, then? There's a special class for deserialized JSON objects, `stdClass`. It's basically a hashmap too, but it's an object, not an array, and all functions that would normally accept arrays won't work with it. So basically its only use is JSON (de)serialization. You can even cast arrays to objects, producing `stdClass`.

Bonus PHP trivia:

Many functions return nonsensical values. `preg_match`, the regex matching function, returns 1 for success, 0 for no matches and false for malformed regular expression. PHP supports exceptions, so it could just throw one on errors. It would even make more sense to return true, false and null for these three cases. But no, 1, 0 and false. And actual matches are returned by output arg.

`array_walk_recursive`, a function supposed to recursively apply callback to each element of an array. That's what docs say. It actually applies it to leafs only. It will also silently accept object instead of array and "walk" it, but without recursing into deeper objects.

Runtime type enforcing is supported for function arguments and returned values. You can use scalar types, classes, array, null and a few special keywords. There's also a `mixed` keyword, which is used in docs and means "anything". It's syntactically valid, the parser will accept it, but it matches no values in runtime. Calling such function will always cause a runtime error.

Strings can be indexed with negative integers. Arrays can't.

ReflectionClass::newInstanceWithoutConstructor: "Creates a new class instance without invoking the constructor". This one needs no commentary.

`array_map` is pretty self-explanatory if you call it with a callback and an array. Or if you provide more arrays of equal length via varargs, callback will be called with more arguments, one from each array. Makes sense so far. Now, you can also call `array_map` with null instead of callback. In that case it treats provided arrays as rows of a matrix and returns that matrix, transposed.

Ocaml, the strictest language when it comes to types i have ever seen, parses a string to a general json object in around 5 loc.

Take notes java.

So, this is probably somewhat esoteric but...

While studying at university I had a "programming paradigms" module, dunno why they called it that, it was more like "introduction to functional programming".

So, it's kinda mind bending, we'd only really started to get our heads around classical object oriented programming and they throw functional programming at us.

It's worse than that though, for do they use an established language, like lisp/scheme, functional Python, or even given Haskell?

No, of course they didn't. They taught us Oz.

You probably won't have heard of it, but this language is burned into the back of my brain, along with a vague understanding of the n-queens problem we had to solve graphically (using qTk, which I dunno if someone took qt and tk and blended them, I stopped asking questions after a while).

To top it off did this language (at the time) have a stand alone interpreter? Did it buggery! It was coupled to the Mozart programming system, which is just Emacs (which has a bloody lisp built into it,so close, yet so far 😭).

It gets worse, though, oh does it get worse, for pause dear reader and consider, have you ever heard of Mozart/oz before, I'd put money on most of you had not heard of it until today.

For, you see, I believe at the time of writing, one, yes, ONE text book exists on this language. When I was doing my assignment there was merely some published conference notes and language design documents.

That's not all, I was not the only one experiencing difficulties with this language, someone in the class ended up pouring through the mailing lists and found the very tutor teaching the class struggling at first to understand the language.

I had to repeat that year. The functional programming class was one semester.

When I retook that year, it was a whole year long. However, halfway through the year, original tutor was fired and a new tutor was hired to teach the language.

He was, understandably, just as confused as we were.

There was a Starbucks and a pub equidistant from the lecture hall, though in opposite directions. From lecture to lecture we had no idea which one we'd end up in.

I have reason to believe Mozart/Oz it some sort of otherworldly abomination designed to give students the occasional nightmare flashback, long after they've left.

My room had post it notes, sheets of paper, print outs, diagrams, doodles and pens, just stuck to the wall, I looked like a raving lunatic three hours away from being institutionalised. There was string connecting one diagram to the next and images of a chess queen all over. As I attempted to solve the n-queens problem.

Madmans knowledge, I call it. I can never unlearn all that, in fact it seeps into much of the code I write. Such information was not meant for the minds of a simple country bumpkin such as myself...

Mozart/Oz... I wouldn't be the programmer I am today without it, and that's frankly terrifying...

I am in love with regex (re in py)!
It is so damn cool!
Unfortunately, Python's find is faster.
But regex is so much better when it comes to complex patterns you would not be able to find with the regular (do you see the pun here? ;)) python's string object find method.

Just got a new job at an old school hardware company. The codebase is giving me heart attack. They don't care about dev experience or code navigation at all. Every attempts to modernize the codebase is so half assed. All patches are so bloated that make the codebase even worse.

Frontend is migrated from prototype-oop-jquery cluster fuck to AngularJS, then finally angular. Holy moly, all business logics are baked into UI "classes" using prototype chain. When they migrated to AngularJS, someone simply added a wrapper to that jQuery cluster fuck class and overwrote all the prototype with a 10k +lines file. Since all the methods are hidden in either prototype, JS object, or callback function, it's impossible to trace the data pipeline using IDE when "go to definition" on update() method gives you all the update methods/string in all objects/classes. And they don't care about immutability. References are taken out, renamed, and mutated everywhere. Finding the source of a bug is fucking guessing game.

I don't know what trick they use that makes cLion static analyzer fail.

And there is no unit test or spec doc.

Fuck me dead

You know what really grinds my gears? When people criticize a programming language but uses edge cases and stuff that can be avoided by using the tried and true "don't be an idiot". Take for instance JavaScript, a language I like and a language that has a lot you can criticize. But I feel like a lot of peoples criticism isn't warranted.

What's that? No ints? Use parseInt or Math.floor.

What are you saying? == works in strange ways? Yes, that's what we have === for.

Type coercion is wonky? Think it's weird how string + int works differently than string - int? Wanna string with number + - + - - + - - etc? Don't! Don't add strings and ints, don't subtract strings and ints. You can't in statically typed languages and you aren't supposed to in dynamically typed

Adding arrays and objects, arrays and arrays, objects and objects etc. is inconsistent? Why are you trying to do that?

Adding floats together gives odd results? Now we're getting somewhere! And Mozilla responded to that with a method called toFixed.

Declaring variables with var doesn't always work that well? Use let and const

Then there's this weird attitude that some people I've met have, where they will complain about the module system and how "well you rely on the community for those packages" as if it's a bad thing. And then coming with the "well you don't know what the (open source) packages do internally" as if I (for the most part) give a shit. Then they'll swear by companies like Zend or Microsoft as if they can't just stop supporting the languages they use. Maybe it's just because I like community content more because of video game mods.

Wanna criticize JS, then there's plenty to talk about. Like the built in date object is basically shit. Or how in NodeJS you can have node_modules in your node_modules. Or how classes don't really have the best syntax. Left-Pad. And so on (it's too late for me to be able to remember much more).

Conversation yesterday (senior dev and the mgr)..

SeniorDev: "Yea, I told Ken when using the service, pass the JSON string and serialize to their object. JSON eliminates the data contract mismatch errors they keep running into."
Mgr: "That sounds really familiar. Didn't we do this before?"
SeniorDev: "Hmmm...no. I doubt anyone has done this before."
Me: "Yea, our business tier processor handled transactions via XML. It allowed the client and server to process business objects regardless of platform. Partners using Perl,
clients using Delphi, website using .aspx, and our SQLServer broker even used it."
Mgr: "Oh yea...why did we stop using it?"
Me: "WCF. Remember, the new dev manager at the time and his team broke up the business processor into individual WCF services."
Mgr: "Boy, that was a crap fest. We're still fighting bugs from the mobile devices. Can't wait until we migrate everything to REST."
SeniorDev: "Yea, that was such a -bleep-ing joke."
Me: "You were on Jake's team at the time. You were the primary developer in the re-write process saying passing strings around wasn't the way true object-oriented developers write code.
So it's OK now because the string is in JSON format or because using a JSON string your idea?"

SeniorDev turns around in his desk and puts his headphones back on.

That's right you lying SOB...I remember exactly the level of personal attacks you spewed on me and other developers behind our backs for using XML as the message format.

Keep your fat ass in your seat and shut the hell up.

The best kinds of comments:

/**
Gets user CC info datas.
*/
public Object getUserCCInfoDatas() {}

If you really want to outdo yourself:

/**
Gets user CC info datas.
@param someshit Outdated docs ftw
*/
public Object getUserCCInfoDatas(String unrelatedToDocAbove) {}

Honestly, no documentation is better in some cases. At least I can't be angry about their shitty quality... And they don't waste my time.

CoWorker: “Yea then just use double inheritance to grab the methods off the two classes.”

Me: “Yea that doesn’t seem right... the first object is a string parser the second object predicts future occurrences...”

CoWorker: [louder] “No trust me, I’m taking a developer course right now. If you inherit both classes your new class can use all the needed methods!”

Me: “Okay, go for it bro.”

So tired of people who think they know what they’re doing...

So for those of you keeping track, I've become a bit of a data munger of late, something that is both interesting and somewhat frustrating.

I work with a variety of enterprise data sources. Those of you who have done enterprise work will know what I mean. Forget lovely Web APIs with proper authentication and JSON fed by well-known open source libraries. No, I've got the output from an AS/400 to deal with (For the youngsters amongst you, AS/400 is a 1980s IBM mainframe-ish operating system that oriiganlly ran on 48-bit computers). I've got EDIFACT to deal with (for the youngsters amongst you: EDIFACT is the 1980s precursor to XML. It's all cryptic codes, + delimited fields and ' delimited lines) and I've got legacy databases to massage into newer formats, all for what is laughably called my "data warehouse".

But of course, the one system that actually gives me serious problems is the most modern one. It's web-based, on internal servers. It's got all the late-naughties buzzowrds in web development, such as AJAX and JQuery. And it now has a "Web Service" interface at the request of the bosses, that I have to use.

The programmers of this system have based it on that very well-known database: Intersystems Caché. This is an Object Database, and doesn't have an SQL driver by default, so I'm basically required to use this "Web Service".

Let's put aside the poor security. I basically pass a hard-coded human readable string as password in a password field in the GET parameters. This is a step up from no security, to be fair, though not much.

It's the fact that the thing lies. All the files it spits out start with that fateful string: '<?xml version="1.0" encoding="ISO-8859-1"?>' and it lies.

It's all UTF-8, which has made some of my parsers choke, when they're expecting latin-1.

But no, the real lie is the fact that IT IS NOT WELL-FORMED XML. Let alone Valid.

THERE IS NO ROOT ELEMENT!

So now, I have to waste my time writing a proxy for this "web service" that rewrites the XML encoding string on these files, and adds a root element, just so I can spit it at an XML parser. This means added infrastructure for my data munging, and more potential bugs introduced or points of failure.

Let's just say that the developers of this system don't really cope with people wanting to integrate with them. It's amazing that they manage to integrate with third parties at all...

python machine learning tutorials:
- import preprocessed dataset in perfect format specially crafted to match the model instead of reading from file like an actual real life would work
- use images data for recurrent neural network and see no problem
- use Conv1D for 2d input data like images
- use two letter variable names that only tutorial creator knows what they mean.
- do 10 data transformation in 1 line with no explanation of what is going on
- just enter these magic words
- okey guys thanks for watching make sure to hit that subscribe button

ehh, the machine learning ecosystem is burning pile of shit let me give you some examples:

- thanks to years of object oriented programming research and most wonderful abstractions we have "loss.backward()" which have no apparent connection to model but it affects the model, good to know
- cannot install the python packages because python must be >= 3.9 and at the same time < 3.9
- runtime error with bullshit cryptic message
- python having no data types but pytorch forces you to specify float32
- lets throw away the module name of a function with these simple tricks:
"import torch.nn.functional as F"
"import torch_geometric.transforms as T"
- tensor.detach().cpu().numpy() ???
- class NeuralNetwork(torch.nn.Module):
def __init__(self):
super(NeuralNetwork, self).__init__() ????
- lets call a function that switches on the tracking of math operations on tensors "model.train()" instead of something more indicative of the function actual effect like "model.set_mode_to_train()"
- what the fuck is ".iloc" ?
- solving environment -/- brings back memories when you could make a breakfast while the computer was turning on
- hey lets choose the slowest, most sloppy and inconsistent language ever created for high performance computing task called "data sCieNcE". but.. but. you can use numpy! I DONT GIVE A SHIT about numpy why don't you motherfuckers create a language that is inherently performant instead of calling some convoluted c++ library that requires 10s of dependencies? Why don't you create a package management system that works without me having to try random bullshit for 3 hours???
- lets set as industry standard a jupyter notebook which is not git compatible and have either 2 second latency of tab completion, no tab completion, no documentation on hover or useless documentation on hover, no way to easily redo the changes, no autosave, no error highlighting and possibility to use variable defined in a cell below in the cell above it
- lets use inconsistent variable names like "read_csv" and "isfile"
- lets pass a boolean variable as a string "true"
- lets contribute to tech enabled authoritarianism and create a face recognition and object detection models that china uses to destroy uyghur minority
- lets create a license plate computer vision system that will help government surveillance everyone, guys what a great idea

I don't want to deal with this bullshit language, bullshit ecosystem and bullshit unethical tech anymore.

Spent 2 hours trying get ng-model working with <select> with no success. Then co worker came and changed model from string to object and moon walked back to his seat.

I like my log messages to indicate automatically where in the code something happened, so that I can easily identify where a message originated from while tracking down problems.

In C/C++ this is nice and easy - write a logging routine, wrap it in macros for the different log levels and have that automatically output __FILE__, __LINE__ etc.

I wanted to do something similar in NodeJS, as I'd found myself manually writing the file name in the log message and then splitting functionality out into new files and it became a mess.

The only way I found to be able to do this was to create an "Error" object and access the "stack" member of it. This is a string containing a stack backtrace, suitable for writing to console/file. I just wanted the filename/line/routine.

So I ended up splitting the string into lines, then for each of the lines, trimming the surrounding spaces (or tabs?), and parsing them to see if the stack entry is inside my logger module. The first entry outside of that module must therefore be the thing that called it, so I then parse out the routine or object and method, filename and line number.

It's a lot of clumsy work but the output is pretty neat. I just wish it were simpler!

Using an api: ok, this url (.../xml/endpoint) gives me an xml-document. Oh, and there is a node whose text contains html markup, interesting.
Using the same endpoint, but requesting json: yep, that's the same data, there even is this big html string and... why is this string in a json object wrapped inside "<![CDATA[...]]>"?

If i ever see a courtroom from the inside i'll plead insanity.

I need some opinions on Rx and MVVM. Its being done in iOS, but I think its fairly general programming question.

The small team I joined is using Rx (I've never used it before) and I'm trying to learn and catch up to them. Looking at the code, I think there are thousands of lines of over-engineered code that could be done so much simpler. From a non Rx point of view, I think we are following some bad practises, from an Rx point of view the guys are saying this is what Rx needs to be. I'm trying to discuss this with them, but they are shooting me down saying I just don't know enough about Rx. Maybe thats true, maybe I just don't get it, but they aren't exactly explaining it, just telling me i'm wrong and they are right. I need another set of eyes on this to see if it is just me.

One of the main points is that there are many places where network errors shouldn't complete the observable (i.e. can't call onError), I understand this concept. I read a response from the RxSwift maintainers that said the way to handle this was to wrap your response type in a class with a generic type (e.g. Result<T>) that contained a property to denote a success or error and maybe an error message. This way errors (such as incorrect password) won't cause it to complete, everything goes through onNext and users can retry / go again, makes sense.

The guys are saying that this breaks Rx principals and MVVM. Instead we need separate observables for every type of response. So we have viewModels that contain:

- isSuccessObservable
- isErrorObservable
- isLoadingObservable
- isRefreshingObservable
- etc. (some have close to 10 different observables)

To me this is overkill to have so many streams all frequently only ever delivering 1 or none messages. I would have aimed for 1 observable, that returns an object holding properties for each of these things, and sending several messages. Is that not what streams are suppose to do? Then the local code can use filters as part of the subscriptions. The major benefit of having 1 is that it becomes easier to make it generic and abstract away, which brings us to point 2.

Currently, due to each viewModel having different numbers of observables and methods of different names (but effectively doing the same thing) the guys create a new custom protocol (equivalent of a java interface) for each viewModel with its N observables. The viewModel creates local variables of PublishSubject, BehavorSubject, Driver etc. Then it implements the procotol / interface and casts all the local's back as observables. e.g.

protocol CarViewModelType {
isSuccessObservable: Observable<Car>
isErrorObservable: Observable<String>
isLoadingObservable: Observable<Void>
}

class CarViewModel {
isSuccessSubject: PublishSubject<Car>
isErrorSubject: PublishSubject<String>
isLoadingSubject: PublishSubject<Void>

// other stuff
}

extension CarViewModel: CarViewModelType {
isSuccessObservable {
return isSuccessSubject.asObservable()
}
isErrorObservable {
return isSuccessSubject.asObservable()
}
isLoadingObservable {
return isSuccessSubject.asObservable()
}
}

This has to be created by hand, for every viewModel, of which there is one for every screen and there is 40+ screens. This same structure is copy / pasted into every viewModel. As mentioned above I would like to make this all generic. Have a generic protocol for all viewModels to define 1 Observable, 1 local variable of generic type and handle the cast back automatically. The method to trigger all the business logic could also have its name standardised ("load", "fetch", "processData" etc.). Maybe we could also figure out a few other bits too. This would remove a lot of code, as well as making the code more readable (less messy), and make unit testing much easier. While it could never do everything automatically we could test the basic responses of each viewModel and have at least some testing done by default and not have everything be very boilerplate-y and copy / paste nature.

The guys think that subscribing to isSuccess and / or isError is perfect Rx + MVVM. But for some reason subscribing to status.filter(success) or status.filter(!success) is a sin of unimaginable proportions. Also the idea of multiple buttons and events all "reacting" to the same method named e.g. "load", is bad Rx (why if they all need to do the same thing?)

My thoughts on this are:

- To me its indentical in meaning and architecture, one way is just significantly less code.
- Lets say I agree its not textbook, is it not worth bending the rules to reduce code.
- We are already breaking the rules of MVVM to introduce coordinators (which I hate, as they are adding even more unnecessary code), so why is breaking it to reduce code such a no no.

Any thoughts on the above? Am I way off the mark or is this classic Rx?

The joy when tools do not have machine parseable output.

I'm looking at you SBT. My favorite pile of poo.

Remove the logging level from each line, then trim the line, then stab around inside the line with regexes, fishing for a possible match which hopefully is right...

Then stripping scala information like the object type, cause yeah...
A line can be for example "[info] Vector(File(...),File(...))" where info is the log level, Vector the wrapping sequence type, File(...) the wrapping element type and the string inside File(...) what yours truly needs.

As this is lot of shitty shabby string stabby stabby, we need to add a fuckton of boiler plate validation cause who knows what we just murdered.

To make it even more fucked up, a multi project project can produce different output for the same key.

:-)

Yeah. So we need to fix that too.

By the way, one can set log output to unbuffered in SBT.

Then the output is in random order :-)

Isn't that fun? Come on, you wanna poke that pile of shit, too.

The SBT plugin way is by the way no alternative, as I need a full Java environment for execution.

Which brings me to the last point:

For fucks sake, writing CLI applications in Java is so much bloody boilerplate code.

There's ugly and then there's the "please kill me" kind of level.

50 lines just to write a basic validation of argc / argv with commons cli.

That's 6 lines in python. Not kidding. :(

I currently hate everything.

Moments where the job sucks: When you have to hotwire two electric cables with high currency by giving both cables the blowjob of your life.

API team: "Hey they have a client demo on Monday! What about changing all of our object IDs from int to string on a Friday night without notifying our only subscribers?"

Other peoples' code... (in C++)
I am finding what some people consider good code is not as described. I found a class that provides strings. Great it gives me paths and stuff. I incorporated it in a new project.

segfaults

Hmmm, it must have an init function... It does, but not in the class. It has a friended init function:
friend init_function(). If this function is not created and called external to the class then the class will segfault...

okay...

I implement this. I use code from another project that implements this correctly. The friend class allows the private constructor to be called to create the main instance of the class. So its a fucking cryptic ass singleton. I look at this class. It uses a macro to decide what to function call in the class. The class already has function names for each call it needs to make. The class is literally a string lookup table. I vow to redo this shitty code, someday...

I start to wonder what other fragile code I will find. Not long later I keep getting errors on malloc. Like any malloc that is called results in a segfault. The malloc is not at fault though. I run valgrind and find a websocket library is returning an object a different size than the header file describes.

WTF...

Somebody has left an old ass highly modified definition of the websocket header in a location in that I include headers (partly my fault). I eliminate that from my include path. All is well, everything behaves. I will be making sure this fucking header is not used and it is going to die. Wasted a bunch of time.

Lessons learned: some code is just fucked and don't leave old ass shit you tried laying around.

Sharing a first look at a prototype Web Components library I am working on for "fun"

TL;DR left side is pivot (grouped) table, right side is declarative code for it (Everything except the custom formatting is done declaratively, but has the option to be imperative as well).

====

TL;DR (Too long, did read):

I'm challenging myself to be creative with the cool new things that browsers offer us. Lani so far has a focus on extreme extensibility, abstraction from dependencies, and optional declarative style.

It's also going to be a micro CSS framework, but that's taking the back-seat.

I wanted to highlight my design here with this table, and the code that is written to produce this result.

First, you can see that the <lani-table> element is reading template, data, and layout information from its child elements. Besides the custom highlighting code (Yellow background in the "Tags" column, and green gradient in the "Score" column), everything can be done without opening even a single script tag.

The <lani-data-source> element is rather special. It's an abstraction of any data source, and you, as a developer can add custom data sources and hook up the handlers to your whim (the element itself uses the "type" attribute to choose a handler. In this case, the handler is "download" which simply sends a fetch request to the server once and downloads the result to memory).

Templates are stored in an html file, not string literals (Which I think really fucks the code) and loaded async, then cached into an object (so that the network tab doesn't get crowded, even if we can count on the HTTP cache). This also has the benefit of allowing me to parse the HTML templates once and then caching the parsed result in memory, so templates are never re-parsed from string no matter how many custom elements are created.

Everything is "compiled" into a single, minified .js file that you include on your page.

I know it's nothing extraordinary, but for something that doesn't need to be compiled, transpiled, packaged, shipped, and kissed goodnight, I think it's a really nice design and I hope to continue work on it and improve it over time

Since becoming an "Architect", I've come to learn "Architect" simply means siphoning your will by eliminating all opportunities to actually write code ...

public class Architect {

static final boolean inMeeting = true;
private int will;

public static void main(String args[]){

Architect me = new Architect();
if(me.inMeeting){
while(meeting.active){
me.reactToEvents();
}
}else{
me.writeCode();
}

}

public void reactToEvents(){
if(new Random().nextInt() % 3 == 0)
will -= this.hit(this.face, this.palm, Force.CONSIDERABLE);
else
will -= this.hit(Office.desk, this.head, Force.MODERATE);
}

public int hit(Object object, Object with, Force force){
object.useForce(force);
object.moveTo(with);
return new Random().nextInt();
}
}

Somehow I wrote some code that ended up comparing a string type object against a pointer of an unrelated class. There are NO functions to convert that class to any kind of string representation. Let alone ones that would allow that to happen in C++.

WhyTF did the compiler not pick up on this?

Converting javascript/ typescript Map to json
or python date to json

or anything complicated to json is mostly ending with implementing serialization patterns

With date it’s so annoying cause we have iso standards that every language implemented or have libraries

so typescript doesn’t recognize Map<string, string> so you have to convert it to array and then to object

with python you need to make your own serializer / deserializer

So much waste of power usage that if only Greta know it she would say ‘how dare you!’
It can stop global warming.

I fucking hate the person that created the ionic timepicker its such a fucking mess if you want to do anything advanced and it's so poorly documented that most of the time you just have to guess what you should do. Best part: this fucking component doesn't even use a Date Object it uses A FUCKING STRING that it parses, so I have to parse, unparse, parse, unparse. Who in their right mind thought this would be a good idea?!

What frustrated me the most was when I tried to use their min, max functionality. I used the component as a timepicker, so I ignored most of the Date Object and just initliazed them at 0. Afterwards set the hr, min, sec and did the same for the max value. Doesn't work... It just bugs out and I can only pick midnight of that day... Okay. I kid you not: tried for two hours to fix this shit. Console logged the crap out of that thing. Everything seemed right. Out of frustration I then just initlialized the max value like normal, so the date is the current date. AND SUDDENLY IT FUCKING WORKED. WHY?! FUCKING NOBODY KNOWS. WHO, WHY, WHAT?!

Generic arguments can't be cast. List<Dog> can't be cast to List<Animal>, because any methods that take Dog as an argument would suddenly have to work with an Animal (same works the other way round with return values).

But there are many situations where this would be okay. For instance, a Date can be cast to a String, so if we know that no method directly or indirectly accessible from a ListView<T> (including accessible property and field setters) will ever take an argument of type T, then ListView<Date> can be cast to a ListView<String>. Conversely, if we know that methods of StreamWriter will only ever take arguments of the generic type and interact in ways that don't change the object, then we can safely cast a StreamWriter<String> to a StreamWriter<Date>.

There could be a pair of generic constraints signifying that the type only crosses the interface boundary in one direction. I think this would be an interesting feature, but I don't know any strict type system that allows it. What do you think?

Safari's developer console seems to truncate the output when a string longer than 140 characters is provided as a non-first argument to console.log or when it's a value in an object that is provided as any argument. You get ellipses, with no way to view the full string. Hovering doesn't get around it. Neither does copying.

Can we take a moment to recognize how absolutely retarded JS' event system is?
Events aren't objects. No, they're managed by an object, and identified by a string.
To subscribe to an event, you call object.addEventListener(name, callback). Because for some reason we can't just have an event object. Events MUST have an owner.
But to unsubscribe you don't call the function addEventListener returned, you don't use the token it returned either. No, you pass the same function to removeEventListener.
Because we don't use serializable tokens like in PP, and we don't return functions like in FP, no, we use functions as tokens, realising idiomatic DFP.

Some Java code I looked through to figure out how to accurately rework a mapping of value intervals to status colors:

• 16 levels of indentation
• Calls an instance method one line before a null-check
• Assigns that same value to a new variable and null-checks it again
• Insistently loops over existing HashMaps' entry sets to find a value by key
• Stringifies a Gson object, parses back the string and then null-checks the result.
• Mixes up the 'leq' and 'geq' comparison operators twice, which is why I went to check the implementation in the first place.

And this wasn't even legacy code. It's from last year.

Wow, function overload is worst than garbage! Is more like toxic waste!

If you run out of names for functions just name them:

DoThatStuff1(int, string)
DoThatStuff2(object, int, string)

instead of:

DoThatStuff(int, string)
DoThatStuff(object, int, string)

Your coworkers and future you will thank you!

I have to go through roughly 700 data entries and adjust the system's data accordingly ... by hand.😑
We don't have the tools to automate this reliably, the only available tool, tells you that an error occured and what happened but not where. That would've been kinda helpful...
The problem is that the respective data object of the data entry often contains a phone number, which has to be in a standardized format ... which it is not. Every number is formatted in like 10 variations!? A dozen different separators like spaces, commas, slashes and hyphens. And it must be edited manually 😖
My solution: built a goddamn chrome extension to format the string on click. Done. Saves a few seconds each time and a lot of headache in future. Of course given the correctness of the extension.

A CASE AGAINST BLUE PRISM
Let's review one of the worst weeks I had with Blue Prism

Monday: Yay! Solved one of the problems we've been carrying around for a week before.

One of the robots suddenly became slow. Like, REAL slow. A process that would take 3 minutes per record now takes 45, and that broke apart all the following schedule.

There were no updates on the application server, the production machine, the robot, it just became slow. And not always slow; a process manually run from console room would work, a process in debug room would work, it's just the scheduled part that caused problems.

It turned out, BP didn't seem to like that particular combination of schedulation + process + machine. Moving the process to a different machine seemingly fixed that. IDK why.

Tuesday: One of our processes waits for a code to appear in the page, and when that happens, it memorizes this code. However, now it is always returning blank. Worked for months, now it breaks every single time.
After half a day of debugging a bug which DIDN'T HAPPEN IN DEBUG MODE YET AGAIN, at 11pm I decided to just place a nonsensical timeout in page before reading and call it a day.

WEDNESDAY: a scheduled process didn't start. "No sessions created". Thanks Blue Prism, very cool.

THURSTAY: This time, schedulation did start, but the process is "waiting". As in: it's 9:30 am, the process has been stuck in the same step since 6:00 am. Turns out, it blocked during a navigate stage; you need to send a string to clipboard using the standard BP action for that, then paste and click "enter", but for some reason the standard BP object sent "ORRCO" instead of "ORRICO" to clipboard, which obviously returned no results and then... the process just didn't feel like doing things anymore. No errors, no logs, nothing: just sitting on its ass. Because fuck you that's why.

Friday: another process uses a very moderate amount of scripts to work. Nothing really fancy, just a couple of lines of code to place in page some IDs and selector to help BP do its thing, otherwise selecting these elements would be a nightmare.

But

Failed while invoking javascript method:Exception from HRESULT: 0x80020101-> at mshtml.HTMLWindow2Class.IHTMLWindow2_execScript(String code, String language)

The same script -it's not dynamically generated-worked yesterday, the day before and the day after. But sometimes it will not. Why? The answer, my friend, is blowin'' in the wind

Just need to pass this json object to the controller as a string. Cool just a quick two minute job.

Three hours later and it's still passing null, what the apple fuckery is your problem?!

(I'm not completely sure of what I'm saying here, so don't take this too seriously)

Settling on a language to write the api for ranterix is hard.

I'm finding a lot of things about elixir to be insanely good for a stable api.

But I'm having a lot of gripes with the most important elixir web framework, phoenix.

Take a look at this piece of code from the phoenix docs:

defmodule Hello.Repo.Migrations.CreateUsers do
use Ecto.Migration
def change do
create table(:users) do
add :name, :string
add :email, :string add :bio, :string
add :number_of_pets, :integer
timestamps()
end
end
end

Jesus christ, I hate this shit.

Wtf are create, add and timestamps. Add is somehow valid inside the create, how the fuck is that considered good code? What happens if you call timestamps twice? It's all obscure "trust me, it works" code.

It appears to be written by a child.

js may have a million problems. But one thing I like about CJS (require) or ESM (import) is that there's nothing unexplained. You know where the fuck most things come from.

You default export an eatShit() function on one file and import it from another, and what do you get?
The goddamn actual eatShit function.

require is a function the same way toString is a function and it returns whatever the fuck you had exported in the target file.

Meanwhile some dynamic langs are like "oh, I'll just export only some lang construct that i expect you to specify and put that shit in fucking global of the importing file".

Js is about the fucking freedom. It won't decide for you what things will files export, you can export whatever the fuck you want, strings, functions, classes, objects or even nothing at all, thanks to module.exports object or export statement.

And in js, you can spy on anything external, for example with (...args) => debugger; fnToSpyOn(...args)

You can spoof console.log this way to see what the fuck is calling it (note: monkey patching for debugging = GOOD, for actual programming = DOGSHIT)

To be fair though, that is possible because of being a dynamic lang and elixir is kind of a hybrid typed lang, fair enough.

But here's where i drop the shit.

Phoenix takes it one step further by following the braindead ruby style of code and pretty DSLs.

I fucking hate DSLs, I fucking hate abstraction addiction.

Get this, we're not writing fucking poetry here. We're writing programs for machines for them to execute.

Machines are not humans with emotions or creativity, nor feel.

We need some level of abstraction to save time understanding source code, sure.

But there has to be a balance. Languages can be ergonomic for humans, but they also need to be ergonomic for algorithms and machines.

Some of the people that write "beautiful" "zen" code are the folks that think that everyone who doesn't push the pretty code agenda is a code elitist that doesn't want "normal" people to get into programming.

Programming is hard, man, there's no fucking way around it.

Sometimes operating system or even hardware details bleed into code.

DSLs are one easy way to make code really really easy to understand, but also make it really fucking hard to debug or to lose "programming meaning".

TLDR: RTFM...

My dad (taught me how to code when I was a kid) was stuck serializing a Java enum/class to XML.... The enum wasn't just a list of string values but more like a Map(String,Object>.

He tried to annotate it with XMLEnum but the moment I saw this enum, I'm thinking that's unlikely to work.... Mapping all that to just a string?

He tried annotating the Fields in it using XMLAttribute but clearly wasnt working...

Also he use XMLEnumValue but from his test run I could clearly see it just replaced whatever the enum value would've been with some fixed String...

Me: Did you read the documentation or when the javadocs?

Dad: no, I don't like reading documentation and the samples didn't work.

I haven't done XML Serialization for years thought did use JSON and my first instinct was... You need a TypeAdapter to convert the enum to a serializable class.

So I do some Googling, read the docs then just played around with the code, figured out how to serialize a class and also how to implement XmlTypeAdapter.... 20 mins ...

Text him back with screenshots and basically:

See it's not that hard if you actually read up on the javadocs and realized ur enum is more like a class so probably the simple way won't work...

Dear God, why are punishing me by another bug report related to Edge?

Console dock freezes commonly for MINUTES, literally. It doesn't support objects, so every object is very usefully converted to "[object Object]" string. And now I am discovering that change event on input is magically not firing?

What a day. This would be solved in Chrome or Firefox in a matter of minutes, for a same time Edge doesn't even manage to render a page with dev tools opened FFS...

nothing new, just another rant about php...
php, PHP, Php, whatever is written, wherever is piled, I hate this thing, in every stack.

stuff that works only according how php itself is compiled, globals superglobals and turbo-globals everywhere, == is not transitive, comparisons are non-deterministic, ?: is freaking left associative, utility functions that returns sometimes -1, sometimes null, sometimes are void, each with different style of usage and naming, lowercase/under_score/camelCase/PascalCase, numbers are 32bit on 32bit cpus and 64bit on 64bit cpus, a ton of silent failing stuff that doesn't warn you, references are actually aliases, nothing has a determined type except references, abuse of mega-global static vars and funcs, you can cast to int in a language where int doesn't even exists, 25236 ways to import/require/include for every different subcase, @ operator, :: parsed to T_PAAMAYIM_NEKUDOTAYIM for no reason in stack traces, you don't know who can throw stuff, fatal errors are sometimes catchable according to nobody knows, closed-over vars are passed as functions unless you use &, functions calls that don't match args signature don't fail, classes are not object and you can refer them only by string name, builtin underlying types cannot be wrapped, subclasses can't override parents' private methods, no overload for equality or ordering, -1 is a valid index for array and doesn't fail, funcs are not data nor objects when clojures instead are objects, there's no way to distinguish between a random string and a function 'reference', php.ini, documentation with comments and flame wars on the side, becomes case sensitive/insensitive according to the filesystem when line break instead is determined according to php.ini, it's freaking sloooooow...

enough. i'm tired of this crap.
it's almost weekend! 🍻

Still as a scholar who has had his intership I decided that I was finally confident enough in my ability to apply for a small part-time programming job. I had an internship at a cool exhausting place with tons of expertise and I've proven myselve over there. So now I wanted a job on the side. Nothing special, just something that would make a little money with programming instead of washing dishes at the restaurant.

So I started at this small internet based startup (2 or 3 progammers) as a backend-oriented programmer. The working hours were amazingly compatible with my school schedule.

The lead dev also sounded like a smart guy. He had worked as a backend guy for years and had code running on verry critical public infrastructure that if it were to fail we'd be evacuated from our homes.

As a first asignment I got an isolated task to make an importer for some kind of file format that needed integration. So I asked for access to the code. I didn't get it since they were going to re-do the entire backend based on the code I wrote. I just needed to parse the file in a usable object structure. So I found out that the file format was horrible and made a quite nice set of objects that were nice. At the end of the first week or so I asked if I could get access to the code again, so I could integrate it. Answer was no. The lead dev would do that. I could however get access to my private repository.

Next week a new intern was taken to build a multiplatform responsive app. Only downside was that all the stuff he had ever done was php based websites. It wasn't going anywhere anytime soon, but I figured that that was where internships were for. So I ended up helping him a lot and taught him some concepts of OOP and S.O.L.I.D. and the occasional 30 minute rants of IndexOutOfRangeException, ArgumentException and such.

So one day he asked me how to parse a json string and retrieve a specific field out of it.
I gave him something like the following to start with:
"
JObject json;
if(!JObject.TryParse(jsonString, out json))
{
//handle error
}
string value;
if(!json.tryget("foo", out value).../// code continues
"

but then the main dev stepped in and proposed the following since it wouldn't crash on an API change:
"
dynamic json = new JObject(jsonString);
string value = json.myJsonValue;
"

After me trying to explain to him that this was a bad choise for about 15 minutes because of all kinds of reasons I just gave up. I was verry mad that this young boy was forced to use bad programming pracises while he was clearly still learning. I know I shouldn't pick up certain practises. But that boy didn't.

Almost everytime the main dev was at the office I had such a mindboggling experience.

After that I got a new assignment.
I had to write another xml file format parser.
Of course I couldn't have any access to our current code because... it was unnecesary. We were going to use my code as a total replacement for the backend again.
And for some reason classes generated from XSD weren't clear enough so after carefull research I literally wrapped xsd generated code in equivalent classes.

At that moment, I realized I made some code that was totally useless since it wasn't compatible with any form of their API or any of the other backend code. (I haven't seen their API. I didn't have access to the source.) And since I could've just pushed them generated XSD's that would've produced thesame datastructure I felt like I was a cheat. I also didn't like that I wasn't allowed to install even the most basic tooling. (git client or, Ide refactoring plugins, spelling checker etc...)

Now I was also told that I couldn't discuss issues with the new guy anymore since it was a waste of my valuable time, and they were afraid that I taught him wrong concepts.

This was the time that my first paycheck came in so I quitted my job.
I haven't seen any of the features that I've worked on. :)

taking hours for string/object conversion to handle cookie based favourites depending on number of clicks left me considering being dumb, finally seeing just 15 lines of code.

The bug: Some string values for an identifier property in the data objects are being sent from our frontend prefixed with a '0'. Sometimes. When it happens, it usually gets stripped away again by the time it's passed to our backend. But not always.

This 0 is never explicitly set anywhere. I even searched for a few variants of " = 0" in both the frontend and backend projects without receiving any results. You might already be suspecting where this is going.

So it turns out.

The data object which holds this value is being initialized in the aspnet (don't ask) backend and passed to the frontend, which then hydrates it. This value is always an integer number, albeit incidentally so which is why string is used as the actual type. When this object is initialized, it's hardcoded with an anonymous type where this property is set as int because I guess someone figured "it's always an int though". Being a typed language, primitive scalars can't be null objects which means the property's value becomes the concrete int 0.

Okay weird. I can think of better ways of doing this but let's just set it to string as I can't start overhauling things right now. Let's just go find where this value is somehow concatenated into the incoming parameter.

You see, this happens because at the point where the frontend sets this value, it may be an int or string depending on where it came from, and I guess someone figured that in order to cast it to string you just go prop += arg seeing as the prop is empty string and all. Because explicitly casting it or - as much as I get a rash whenever I see it - going prop = "" + arg would be too verbose and unoriginal.

Bonus round: How come the 0 only sometimes made it all the way to our backend? The thing is that this bug has been fixed before. The fix is that because this string is "always" an int, you can parse it to int before passing it to the backend in case it has leading zeroes. This path is only taken in certain views because someone forgot to copypaste their fix into all the places this is repeated.

Sometimes you find a bug and you are just somehow more grumpy after fixing it.

Horror story and rant time I guess...

I haven't seen the main developer of this MVC project that I've been working on but I can totally assure that his seniority isn't in frontend development 😠 and I doubt the backend too... Fucking DataTables converted to IDictionaries<string,object>

Guess who need to build on top of the pile of shit!

Anyway, I wasn't really careful about what kind of template I was given to work on a new SPA page, so I'm doing the job given the time, but it's fucking gory:
- matrioska style layers (n.3) without documentation
- partials everywhere
- too much inline styling
- too many <style> sections (n per layer or partial)
- too many <script> sections (n per layer or partial)
- poor CSS styling or no styling at all! (classes without any style nor js association)😠

He's just been lucky that the browser is capable of handling his shit

Now that at the end of this year I'll leave this project (solo fullstack) and need to provide documentation for the next poor souls I was thinking to leave behind something at par of my skills and capabilities but analysing the current mess ticks my brain in a bad way, fuck you Marco!
Fuck you
and your seniority
and the Italian way of perceiving seniority that gives you a higher living in the wrong side of the field 🤬🤬🤬

JS has
dynamic object literal keys
String object literal keys
Why aren't template literal keys allowed, and _why_ isn't there a proper error message for them?

Since I have learned Java I was taught that Java only passes by value. But my Uni Professor discusses that Java is passed by reference for object and string. I am really confused right now and need some advice.

I implore ANYONE... please...

Have you EVER written a SINGLE Jest test that didn't have some sort of bullshit spewing stuff like this:

"ReferenceError: You are trying to `import` a file after the Jest environment has been torn down."

"Warning: React.createElement: type is invalid -- expected a string (for built-in components) or a class/function (for composite components) but got: object. You likely forgot to export your component from the file it's defined in, or you might have mixed up default and named imports."

and yet running on a device, features work flawlessly and quite well, no errors or even warnings in sight logged

This is the most fragile pile of garbage I have ever seen.

I hate this.

inb4 your stupid ass todo boilerplate garbage you wrote tests for in freshman year. i'm talking about a REAL app with HUNDREDS of components.

where the grownup testing tools at? it's a question I've still not answered after a year of fucking around with this framework

Holy f-ing hell!
Why do the small things have such fucked up corner cases?
This is very likely a giant bug with Qt, but how does this even happen?
I am using Qt with QML and sending data to a database on the backend. I call functions in QML from a Date JS object (property actually, but it calls functions) to set the date as a QDate. This is stored in the database as yyyy/MM/dd. This is fine. When I read the date out I convert it back from string to QDate and send this object to QML. Which then converts this to a Date object in JS in QML.

But at the point where it converts from a QDate to a Date object it loses an entire day. Seriously? You didn't gain a day going from Date -> QDate, but you lose a day going from QDate -> Date?

How long has QML and Qt been around? At least 5 to 10 years. How has this bug lasted this long? I don't want to do a bug report. I will, but I don't want to .

Dear C++ / Java developers.

Please do not write Python, or do utilize helper libraries / pythonisms.

Not EVERYTHING has to be done by hand, it's not CS class anymore. Classes are fine too, not everything has to be passed as comma separated string. Python is proper Object-Oriented language, not scripting tool like Bash.

How should you override Equals in Java?

We have model classes with lots of fields and the we override the details equals to compare all the fields. I guess like doing a deep comparison.

And in all these classes SONAR is complaining about lots of ifs, complexity, etc.

And it's killing the analysis time... Old issue never fixed because whoever setup sonar was too incompetent to bother asking, researching, or fixing...

Is there a better way to override the equals to get the same result but without triggering SONAR issues?

Pretty sure this is a solved problem. And well if the top of my head, is just create a Util method that uses reflection like
Boolean equals( Object a, b, Class class)
foreach (String f: class, getFields()) {if !compare(class.getField(f,a),class.getField(f,b)) return false; }

return true

Can someone please explain me why the fuck the following doesn't work in C#?

Dictionary<string,object> mex=Json.Deserialize(data.ToString())as Dictionary<string,object>;
if(mex.ContainsKey("sessionId") && mex["sessionId"].ToString()!=tkn.GetSessionID())

the previous condition is NEVER true, but the following:

Dictionary<string,object> mex=Json.Deserialize(data.ToString())as Dictionary<string,object>;
if(mex.ContainsKey("sessionId") && mex["sessionId"]!=tkn.GetSessionID())

will actually work as expected.

tkn.GetSessionID() returns a string, of course
and mex["sessionId"] is always an object that contains a string in form of a number, like 152, 552, 6246 and so on and no, it won't ever contain any blank or other character

Fuck this fucking shit from C#

programming:

cc hello_world.c -o hello world

object-oriented programming:

cc -c hello_world.c -o hello_world.o
cc hello_world.o -o hello world

some weird shit:

class HelloWorld {
public static void main(String[] args) {
System.out.println("Hello World!");
}
}

“Just a quick fix” Classic start to a reported ticket. Ticket states that a form field was not doing anything.

Think to myself ok this sounds like a nice easy one for the morning. A few hours later I find something like the following written by a senior member of the dev team.

SearchClass {
//...list of getter and setters

Private $snakeCaseName;

SearchFunction() {
Foreach($this as $property => $value){
//... if property keys = string for each object property then do code

If($property == “snakecaseProperty”){
//...do stuff
}
}
}
}

Why does this loop exist!!!!! All it does is remove any error checking if a getter method is misspelt...

To make matters worse the entire search method was over 300 lines building a MySQL query string.... even though there was an ORM and entity classes available!!!

My new favourite RUNTIME c# error, evidently thrown by the standard library because the string is not present in the project:

"no parameterless constructor defined for this object"

🐟💩The image i fetch from s3 is of type byte array

I return it to angular as an ArrayBuffer

Which then needs to be somehow converted to an image so i can fucking show it

Then after research i had to convert ArrayBuffer to Blob

And from Blob to URL encoded object which returns a string that now shows the full image in img tag

Somehow, by a sheer of trials and error i have just accidentally made a very secure way of fetching a very sensitive piece of document (verification document with user's personal data on it) and now in browser this is shown as blob:shit-image/random-hash. Not even the file extension. This means nobody can download this image. You fucking cant. Its a Blob motherfucker! Like a Blob Fish. It saves either a .txt when you try to save it (no idea how) and if you try to open the image in new tab it shows gibberish text. This means you can read-only this highly sensitive document image and not manipulate it, not even download it. Perfect. I have just made a very secure software by accident.

(this blob fish looks like my shit)