https://news.ycombinator.com/item?id=39942397&p=2

https://news.ycombinator.com/item?id=39942397&p=3

Having no understanding as to the technicalities involved, the project was given the go ahead by the directors after several meetings with a vendor. After the CTO and I expressed our concerns about the scale of the project and the sheer amount of functionality involved, the vendor gleefully assured us that they were experienced with "migrations of this scale" and were more than prepared, which was music to the ears of the CFO.

Daily 2-3 hour meetings followed (for many months) to define the scope of the project. Within each meeting I sort of zoned out because it became very obvious that no only did the vendor not understand the scale of the work involved, but had started cutting corners everywhere/leaving out crucial functionality, and this was just the scoping stage, no development had even started yet.

I eventually departed the company but kept in contact with the CTO and learned that after 5 years (project was scoped for 2), the migration was abandoned costing multiple millions of dollars with nothing to show for it.

the key thing to understand about ERP systems is that this is their primary purpose. anything else they claim to do is secondary.

Perfect scheme.

Story time: people are dense as hell. This one idiot put in his personal SSN to open $XX/m a year POS account. This account currently manages just over a hundred locations, and this fool's TIN has just been raking in millions for a few years all because 'I needed to get us signed up!' And what's better!? They still can't get out of the wet paper bag, because otherwise "we'll someone might start selling candy or something unapproved" > we'll isn't that's what your franchise liaisons and agreements are for?! I'm just beside myself.

I'm so hyped on this because this one 'immovable problem' prevents me from having API access to my locations accounts. Everything just hurts. /rant

And then live forever with two systems running side by side.

In this case it sounds like a midsize business is using some random guy’s number, so the tax authorities are under the impression there’s a guy pulling in millions personally, and a company with no income. Given that individuals and companies get taxed differently in many different jurisdictions concurrently this will be a mess to untangle.

Sales is mostly just lying to collect a commission check.

French poet Paul Valery once said: "A poem is never finished, only abandoned".

It is the same for those projects. You just gotta keep paying forever.

Overall the product isn't really needed, and sorta sucks too. If I was an typical developer trying to pitch a startup idea to businesses, it probably would have never got off the ground and nobody would have wasted any time. Maybe eventually the developer would have landed on an idea so good it had REAL PMF, and made that.

But no, instead I sold some garbage and now I'm stuck working on it. There is such a thing as being too good at sales. You don't want sales talking people into bad ideas.

Is there a world where you hire someone to build a more maintainable replacement?

Why not put those sales skills to work and sell the business to someone else?

Deep down I know the product isn't that well liked, and a lot of the renewals are based on my salesmanship. I know I have a chance at fixing the bugs, because I wrote everything, while someone else would be screwed IMO. It just wouldn't be fair to sell it to someone because I know they'd be making a mistake.

When people keep using software, it is because it solves a problem. They might not like the software but begrudgingly use it anyway because it solves that problem. So sure, pat yourself on the back for good salesmanship, but also realize that they keep using it because even if it is covered in warts, it still does the job.

It doesn't really solve a problem. Not really atleast. It's a little better than using 6 Excel sheets via a google drive honestly. The reason it got sold is because in theory it sounds a lot better than Google docs. But it's only a little better. And it has its own host of difficulties (can't modify anything for example.. don't even ask me to add a column or rearrange a table for you).

I know it's not great because if I get a company using it, but don't hand hold onboarding to an extreme level, they never even really get going with it.

When selling it, if I don't really lay it on thick and do a spectacular job, they don't buy.

The reason they don't quit once on it isn't that it solves a great problem or is much better than their previous solution (almost always just Excel sheets), but because switching back to their abandoned system is harder than the $300 a month is worth. So they just keep paying.

Also, I've contacted and pitched wayyy more than 34 companies. Many companies are/were smart enough to pass on it.

I'm not writing for a sob story or to talk up MY salesmanship for whatever reason, but I wanted to share my main takeaway from this project.

When doing lean startup and trying to get your first customers, don't think youve struck gold just because you had a great person schmooze a few people into using your MVP. It can be a false start just based on their persuasion. At this point I'd rather tone down salesmanship a lot and see if I can find a product that really strikes a cord with them instead. Less false starts that way IMO. I'd much rather have gotten 1-2 costumers only, not 34, and realized this isn't worth pursuing.

It sounds like bringing them and you together would basically be a guarantees success.

By the way; I am working as a self-employed developer and my favourite project is a software that has initially been written by a self-taught programmer who both started and ended his career with that software. As far as I heard (he was already out when I joined) he got burned out and said that he never wanted to work with anything software related again. He used PHP and jQuery. I've mainly added new stuff instead of touching the old code, while only only refactoring the old stuff where necessary in very small increments. Most other devs really hate to touch it, but I don't understand why. Of course it would be better to replace it with a new version that has been built on top of a proper framework, but their management is too stubborn to understand that an incremental approach is the better way of handling this. So instead, they try to get a "complete understanding" of the project and try to create a completely new version in a "big bang" approach. This usually takes a few months or even a year until this new replacement project is considered a failure while I keep maintaining and cleaning up the old project. It's been six years for now and even their most "optimistic" people currently say that the old software will be running for at least two more years. Having built up a lot of knowledge over that time, I could easily create a new version in less than half a year (I actually think two months, but I'm tripling my estimation for safety), but that would make their management look bad (long story), so I don't get the green light to do that ¯\_(ツ)_/¯

"Two month sound too optimistic" you say? Well it's an inbox, an outbox and one form with a few calculations in between. I have created way more complex software than that in the last years.

Reality would be that a combo of a good developer and a good salesperson would get into a similar situation that I got myself into, but with a better more manageable code base. Not the worst outcome in the world, that's basically most b2b saas technically.

Long ago i made a few html websites for businesses then explained basic html and ftp to the owner. It was perfect except from other webdesign shops hammering my clients.

Depends on the type of sales. A pretty good indicator is how many times a customer makes a purchase from the same salesperson. If it's just one purchase (like ERP consulting) it could be grifty/etc like in this case. If they're buying from the same guy for years, there's often very different types of salespeople. I used to think all salespeople fit the sketchy used car salesman type, but after working with great salespeople I know better. This is a big blind spot for us techical/engineering types.

Estimation is hard but it seems it's far too easy for the vendor to blatantly lie with all the risk on the customer.

the #1 cause of failure is summed up greatly by Isaiah Bollinger, paraphrasing "most bad implementations are because people are trying to buck the system they bought, rather than work with it, understand how your ERP, eCommerce or other system does a workflow and match it. There's billions of dollars going in and out of Shopify (or x system) daily, and you are not that special. You will spend 10x as much trying to NOT use the system rather than trying to use it".

Overcoming process inertia is profoundly expensive and often demoralizing to teams. The project budget for a new system is often pitched as vendor price plus some internal oversight, but this fails to represent the cost of the project exactly because adapting the workflow of a whole division or organization inevitavly costs some multiple of that budget while vendors, consultants, and internal spearheaders all pretend it's negligible.

You're right, ultimately, that failure to adapt is the final damning issue in many of these projects but the root cause of the failure is often that nobody sincerely quantified just how costly and disruptive it will be.

What makes the clients we work with great and unique and what I love is the products they make or the problems they solve for customers, but they're not tech companies or banks. The ones who succeed are the ones who focus on thier core value and core skills and not random accounting process X Y or Z.

Besides the ERP consulting bit we sell niche solutions in our product space. Mostly comms(EDI, ecommerce,etc) or payroll/bookkeeping addons.

This is insane cope. We make technology to assist end users perform the tasks they do. To say "well you're doing the task wrong, the tool is made so you do task X way instead" is to put the cart in front of the horse.

Those big ERPs are not really blank canvases, but when you need those micro-customizations in the wrong place, they can become one.

End users involved in the integration don't really want to learn new processes or even do things slightly different, as they know that changing process often involves burning a lot of political capital and they often lack awareness to know that "just using the ERP the way it's intended" is cheaper. And consultants are experts in finding a chance to perform those micro changes. It's a perfect marriage.

Its why picking the right ERP is very iimportat

If you do things the "native" way in Azure or AWS, you'll be fine, just like millions of other customers.

If you try and make the cloud work like your old data centre platform, then you'll have a bad time.

I just watched a customer spend $2M to deploy software routers to replace the "bad" cloud-native routers. Now everything is more difficult, slower, and just all-round bad. But they "had" to do it. (Narrator: No, they didn't.)

In your example, the customer may have had software that depended on the routers having some functionality that the cloud native routers didn't. Sure if they had designed for that cloud from the beginning it wouldn't be a big deal. But now, that $2M might be less than the cost of changing all their other systems to work around the limitations of the cloud native router. I've seen situations play out like that a few times.

In this particular case they were sold a product that serves one purpose: multi-cloud solutions across international boundaries where no single telco can connect all of the data centre locations.

Their handful of locations are all in one city and well-connected by multiple telcos because... they're a state government, not a multi-national corporation. They're blocked by the constitution from expanding inter-state, let alone internationally. That would be a literal act of war.

That didn't stop the vendor's sales team showing slides with titles like: "What if you need to expand into the Chinese market?".

E.g. many professionals jump from technology to technology, rather than mastering how get stuff done with one so they end up being mediocre their whole career.

I have yet to meet a single company which works like ERP are designed to work. This is their edge over competition, their reason for existence on brutal market. And ERP wipe that out, with the most expensive wiper you can imagine, while selling various bullshit left and right, and constantly lying to given company how everything is fine and under control.

Truly, a way to kill a company. The fact that some survived it all to tell a story just shows how resilient whole such org is to such a massive stressor that ERP migration always is.

They sell first and foremostly a lie - that you can have cover-it-all system just like competition, to match your unique way of working, without suffering tremendously, fit like a glove out-of-box (when reality is exactly opposite). It just never works, more like hammering a concrete glove on your progressively more disfigured hand, while being told how rosy your future will be.

"We want the new system to work exactly like our old one"

It's probably not worth the money...

I had one guy want me to recreate Microsoft word, worts and all, because his current Word wasn't doing it for him

I never worked in those ERP companies, but a few times I've been on the receiving end, working at companies undergoing a large migration.

It is very often stuff that doesn't really matter, is highly inefficient, and requires small changes everywhere in the system. Death by thousand papercuts.

There is no incentive from both sides to change: the company wants to keep modifying to get $$$, employees don't want to change how they work (because change is often stressful), and the person paying the bills is not getting the full picture.

If there's anything that is actually really "unique" (in a good way), then you spend money. Often this means not customizing the ERP, but actually writing new software that integrates with it.

But the RFP said I had to do it, and the customer insisted that I follow the letter of the RFP, rather than hit the high value parts of the project.

The two problems with this are that the cost of automation far exceeded the possible future cost of the manual process, and working on this automation took oxygen away from the stuff that would help people.

Needless to say, it was a disaster.

Trying to build a total second system is nearly impossible if it's big and old.

I guess success and failure are both group efforts.

If you're not part of the solution, be part of the problem.

128k/ year, while RT still runs in a VM for all the other ticketing queues.

Lots of powerpoints.

I feel this on a deep level since the ERP system at our company is in the middle of moving to one of these big vendors and so much is simply not working out, extending timelines and development cycles. Cost only goes up and ROI is pushed out beyond the horizon.

PLM is equally bad and a horror show of legacy vendors trying to sell their solutions with the promise of customizability. Again the small, more modern players in the PLM space are constantly ignored for the big legacy ones and it turns out those legacy platforms even in their latest iterations are inflexible at best or downright hellish at worst. The reason behind all of this? Service contracts and vendor lock-in are the main drivers of value for these vendors, rather than quality (and modern) engineering.

I read this and immediately became sick to my stomach. I despise working in corporate america.

America might lead the way, but corporate anything, anywhere is pretty terrible.

I have a current client that is a mirror of this exactly. They had one new executive go to a sales demo, and the sales guy told him it would be "live" in a few months, and he believed it. It was so laughable and downright embarrassing for the client. In this case, I know it is at least a 3-year project if you have competent team members and vendors. It also depends on how you define "success" and how many business processes you are willing to break. However, these can stretch into 5 year projects when you are working with mission critical systems running accounting and payroll. In those cases, you cannot move fast and break things.

Most organizations don't have competent teams, and the vendors are often made up of low-skilled offshore workers, which makes nuanced, complex projects very complicated. The sales side will promise the world then the implementation team is low skilled and underdelivers.

I have seen this with several other clients, replacing pieces of PeopleSoft modules with different applications. Sometimes they come back and sometimes they just suffer through it and then find a third option after the new system fails to deliver.

Most ERP systems are just CRUD systems. So, you end up with a working CRUD system being replaced by a new CRUD system. The old system had its strengths and weaknesses. The new system will solve some problems and break others.

You spend 3-5 years replacing a system only to end up in the same place you were before: A CRUD system. This is maddening unless you are the new vendor and getting enriched.

Anyway, the only task they ever gave me was to build Cisco router ACLs to match existing traffic in the stores; they needed to implement access control but wanted to make sure they weren't blocking anything important. Because many of the stores had been opened years before, there wasn't necessarily a consistent IT stack in the store; there were a lot of one-off solutions so they wanted a universal set of rules with per-store exception lists.

So, every week, someone would drop a few terabytes of network logs in an FTP server, separated by store, and I would distill them down to a set of rules; if something was being used in more than X number of stores, it went in the universal rule list, if it wasn't it went in a store exception list. At first I was doing it semi-manually, but eventually I built a database and wrote some SQL to mostly automate it - got it down to about two hours a week, most of it waiting for the DB, by the end. Once it was done, I would send all of my updates to a network engineer who was the designated point of contact for the access control project.

They had a lot of stores, and for some reason could only supply a certain amount of logs weekly, so when my contract was up there were still a number remaining. I tried to set up a meeting with the network engineer to go over the work to be done and the automation I'd built, but he never responded. Eventually I tracked down his desk in the vast corporate complex and paid him a visit. He was pleasant but told me that he wasn't even on the access control project and he thought it had probably been canceled at some point. He had been dutifully copying my updates to a network share somewhere in case they were ever needed. I gave him the SQL scripts and the database info and he put them out on the share where they probably still are today, a decade and a half later.

So that was six months of my working life.

I love it. A co-worker of mine told a similar story. For at least six months, his only regular responsibility was to compile a report every week and email it to a distribution list for the headquarters one echelon above us. It wasn't a ton of work, but it was all manual and he was meticulous. He dutifully did this without any feedback until someone finally replied to simply say, "We do not require this report, and have never required it. You may choose to either continue or discontinue sending it."

BP: 'We always like to avoid confrontation, whenever possible. Problem solved from your end.'

What a line.

https://wm-help.net/lib/b/book/2895812152/59

Added random sleeps to slow down performance. Random alert messages about fake errors. It was weird.

EDIT: Since this is getting some votes I'll add some more details. He would also come by to tell me how happy he was about all the complaints he was getting about it.

Long story short, last time I heard, both the original hack and my partial replacement were still in use ten years after I left. The big project ran for almost two years without replicating a single feature of the original hack. Right up to the last, the fancy guys produced an impressive series of complaints and excuses that basically said that they were doing everything right but doing things right didn't work because the problem and the context were wrong.

They hated it.

I added moderate, random delays before emitting each number. They loved it, and used it every week for quite some time :)

Back in the nineties my dad had the notion that if you make the computer select numbers at random many times, and you run statistics on the results you gain some "legitimacy" to these selections. So he asked me to write a "lotto number selector program", but it needed to run for a few hours and select many numbers and then output the ones that were selected the most or some such. Maybe it was more sophisticated, but I can't really remember the details. I guess I could just add a delay instead of actually selecting the numbers, but I wouldn't lie to my dad :)

It was super silly, but like you said, why not, I was a teen/tween and I didn't mind playing around with silly software.

He actually used it and actually filled out the lotto numbers based on it. No, we never won the millions :)

Also - I just visited a casino in Spokane, WA for the first time ever. Isn't that what all the machines there are doing? A random number emitter thingy with random delays, animation, music and flashing lights?

This is the sort of attitude that ensures you don't win!

A lot of these "certain delay for 'correct' response" rat feeder types of mechanics. Even when you know "the hit" is pointless or a waste, still heavily compelling anyways.

Cigarettes have a similar experience if you smoke too long. Days of "I don't 'actually' enjoy this anymore", yet my body still demands something that makes me vaguely nauseated now. Except, per the original comment, it is still the "suspense for a reward" that no longer exists.

I still don't really know what there isn't to trust about generating random numbers for the lottery.

Usually you have to spend years working away and trying to guess why everything you're told to do makes the software worse.

frankly it sounds kind of fun, but i'm sure it was not in reality.

Where do you draw the line?

Did a lot of great work with good people there too. Projects that lasted for nearly a decade.

Previous user took a very huge risk. I've seen similar stuff happen, you can get sued (along whoever told you, but you need proof) just for the sake of making an example in front of the rest of the company.

That said, I think it'd be a pretty tough argument to make that an employee following the instructions of their superior was breaching fiduciary duty. If they weren't acting to their employer's benefit, then whose? Employee fiduciary duty cases tend to be more about things like embezzling or competing with your own company while you're still working there.

An example of Fox getting shut down for trying to claim that sexual harassment was a breach of fiduciary duty: https://www.barnespc.com/news-articles/limitations-on-the-sc...

There has to be proof about that. If everything was verbal how do you prove it?

There's no way around the fact that sabotaging your company is illegal and a breach of contract.

Also, still it is not illegal. Just that the shareholders need to sue him to matter at all.

Incentives and adoption goals for internal tools are weird at large corps. Intentional suckage is just another tool to drive migration

Heaps of people online have stories about how Microsoft tricked them into upgrading Windows.

I was late to join the team, but when I did, they'd already bought and branded thousands of meters that were boxed and ready to go in a nearby warehouse. The team had already built a number of APIs exposing controls for these meters as well as various monitoring and reporting interfaces. A UI already existed but it had issues, my job was to come in and fix that and get us ready for release.

We worked hard for a couple of months and whipped this thing into shape. Meanwhile a multi million budget was lined up for the marketing launch. Adverts where drafted, installation technicians where trained and merchandising was branded. All systems go.

It all drove towards a set in stone deadline and we busted our guts to get there... When the day arrived, there we sat. Ready. All features built, no bugs that we knew of. Ready to hit the 'go live' button. Honestly in 30 years working in the industry, that was the only time I think I've ever been in that situation.

Our product owner walks into the room and says, "erm, there's a couple of issues we need to discuss at board level. Hold tight guys I'll be back".

So we sit...

Two days later he returns, "erm, guys... the board aren't sure that this product is on brand. And they are concerned that if it fails it could be bad for reputation. So, we're not launching."

So we sat... for a month... working down our contracts whilst I taught myself Node.js

Ouch, I like to think it's the other way around.

Most executives in big tech are firmly in the ego-driven camp and get high on their own supply of hot air. And they really believe that everyone’s function in the company is to serve them.

Notably, there are some leaders who are very mission-driven and see their people as peers. Usually they are quite known for it, as it’s rare.

There were four control boards per panel and four panels per satellite and a projected order of 12 satellites so were going to have over 200 boards (enough plus spares) to test and last go-around it took two days to test a single board with failures due to connector issues and I wanted to get that down to two boards per day and no failures.

It was my masterpiece. Place a board on the fixture, pull down an assembly, and six micro-D, two nano-D, MCU and FPGA programming connections, remote thermal and multimeter probes, and a canbus connection with onboard amplifier to traverse the absurdly long cable all slotted with perfect precision into the board and were routed back to a test rack via a single cable bundle. Automated test bench scripts powered on the board, took voltage measurements, programmed the devices, and ran a suite of tests automatically and then printed out a report with a signature block for me to sign at the end.

The board experienced no stress, every connector was mated with perfect force at the perfect angles, and it would cut down the amount of time needed to test the hundreds of boards we were going to build by months.

It was featured in marketing materials for my employer.

We built four of them and with all of the labor they probably cost a quarter million each but would put the program back on track as it was delayed.

Additional delays in another part of the program led to the government cancelling the entire thing and we turned over all of our tooling and prototypes to the customer who probably put my fixtures in a warehouse for fifteen years before selling them for scrap value.

What they asked for: Refreshing their suit of unit tests, 10,000 of them, that had gone stale (that applied to their widely used C++ telephony library)

Budget NZ$1,000,000

(Don't ask what, why, really?)

I joined half way through

Half the budget spent, nothing to show, mission mysteriously changed to producing a deep packet inspection tool. (What happened to the unit tests you ask? Be quiet! Take the money)

My job was to reach deep into their system and bring the data to the surface, roughly speaking

Thing is nobody that worked on the library wanted, needed, or asked for any aspect of what were doing

The design for it was written on a whiteboard by the project lead before I joined

Then the whiteboard was cleaned.

So the design was in one man's head

That man, the project lead, quit 3 weeks before delivery

On handover day I was a bit afraid, I had taken care to keep good records of my obeying (increasingly unhinged) instructions, so I was confident I could not be blamed (I was younger and naive-not part of the story)

But I fully expected an unpleasant "you did what with our money? What were you thinking?" type of unpleasant meeting

Instead the corporate types sat around a table gushing how wonderful everything was, what a brilliant job we'd all done...

I was stunned. I understand now, I learnt that day, that it was in no-one's interest to acknowledge the waste of a million bucks

Coming from academia, as I was more or less, this was a valuable lesson.

But utterly worthless software for a million dollars

I still use some of the furniture I bought with the money....

I just left a company where I was contracting with a few other people for about a year. About $1.2MM down the drain. There is no hope of spinning that product up. The owners "suspended" the project and the entire brain trust evaporated. Imagine the bank owners from the movie Mary Poppins interfacing with technologists and startup founders. Whoosh...

Academia, where there is a reasonable budget but it’s completely impossible to spend is as the cycle of meetings, emails and fucking around is eternal.

I heard Spark was worse, but this was not it.

The funny thing is that in my current job, they use something called "Dailybot" or something integrated to google chat where you check-in and once a week, give an emoji between 1-5 about my productivity or something... Unhappy face for me.

What value is this supposed to provide? If you take part, it will be 8-9-10 regardless to "look good" and if you don't take part like me, it drags the entire team to an average of 4-5-6 which is the same as not having the system at all and yields the real outcome of average.

Almost everybody knows its BS but nobody is able to mention it

In other word pointless. The only time you can even halfheartedly trust the results is when the peoples surveyed are almost adversarial- like customers.

This is often sad and depressing, but sometimes it's just about doing your job, and getting some good food after work to counterbalance the bullshit.

Lean, well-run businesses rely on people who are sharp and willing to search out problems in need of fixing. This is necessary for real innovation and competence.

You see this pattern not just in business, but it all kinds of situations. For example, governments that harshly suppress their populations' speech gradually turn into economic backwaters (requiring them to suppress even harder and harder to stay afloat), while governments that foster the exchange of ideas and roll with the punches become the world's economic and geopolitical kingpins.

If management is unable to deal with people that rock the boat (as in firing should there be actual problems) then they seriously lack backbone and leadership qualities since spouting corporate propaganda and playing silly conformance games with employees is probably all they can do.

I understand if the data is not used or is used to falsely claim high employee satisfaction, but nothing you've said suggested these systems didn't start out as good faith efforts on the part of the company to track and improve employee job satisfaction.

I don't see how your behavior isn't just childish insubordination.

Thank you.

After acquisition - we were handed down the order to migrate to AWS.

This was after (in the mess of the merger) the colo contracts were basically ignored and not renewed. Once someone within the company realized the issue, it was the 11th hour.

After many, many attempts to discuss our (Operations team) concerns, we abandoned our protests. It was clear the new CTO wouldn't cave and sign the contract.

Some superficial testing was conducted and the order came down to move...NOW.

We began moving hundreds (maybe thousands) of very resource hungry DB servers first (there was no way to use something like RDS without major app/config changes).

Once the AWS bill came in, the CFO blew their lid and within 90 days we were migrating BACK to our DCs (and the millions of dollars of hardware we left idling).

He holds up his hand to stop me speaking, says “you have an A. Don’t come to class anymore.”

So, I never went to class or did any of the assigned work, instead working on the simulator in my dormroom. At the end of the term I got a failing mark. I went to Tewksbury, and he had no recollection of having told me that I shouldn’t come to class. I tried explaining myself about the simulator, and showed him. He grudgingly agreed to change my mark, seeming suspicious.

Despite my attempts to work with the dept, the simulator never got used.

My "filesystems and database design" class was basically howto use mysql. It's a shame, I was more interested in actual file system and database design

Just stupid stuff like that made me drop the academic bullshit and skip into the real world.

The database class at my uni was mostly relational algebra. Likewise, many other classes were mostly foundational theory (though we did have project-based applicative classes too).

The final nail in the coffin was me failing calc 3 twice so I don't think they were unacademic

This failure also may be a sign that it was the wrong school. A course taught badly or indifferently can make a subject tedious or difficult, that when taught well is exciting and easy.

In my first exam I got a 90/100. I got all the queries right (don’t get me started on programming/writing queries on paper and how stupid that is), but I lost 1 point on each of the 10 questions. Why? Because I didn’t put a semicolon at the end of my query. Something I had never done in any database tool, never done in my code, and only done 1-2 times on the CLI if that.

IMHO computer science in college is a joke at most places. Teaching things 10+ years out of date by people who have a chip on this shoulder towards anything new. I had an EE professor who literally did not go a single class without find some way to denigrate web developers and “not real developers”. Fun times.

Still a dumb reason to deduct points though

What it really was was a forcing function to figure out auto formatting tools, so in the end I didn't mind.

I hate to agree, but yeah.

I'm watching my husband go through a CS degree course right now, and the stuff he talks about I'm just like "why are they focusing on this, you literally never use this."

His "computer graphics" class was all about using the original GLUT library. The school specifically highlighted M1 Macs as good computers for their CS department students to use. Think about this combo for about 5s.

Yeah, the lib wasn't available in arm format...

Outside of my university coursework, I have never used the Pumping Lemma (Theory of Computation), balanced a red-black tree (Algorithms & Data Structures), relational algebra (Databases), and all sorts of other things that someone focused on the software development aspect of computers would literally never use and would wonder why we're focusing so much on these things.

I grew up very poor, so finishing uni was like, an achievement I had to do, but I learnt C++ and HL2 modding and other computer stuff back in highschool.

My uni experience was 7 years of hell (multiple gap years to go do real work before coming back). Getting stuff like "pseudocode can't have an equal sign in it, so NC" on assignments.

"Even if you finish all the assignments you must attend every tutorial or you pass"

Ive told many academics to go to hell. I'm not paid to be there like they are.

My grades go from "barely passed" to "high distinction" like a rollercoaster each year.

- and yet, I still really wanted to like uni. I do still want to like it, it's just a shame about the academics

My boss came by one Friday morning and informed me i'd be onsite working all weekend and told me to "ensure not work over 40hrs" this week. This is an impossibility, I had been working late all week and was already at 45 or 50hrs for the week, and I still had Friday/Saturday/Sunday to go.

So after the discussion, I sent him an email, casual, saying I've cancelled my flight back home for the weekend, confirming i'll be onsite all weekend and that I'd ensure to "only log 40hrs this week".

He came by my desk furious and said if I "ever pulled a stunt like that again there would be consequences"

So I re-plied to the original email, took him off, put on my personal gmail and just recounted the entire episode and sent the email again.

It was a great story until that point and I want to know what happened next, I feel I’m missing something.

> they would put pressure on their workers to work overtime but not log the hours.

The email was to get in writing that his boss told him to lie about how many hours he worked.

Another kind of sad Tewks story is, once I went to see him in his office, and I witnessed him giving a grad student and early edition copy of a Claude Shanon tome. Clearly was intended as a nice gift. The grad student didn’t get it, and Tewks had to explain who Shanon was to a EE PhD candidate. I wanted to jump up and down and get his attention but I was just that pesky undergraduate with some scheme to get a free A…

Anyways he was a Bell Labs guy and full of lore and wisdom, but I would have loved to have known him as a younger man. Same with Harold Stalwen.

I did have the great honor of taking courses with Roger Pinkham and Norman J Morgenstein Horing who despite both being in their 80s were possessed of incredible intellect and didactic powers. Pinkham in particular was the closest I’ve ever come to meeting Gandalf.

Turns out the customer didn't read the (enormously expensive) SOW. They don't want the product, can't ingest its output, don't want to do the work necessary to implement it, and on a recent roadmap review listed its function as their absolute last priority. I am not sure whether there's been a change in management, or some salesperson talked really fast, or what.

I am trying to appreciate this as 'salary for nothing' and use the time to study for other things but it turns out that for me this is an anxiety-inducing and unhappy experience.

It's like Bret Victor's story of Puddles. Imagine you adopt a puppy. You name him Puddles. You take puddles home and you give them a little snack and you stick puddles in the cage and you lock the door forever and never open it again. Most, although not all, would agree this is cruel.

We have a notion of what it means to live a full doggy life. Dogs have to run around. They sniff other dogs and they pee on things. That's kind of what it means to be a dog. Dogs have a set of capabilities and we recognize that a dog has to be allowed the full free expression of its entire range of capabilities in order to be a dog.

The same is true for people. Like a dog needs to hunt, people seek meaning. They wither away without it.

https://vimeo.com/115154289

They could get software consultants anywhere, especially where they were not needed.

A former colleague of mine got assigned along another to assist an Italian bank with their software projects.

He told me that in his office there were 6/8 people, barely doing any work. Two of them were leads in that office and did their best to never ever be seen or meet anyone. They would just close themselves in the office the whole day doing god knows what. If you knocked they would pretend to not be there.

My colleagues tried to get some work to do desperately for months, but consultancy said to shut up because the contract was huge.

He says it was one of the most miserable periods of his life because "there's enough Youtube a person can watch for 18 months" and there were more than 2 hours per day of commuting on top of that.

You have two choices.

1. Start a side gig. Be busy at work make extra money.

2. Work on open source. Find a project you like or want to work on and do that.

PS: set up a home lab and learn to love SSH tunnels.

Doing what depends on the business - it can take some detective work to figure it out.

Then again, the detective work could quickly show this is someplace you don't want to be long-term.

In which case yeah, find something productive to do, or treat this as that paid learning time to build some new skills.

Best part of learning is test it's defensible ... "oh, I'm busy doing an assessment on the pros and cons of rewriting the system using node.js. "

Don't measure yourself by goofy corporate politics. It's not your fault (and ignore their political blame games). You can make the world a better place by the pay you are receiving.

Hell, maybe try work two jobs if you can!

After a quick debug session, the problem was clear: the methods to call was being searched using a linear search. I changed the search to use a hashtable and the function disappeared from the list of most CPU consuming functions.

After running both implementations for a couple of weeks to make sure my implementation was right, I made a PR, which wasn't approved. The manager said my change was too risky to go to production, even though the implementation was simple and I spent two weeks making sure it was ok. They're probably still using the linear search.

I was a junior programmer helping out with a Java applet game that ran in the browser. It had performance issues, which I narrowed down to a “ticker tape” text animation. The position of the text in pixels was being updated by a background thread, and that thread would leak each time when the user switched screens. If you clicked around enough there would be hundreds of threads all updating the same shared variable.

I replaced a thousand lines of that nonsense with a single line of code:

    return getTicks() * velocity % tickerWidth;

I got in trouble for “making a mess”. The developer responsible for the previous code had been busy (for weeks!) “fixing” this code and couldn’t merge his change. So it got reverted and they spent another month tuning the threaded code for better performance.

I waited until they got distracted and gave up, then re-merged my one-liner. They were convinced they had “fixed it” and never looked at that code again.

The responsible coder got a pat on the back for his hard work and I got a disciplinary meeting to discuss my “behaviour”.

This was at a startup that burned through ten million dollars and then was shuttered because the software was basically garbage and couldn’t be sold to anyone.

Even insecure people will take better code in most circumstances. Maybe after adding their own touches to it so it doesn’t seem like you got it right or came up with a good solution on your own.

It’s the ones that don’t understand the change and don’t want to admit it that are the real problem. “Too disruptive = I don’t understand this and don’t know how to support it”, if, in fact, it was a simple search->hash map change which any developer should be able to grok.

I've never shared it with the magic community because I think wizards of the coast would just send me a cease and desist, so three years of work is just sitting idle.

I'm still not sure how I feel about it.

You never know what opportunities this adventure might bring, but certainly more than just keeping it on your hard drive. Go for it mate.

Heroku restarts their servers once every 24 hours with only thirty seconds' warning. Since the games are in-memory, this of course kicks everyone out of whatever games they may be in the middle of.

I guess the solution is to have the games be on Redis instead of in-memory? I'm a bit more front-end oriented so I was bewildered to learn this was how it worked in the first place and I'm not 100% confident in my solution.

Redis would work but I'd be afraid of race conditions. There probably wouldn't be any, but it's something to be aware of.

If you wanted to test the waters with at least the relative public, I'd consider doing step 1 of renting a server, but then hiding it behind a Tor hidden service. Inconvenient to connect to? Yeah. (Probably) safe from the Hasbro demons? Also yeah.

I'd be happy to help if you need anything! Whether hosting or just backend related questions!

Also depending on how hard coded your architecture is, maybe you can separate out the Magic specific data from the app itself? So the MTG specific data could just be loaded as a datapack?

Feel free to message me and I’d be happy to help with server stuff.

[0] https://dokku.com/

An open-sourced codedump will survive best if people (which effectively means "average technically-minded MTG player" in this case) can run it wherever they're most comfortable - which will almost certainly be something that smells like a VPS. Mayyybe a container.