Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
cool av
Mar 2, 2013

Bloody posted:

im hitting a debug assert in my c# app but cannot attach a debugger for some reason is there a way to get the full stacktrace outta the stupid lil abort/retry/quit window

"retry" will ask you to attach a debugger, generally. if you can't attach because it's on a machine w/o a debugger or something, you can create a dump file from task mgr and debug that later.

Adbot
ADBOT LOVES YOU

cool av
Mar 2, 2013

Soricidus posted:

just use csv and awk, your data isnít that big or important

there must be some dataset out there that is both Big and important?


but I can't think of one

cool av
Mar 2, 2013

it lets the compiler check your code for errors that otherwise would be caught at runtime.

it provides built-in documentation for what variables are meant to be that you'd otherwise need to have the discipline to document manually.

cool av
Mar 2, 2013

you know what, I have decided I'm OK with mixed line endings in a single project. hell, in a single file.

join me; set yourself free.

cool av
Mar 2, 2013

Shaggar posted:

this is basically how entity framework works and its the loving worst thing ever.

you mean for joined collections on the orm objects? yeah those are bad and I just stopped using them after a while

cool av
Mar 2, 2013

I remember arguing that we were clearly wrong for storing a time in *local* time as a 'source of truth' in the database; it made calculations harder and isn't even unambiguous, it should be UTC.

But it was a record of when an event happened, entered by a user in local time, and i now realize that storing it that way was more correct. That's my story.

cool av
Mar 2, 2013

prisoner of waffles posted:

everything in this post is correct

none of it speaks to whether/when you should, e.g., partially squash history, make a merge commit vs a fast-forward, etc.

in a good VCS you'd be able to view and search the history however you like, instead of relying on committers to commit the "right way" so that the history-searching method your team prefers works.

cool av
Mar 2, 2013

Krankenstyle posted:

easier to ctrl-f in a single file :shepface:

this is true and good advice if people are using partial classes in c#

cool av
Mar 2, 2013

IMO he's right that pushing straight to master is better for small teams/projects (I would say 3 or 4 people; 6 is probably pushing it).

PRs mostly just waste everyone's time and increase merge conflict risk under those circumstances.

cool av
Mar 2, 2013

of course "permission to push to master" in git actually means "permission to irreparably destroy your entire repo" but that's a git problem not a workflow one

cool av
Mar 2, 2013

kinda like when i bought a sega saturn, i had so many reasons it was great & better than playstation

cool av
Mar 2, 2013

zokie posted:

Donyou really think that when I describe a system that builds, tests, lints, and looks for breaks in GraphQL and OpenAPI schemas, thatís used create small feedback cycles that it there isnít a CI/CD pipeline for the trunk also? But that pipeline has additional tests that take to long to run.

Like holy poo poo, just because you work with a bunch of morons that have PRs open for years that doesnít mean PRs are bad.

The point is to get review from your peers and having automated checks just makes things better. You want to keep the PRs small, no way can you [b]properly[/] review a PR for a branch someone has worked on for a month.

All of these things can be done, oftentimes better and more efficiently, without PRs.

Some people are misunderstanding nbsd's point about hidden dependencies between branches -- you can't catch those by merging master into your branch before running tests, unless you have a queue of branches to be merged in a specific order, which no one does. Your "additional tests that take too long" to run on branches is another point against a PR-based workflow (you're still merging things that you can't deploy).

The typical PR workflow is designed to slow things down and enforce process and discipline. This is often very good when you're working on a large/critical system or with people who aren't disciplined, but saying they're universally the better choice is just cargo culting.

cool av
Mar 2, 2013

I guess I think of them a sometimes-necessary evil, that should be avoided if you can get away with it, rather than something to strive towards.

cool av
Mar 2, 2013

Aramoro posted:

As soon as you say 'If you don't want to you can just commit to master directly' then that becomes the only process because people are vain and arrogant and won't follow a process with even a single more step if they don't have to. Obviously 'they' know thier change is all fine and good, it was the other rear end in a top hat that broke it.

They were going to break everything with their crappy commit either way.

Without PRs you have to fix it immediately. With PRs you don't. There are many situations where the former is better.

cool av
Mar 2, 2013

it was kind of cool working in .NET before it had a package manager / package ecosystem. if you wanted to ship a library, you had to think very carefully about which dependencies to require and most libs were just standalone and you could easily check them in or stick them on a shared drive somewhere.

package-manager-based ecosystems make it so easy that devs get lazy and any library you depend on is likely to bring in a tree of 100 other dependencies.


I guess what I'm saying is, don't release a new language without a robust and feature-ful standard library plz.

cool av
Mar 2, 2013

Ciaphas posted:

We currently have a monorepo containing about a dozen separately-built projects all depending on some common includes. We're transitioning soon from P4 to Git so I've got a window to reorganize this

My problem is how to actually do that; the only tool I know well for the task is git submodule and that particular tool sucks arse

(ed) or if I even want to do so; a lot of the Googleable Internet seems to think monorepos are a good thing for various reasons that only halfway make sense to me

what're you trying to accomplish by reorganizing it. smaller checkouts? preventing accidental circular dependency commits?

cool av
Mar 2, 2013

Doom Mathematic posted:

If the master branch changes while the PR is open - for example, if a different PR is merged first - the CI pass is invalidated and has to be rerun.) If the reviewer requests changes, the PR cannot be merged. If the CI build fails, the PR cannot be merged.

The temptation in this thought experiment here is to make one of the implicit promises of a PR-based workflow -- that master is never broken -- actually real. But it's not practical at anything but the smallest scale.

You're giving up all horizontal scalability in your build process, and the smaller and more frequent your PRs are, the worse it becomes.

cool av
Mar 2, 2013

null was a mistake

cool av
Mar 2, 2013

can I get a job in embedded dev if i beat shenzen I/O?

cool av
Mar 2, 2013

Sapozhnik posted:

of COURSE this was written by some breathtakingly arrogant Apple fuckstain, how am I not surprised

"I would have unilaterally imposed my own opinions onto a very widely used piece of software, drat the consequences, were it not for the fact that Daddy would have been inconvenienced"

Apple delenda est

are you sure he was from apple, it sounded to me like he just didn't want to break compatibility for a major player

but lol breaking compatibility sure as hell wouldn't ever stop apple from anything ever


are c++ modules going to make this stupid problem go away?

cool av
Mar 2, 2013

Gaukler posted:

I love that, as an industry, we decided reproducibility is so hard that we better just schlep around entire filesystems

recently at work we discussed making some methods that operate on a ~30MB state thread-safe without needing to duplicate the state.

it' s like uhhh okay but we've basically decided as an industry that even sharing an operating system with another app is too hard, and we're going to very carefully try to minimize locking on this 30MB state object?

cool av
Mar 2, 2013

Plorkyeran posted:

there's no inherent reason why async/await can't give useful stack traces. i don't know if it doesn't in js because the spec is poo poo and accidentally made it impossible or if v8 just doesn't give a gently caress. probably both.

In chrome dev tools you can't add variables to the watch window or see their value in the console for source-mapped JS code.

You can mouse-over them though.

I don't know how people put up with it. Maybe part of why JS devs are so obsessed with filesystem watchers that auto-rebuild and restart their app.

cool av
Mar 2, 2013

Powerful Two-Hander posted:

terrible programming: looks like the vendor never considered that someone might use a session to download hundreds of thousands of items and didn't bother implementing any temp file cleanup except on session shutdown.

:rip: the available disk space on those servers.

temp file cleanup of some kind was implemented, sounds like a good vendor.

cool av
Mar 2, 2013

Twerk from Home posted:

actually, one single commit merged back to master will result in multiple new tags

each package has its own monotonically increasing version number in tag, and when you update a package that other packages depend on, the entire tree will get new tags

everyone seems happy with our ramshackle monorepo that barely works, though?

does it cause any problems?

cool av
Mar 2, 2013

it's always namespaces. nobody can understand those things. and even if you understand them, save everybody else the trouble please.

cool av
Mar 2, 2013

Shaggar posted:

namespaces are ez and good

im glad your app can automatically parse and process XML from all kinds of different sources and understand the context and common data between them with no extra work thanks to the metadata included in the documents, but somehow mine still can't

cool av
Mar 2, 2013

love to

quote:

XmlNamespaceManager.AddNamespace("woop", "http://some-dudes-awesome-namespace.com")
so i can query the elements out of some dude's awesome xml that he put his namespace in, presumably so he could participate in the amazing ecosystem of standardized namespaced XML elements out there



edit: even for something like XMLSchema-instance... i think an XSL parser could manage to figure out what a <xsl:template> is even if the element name was actually "xsl:template" and this namespace technology never existed

cool av fucked around with this message at 19:32 on Nov 20, 2020

cool av
Mar 2, 2013

DELETE CASCADE posted:

our current solution: just throw the loving namespace away and compare only the local name :shrug:

same

cool av
Mar 2, 2013

Achmed Jones posted:

hot take coming in

here it comes

hope ur ready

ive never had a problem with protobufs. never used them outside of the environment they were made for though

the C++ version library adds in the neighborhood of 50MB to your binary size like lol how hard is it to serialize ints and strings? they have a "lite" version that's only like 10MB.

also i bet most people using it are transferring it over HTTPS

cool av
Mar 2, 2013

jesus WEP posted:

oh wow i am so worried about my xml being human readable because itís not like thereís a million tools that you can paste it into to display it nicely

internet explorer all the way.

please assist my issue here is my input file:

quote:

<?xml version="1.0"?>

-<Catalog>


-<Book id="bk101">

<Author>Garghentini, Davide</Author>

<Title>XML Developer's Guide</Title>

<Genre>Computer</Genre>

<Price>44.95</Price>

<PublishDate>2000-10-01</PublishDate>

<Description>An in-depth look at creating applicationswith XML.</Description>

</Book>


-<Book id="bk102">

<Author>Garcia, Debra</Author>

<Title>Midnight Rain</Title>

<Genre>Fantasy</Genre>

<Price>5.95</Price>

<PublishDate>2000-12-16</PublishDate>

<Description>A former architect battles corporate zombies,an evil sorceress, and her own childhood to become queenof the world.</Description>

</Book>

</Catalog>

cool av
Mar 2, 2013

the only thing that sounds kinda off is the back-end pushing a new token to the client? (is there always just one client? will they definitely be logged in when the manager approves it?) and I guess if they need 2 temp. roles they have to request them both in a single transaction.

cool av
Mar 2, 2013

the solution to the problems caused by a queue is, of course, another queue

cool av
Mar 2, 2013

actually every programming language is bad

cool av
Mar 2, 2013

love debugging in containers & especially k8s

cool av
Mar 2, 2013

kinda cool that the solution to dependency hell was to store every permutation of all versions of every dependency ever used by every app forever as container images

cool av
Mar 2, 2013

hm i recently implemented serilog in a project and it was pretty painless but as I was doing so I noted the only useful thing it has is a standard configuration for enabling different log levels for differently named logs.

(wow it can json serialize the arguments to my log function what a world lol)

is it going to bite me later on?

cool av
Mar 2, 2013

going on 3 weeks now trying to fix a client on a 12-hour timezone difference that has 2 problems: (1) they are using an outdated version of a "plugin" DLL that they wrote and provide themselves and (2) they have a missing root cert on some of their machines that breaks the TLS connection between that DLL and their web service

the combination of 2 problems and one of them only affecting some machines is just too much and now i have to get up early to meet with them again ugh

cool av
Mar 2, 2013

allowing both single- and double-quotes for strings in a language is wrong

cool av
Mar 2, 2013

Shaggar posted:

with our app at some point in the next year or so we're gonna start overhauling it and im considering converting it to a webapp with fully server side application state. the idea being when dumbass users close the browser they can just go back to the webapp and be immediately back where they were.

i dont know if the current team is right for it tho.

you gonna store the state in sql server? ive done this in an app layer and it's ok but kind of annoying to do updates

Adbot
ADBOT LOVES YOU

cool av
Mar 2, 2013

it's so sad that git became the predominant source control tool, we should've waited for a 2nd generation of dvcs tools to come out before settling down.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply