|
mystes posted:Capitalist/proletariat i am v late to this discussion but wholeheartedly suggest that we use chatlanin/patsak: https://www.youtube.com/watch?v=Pe66FcnlYpA base leader election on the color of the database's pants
|
#
¿
Sep 15, 2018 05:22
|
|
|
|
| # ¿ Dec 10, 2025 14:42 |
|
|
can we talk about how wonderful the 506 word salad isthe IETF posted:The 506 status code indicates that the server has an internal explain what the gently caress that means without context
|
|
#
¿
Sep 21, 2018 07:44
|
|
|
MononcQc posted:HTTP revisions are kind of poo poo, 4 of them all use 1.1 as a version but coexist in at least 4 distinct standards. thankfully we now have HTTP/2, aka "content-length mismatches are now a fatal error, get hosed" not that the mismatches didn't not cause issues in http/1.1, but half the time it would cause request/response handling to only degrade somewhat and still more or less succeed im curious if the [hard fail in http/2](https://httpwg.org/specs/rfc7540.html#malformed) was actually necessary to implement the stream multiplexing stuff (decent chance that it is, idk, don't want to wander through a bunch of old ietf mailing lists) or just that one of the authors was so fed up with handling weird length mismatch gremlin bugs that they wrote the spec that way (much more amusing)
|
|
#
¿
Sep 22, 2018 04:56
|
|
|
Sapozhnik posted:Still don't understand why people use vue instead of react i assume react is probably not any better, but my knowledge of vue is that it turns any js traceback into a giant pile of inscrutable utility loader functions, half of which are anonymous. trying to figure out which levels are poo poo i should care about (our code rather than vue's poo poo) is a painful and tedious process, and the sentiment of our frontend devs is that they have no idea what to pay attention to either. i do not understand how or why the stack that arguably has the best live debugging tools around has apparently gone out of their way to then make those tools unusable
|
|
#
¿
Oct 17, 2018 07:50
|
|
|
ThePeavstenator posted:Taking about anything related to sorting, paging, UI appearance, etc of search/query results is a huge waste of time that could have been better spent on important topics, like which of 200 or 404 is the real RESTful response code for a webservice to respond with for a valid, successfully processed request that contains no results for the query parameters supplied. did nobody raise that 204 exists for this specific purpose
|
|
#
¿
Dec 31, 2018 20:23
|
|
|
Oneiros posted:
we're migrating from zendesk to service cloud and about to move onto our third contractor after the first produced a system that worked poorly and the second seemingly threw out the first's work entirely, producing a similar system that barely works at all. it's okay tho, management thinks we're saving money since we've now shipped the completely broken version (with assurances that the next contractor will fix it) and can stop paying for the zendesk seats! Maximum Leader posted:one great example is a colleague who made a button with a basic condition that simply couldn’t ever evaluate to true. a success message was always shown to the user though and somehow this made it through testing. when I removed the button (because it was confusing and never worked) there was a great amount of pushback from the client because they thought the success message was legitimate. this appears to be in line with sfdc's overall model, where despite basically nothing working, every HTTP response is a 200.
|
|
#
¿
Jan 4, 2019 05:14
|
|
|
staring at modern webapps (including ours) saddens me. i have fond memories of a time long long ago when i was briefly employed as a web dev, and in absence of any actual education or training from my employer on how to do web dev, watching some nice video made by google engineers about how html provided the content, css provided formatting, and javascript provided dynamic modifications of the other two based on user action. it was such a beautiful separation of concerns. now html is a stub document with a bunch of script tags and a splash logo to display while said scripts construct a document with actual content and styling using some nonsense involving 14 layers of functions with single-character names. i have become angry outdated old so much sooner than i thought.
|
|
#
¿
Jan 12, 2019 10:03
|
|
|
MrMoo posted:Turns out they have terrible coders who were drunk when reading the docs or something, I think someone just copy & pasted terrible code from the Internets. doesnt this just describe regular coders
|
|
#
¿
Jan 20, 2019 23:38
|
|
|
Soricidus posted:lua status: ugh why do metatables let you override equality but not hashing, so it is literally impossible to create a type that is suitable for use as a table index constant lua status: dear god what were the designers thinking
|
|
#
¿
Feb 6, 2019 05:14
|
|
|
my new boss is infected with 'excel is the solution to all business software problems' sickness, except google sheets, so worse. i had never seen someone attempt to do web UI mockups in a spreadsheet before but whelp
|
|
#
¿
Feb 15, 2019 16:58
|
|
|
VikingofRock posted:I might have told this story here before, but my undergrad nuclear physics professor wrote a nuclear reactor simulation in excel. The core of the spreadsheet was a 2D overhead view of the reactor, with each cell representing either a control rod and its insertion percentage, or water and its neutron flux. You could change the values for the control rods and make the reactor run hotter if you wanted, and presumably if you pulled out the control rods entirely you could watch Excel itself meltdown as the neutron flux shot up exponentially. this actually seems like a decent use of excel in that it's just a bunch of numbers that change based on other numbers, where you want to see all (or at least most) of the numbers at once. im much less fond of things where people have created excel sheets with multi-paragraph text in every field with no calculations being performed whatsoever, doubly so when this is duplicating information found elsewhere
|
|
#
¿
Feb 15, 2019 22:44
|
|
|
my current employer has a head of engineering that believes engineers should write documentation, as "they are the ones that understand the code". this has * people good at writing code aren't necessarily, and often are not, particularly good with writing clear, readable prose. granted, half our engineering team are non-native english speakers and can't be blamed for this, but the native english speakers aren't much better. * the original designers of a user interface/workflow tend to bake a lot of assumptions about how that interface should be used and what knowledge its users will already have into their documentation. this makes the documentation bad because those assumptions rarely line up with reality. * any role whose primary responsibility is not documentation will focus on their primary responsibilities first. documentation is written in a rush before a release, exacerbating the above two issues. i pray for us to hire a technical writer
|
|
#
¿
Feb 17, 2019 03:31
|
|
|
Soricidus posted:tfw you ask someone to submit a pr and they email you a zip file of their changes instead same but people who are ostensibly sysadmins who send screenshots of text-only terminal contents. sure do love manually transcribing poo poo elsewhere and only having as many logs as fit in whatever window size goofus prefers for their terms
|
|
#
¿
Mar 14, 2019 04:54
|
|
|
Powerful Two-Hander posted:why is jira so slow anyway, we keep getting told we have to migrate instances because "the current one is overloaded" but that implies it's due to the number of actual jiras which is ridiculous unless the underlying data storage/model is complete garbage ill forgive jira's slowness for its having a sane UI. im now stuck doing poo poo in parts of salesforce, whose UX team is somehow incapable of understanding things like "maybe 3 levels of nested tabs isn't a great idea" and "perhaps you shouldn't use a modal for loving EVERYTHING". it's not particularly fast either to boot.
|
|
#
¿
Mar 15, 2019 02:04
|
|
|
my stepdads beer posted:ya i like docker. the worst part of it is trying to run your own k8s and fighting with the weird container networking. they should have just used ipv6 instead of NAT then you can trade bullshit nat weirdness for bullshit apps not handling ipv6 correctly behavior!
|
|
#
¿
Mar 15, 2019 05:52
|
|
|
Beamed posted:who the hell's high school had programming courses we got BASIC and vb6. quality stuff.
|
|
#
¿
Apr 7, 2019 07:32
|
|
|
gonadic io posted:i'm having increasing amount of issues with discord's formatting especially on mobile. how dumb is it for my bot to dynamically generate a png or svg instead of responding in discord's half-assed markup? what the gently caress are you doing in discord
|
|
#
¿
Apr 16, 2019 02:58
|
|
|
gonadic io posted:the junior we hired today astounded me. i left him with: the structures seem sound - get it compiling and then we can start to test it plz stop insulting the very good and enlightened way that lua works by default. lol if you're shameful languages doesn't automatically yield null when accessing an undefined variable
|
|
#
¿
Apr 18, 2019 04:27
|
|
|
HoboMan posted:tps: i be all like "GET /" and the client app says "cant GET /" and i just have to sit here and imagine why i can't GET / because my browser's certainly not gonna tell me, and there's no precious logging to help me out either. https://developer.mozilla.org/en-US/docs/Web/HTTP/CORS the quote:Who should read this article? really should be followed by quote:Who actually reads this article? it may well be some other reason but i am astounded by how many people who work with web apps have just somehow never heard of cors and have never looked at dev tools to see the big angry 'hey you should check out cors' warning that pops up when its constraints arent satisfied.
|
|
#
¿
Apr 30, 2019 21:39
|
|
|
pseudorandom posted:Maybe I'm dumb and stupid, but I loving hate CORS so much. As far as I can tell, it is only good for awful legacy apps that accepted requests like `GET /delete-everything`. But, GOD FORBID anyone like me wants to implement a `DELETE` route or use a non-standard header, because now I need to manually whitelist poo poo and endure a preflight `OPTIONS` request before anything can happen. cors does protect against some very real issues but the design and usability could maybe use some work in my case tho cors was "protecting" requests for fonts, which apparently only require cors headers because some font authors were real upset about the possibility of people stealing their copyrighted works and were apparently oblivious to the idea of requests that don't enforce cors and/or rehosting font files elsewhere.
|
|
#
¿
May 2, 2019 05:37
|
|
|
leper khan posted:I found GPL code in a codebase once and had to explain to exec team what that meant and how to remediate. that was fun this describes my current employer's code base, and is doubly fun because it was checked in *by* the exec team founders, amirite
|
|
#
¿
May 7, 2019 05:07
|
|
|
c tp s: our app has a status endpoint that basically serves no purpose, because the only time it should actually return anything other than a 200 OK is if the underlying infrastructure collapses spectacularly and would likely be unable to return a well-formed error response anyway. it reports if it has lost db connectivity in the body, but doesn't consider this a failure condition because it can still generally serve most requests from a local cache. if rbac is enabled, requests to the service that the status endpoint lives on are pre-checked to ensure the requester has the correct permissions to access an endpoint. this check requires a database request. if that database request fails, the pre-check phase will immediately fail the request and return a 500. a bug in our health monitoring system causes it to open LOTS of db connections in some scenarios, enough that typical postgres installations will start rejecting new connections. as one might expect, this causes the failure above, which is real good if you use the status endpoint for kubernetes liveness checks like we suggest. nothing like restarting every pod in your cluster simultaneously for no good reason.
|
|
#
¿
May 11, 2019 02:45
|
|
|
OP definitely needs SOAP. SOAP makes everything better.
|
|
#
¿
May 22, 2019 10:19
|
|
|
ctps: banging my head against the same version gdb loading the same set of debug symbols for identical binaries on one system as another, but the system i care about produces cores with no translation from offsets to functions. why gdb. why.
|
|
#
¿
Jun 12, 2019 07:25
|
|
|
Sagebrush posted:i loathe it so much. writing it feels like building a tower out of playing cards yeah thats why you template the gently caress out of it to make it easier lol gently caress helm.
|
|
#
¿
Jul 18, 2019 02:39
|
|
|
CORS sucks, but I'm consistently amazed by how many people who work on HTTP APIs have seemingly never heard of it and/or have no idea how to diagnose issues with it. The console has a very obvious "CORS needs ACAO to be this, but it's actually this" error and nobody ever bothers to even look half the time.
|
|
#
¿
Jul 30, 2019 04:36
|
|
|
here in lua land, i've discovered that lua allows using tables as indices for other tables. i continue to hate lua.
|
|
#
¿
Aug 2, 2019 02:33
|
|
|
yall are some sick fucks. my life is worse for having programmed in both lua and mumps
|
|
#
¿
Aug 2, 2019 06:25
|
|
|
necrotic posted:im also an idiot who continues to use vim. ive spent too much time getting all my terminal poo poo just how I like it to migrate to other options at this point. my excuse for still using vim is that i spent a long while working with locked-down BSD systems where vi was the only option other than ed. that and my debugging strategy of hot-patching within running containers because lol interpreted languages
|
|
#
¿
Aug 10, 2019 02:02
|
|
|
Corla Plankun posted:grad school didn't teach me anything magical about reading papers, just persistence truly this is the secret to me doing my job management always like "how on earth have you done this magic" and im like "idk shrug i read the relevant bits of the OIDC/IPSec/x509/insert bullshit overcomplicated spec of choice repeatedly until it was clear what was wrong" that said i do not and will never understand x509. x509 is bullshit.
|
|
#
¿
Aug 30, 2019 04:37
|
|
|
Ciaphas posted:What situations would you want to write your own memory allocator for in C? in nginx, at least, GCing memory allocated during a request after the request completes
|
|
#
¿
Sep 15, 2019 21:40
|
|
), all the asserts I was seeing before are long gone; gotta let it run a few hours to be sure
|
gonadic io posted:Strongly disagree. The number of bugs I've had with accidentally mutating shared state are incredible and they're always really annoying to diagnose too something like 50% of our app's state is handled via global (albeit request-scoped) poo poo love 2 search through every use of a particular global name to try and reconstruct all the possible things that could have hosed with it at any given point in the request lifetime
|
|
#
¿
Sep 20, 2019 10:56
|
|
|
love that someone wrote an (admittedly quite useful, when it works) debugging tool that hooks deep into the core of our poo poo with jack poo poo for tests as a side project and hosed off to another department. nobody else on their previous team wants to deal with it. people try to use it and discover interesting quirks that it has developed due to changes elsewhere in the system, like "causes all HTTPS requests to fail" and "causes ALL requests to fail"
|
|
#
¿
Oct 2, 2019 07:34
|
|
|
NihilCredo posted:to defend themselves, the devs hosting said api sent me ten postman screenshots inside a powerpoint presentation i can never figure out if people still doing this poo poo in this day and age are intentionally trolling or just somehow haven't learned that this is loving stupid yet
|
|
#
¿
Oct 19, 2019 01:54
|
|
|
Blinkz0rz posted:avoid dogmatism in software engineering because it's loving dumb sadly, software engineering is a field largely populated by software engineers, so this will never happen
|
|
#
¿
Nov 2, 2019 23:35
|
|
|
inheritance mostly reminds me of reading through the cassandra codebase and trying to figure out which thing in a stack of 4 classes contained the method i cared about i expect maybe ides make this easier but i almost never read java so gently caress if i know
|
|
#
¿
Nov 15, 2019 08:16
|
|
|
Mahatma Goonsay posted:Me: hey vendor how do i get a token to access your api? unironically how we provide access to build artifacts. we don't bother doing anything better since it's trivial to put out pirated redistributions of our poo poo regardless of what we do and the key is just a basic "don't leave this open to the world by default" layer on top of the actual, meaningful threat of suing someone. elcannon posted:This except instead of them giving us a token they gave us an MSSQL connection string (to a server just listening on the open internet) and a bunch of pinvoke signatures for a dll that shipped with the GUI client.  quote:an EHR oh, okay, now it makes sense, carry on
|
|
#
¿
Nov 20, 2019 09:12
|
|
|
i have wasted years staring at salesforce and jira loading spinners
|
|
#
¿
Nov 21, 2019 07:46
|
|
|
today in terrible programming and sysadmin-ing: while we perhaps should not have treated timestamps from random nodes in a distributed system as absolute, i am surprised how many people simply don't run ntpd on their servers
|
|
#
¿
Nov 22, 2019 05:42
|
|
|
|
| # ¿ Dec 10, 2025 14:42 |
|
|
Private Speech posted:Though TBH I've had more issues with people not being able to deal with hand-editing config files. would give my left nut to never have to deal with sysadmins that can't use vi again i swear to god
|
|
#
¿
Nov 23, 2019 03:04
|
|