|
e: wtf is going on with the forums. before I posted this post, loading this page did this on the topic list it also said last post was from mononcqc but I don't see a post from him Flat Daddy fucked around with this message at 18:59 on Oct 6, 2017 |
# ? Oct 6, 2017 18:56 |
|
|
# ? May 25, 2024 11:06 |
|
zdrs deleted posts are messing up the forums
|
# ? Oct 6, 2017 18:57 |
|
Shaggar posted:server side rendering is better than javascript any day of the year.
|
# ? Oct 6, 2017 18:58 |
|
or once webassembly comes out, client side rendering in c#
|
# ? Oct 6, 2017 18:59 |
|
Flat Daddy posted:on the topic list it also said last post was from mononcqc but I don't see a post from him yeah I posted but it never appeared. maybe it's for the best though
|
# ? Oct 6, 2017 19:05 |
|
server side rendering gives you a nice simple mental model because your DOM contents are determined purely by the contents of your viewbag or w/e and nothing else. i.e. single source of truth. to debug why my HTML looks like this, I look at my http call going to the server, look at the viewbag values coming back, and at my template. jQuery and angular hosed with this by making your DOM contents determined by a combination of viewbag from server, users manipulating the DOM via interaction, your lovely JS code that probably mutates DOM manually and uses global variables, a mess of event handlers, etc. no single source of truth. I'd rather kill myself than debug this. redux is basically the same mental model as server-side rendering, except replace viewbag with redux store and replace http requests with redux actions. to see why my DOM looks a certain way, I just trace what actions were dispatched by my components and see how those actions modified the store. since my DOM is literally a (pure) function of the contents of the store, I can figure out from there exactly what part of my component tree is loving up. not to mention this owns bc every part of this is deterministic so I can literally rewind user actions and play them back and see the store contents at any given time in a session. and when I do notice a bug in, say, a reducer, I can use my time travel slider to put me exactly where the bug occurred, then copy my previous state, last action, and current buggy state as JS objects straight into a unit test Flat Daddy fucked around with this message at 19:21 on Oct 6, 2017 |
# ? Oct 6, 2017 19:19 |
|
I'm still very tepid on Redux. I use it, but only for things that need to interact with Sagas. Pretty much I use it to hold mapped responses to GET requests, but then I'll have view components that sort and paginate and display those data sets, and that side of things is handled in local state within higher-order components. Didn't even know "Flux Standard Actions" was a thing either, will look into it. I tried coming up with my own solution to this sort of problem but because I'm dumb and bad I abandoned it.
|
# ? Oct 6, 2017 19:25 |
|
Sapozhnik posted:I'm still very tepid on Redux. I use it, but only for things that need to interact with Sagas. Pretty much I use it to hold mapped responses to GET requests, but then I'll have view components that sort and paginate and display those data sets, and that side of things is handled in local state within higher-order components. check this out too https://github.com/agraboso/redux-api-middleware it's another way to take care of started/done/failed boilerplate. one thing I really want is a swagger codegen that creates RSAA-returning functions to use with this e: RSAA just is another standard action format like FSA but it describes a full http request lifecycle Flat Daddy fucked around with this message at 19:37 on Oct 6, 2017 |
# ? Oct 6, 2017 19:30 |
|
quiggy posted:boy c++ it sure would be nice if you had literally any reflection at all lol you can't even get a printable string with the name of a type
|
# ? Oct 6, 2017 19:51 |
|
hackbunny posted:lol you can't even get a printable string with the name of a type i have a suite of classes that encapsulate three primitive types and one stl type: bool, double, int, and std::string. i had to create a Type enum with values BOOL, DOUBLE, INT, and STRING to do some of what i want
|
# ? Oct 6, 2017 19:55 |
|
Flat Daddy posted:server side rendering gives you a nice simple mental model because your DOM contents are determined purely by the contents of your viewbag or w/e and nothing else. i.e. single source of truth. to debug why my HTML looks like this, I look at my http call going to the server, look at the viewbag values coming back, and at my template. what tool are you using to do this? i got the same tech stack, but i don't know of any tooling for it so debugging and building currently sucks rear end for me.
|
# ? Oct 6, 2017 20:11 |
|
tef posted:pay rent u wanna move to MN and work on crusty old webforms? e: we also got a react thing kind of bolted on to the side of it
|
# ? Oct 6, 2017 20:14 |
|
HoboMan posted:what tool are you using to do this? i got the same tech stack, but i don't know of any tooling for it so debugging and building currently sucks rear end for me. Redux has a browser extension and debug add-on for in app use that let you rewind and all that poo poo.
|
# ? Oct 6, 2017 20:18 |
|
necrotic posted:Redux has a browser extension and debug add-on for in app use that let you rewind and all that poo poo. in chrome? i tried this and it reliably crashed chrome
|
# ? Oct 6, 2017 20:19 |
|
HoboMan posted:what tool are you using to do this? i got the same tech stack, but i don't know of any tooling for it so debugging and building currently sucks rear end for me. https://github.com/gaearon/redux-devtools + the browser extension for it. and the Facebook browser extension for react dev tools.
|
# ? Oct 6, 2017 20:20 |
|
HoboMan posted:in chrome? i tried this and it reliably crashed chrome You can use the entire thing in-app with no browser extension. You gotta configure redux to work with the dev tool in either case. I've not had crash issues.
|
# ? Oct 6, 2017 20:30 |
|
we may be hiring soon. I'm psyched. gonna get a qa manager
|
# ? Oct 6, 2017 20:32 |
|
Flat Daddy posted:e: wtf is going on with the forums. before I posted this post, loading this page did this
|
# ? Oct 6, 2017 20:34 |
|
when ZDRs posts got hosed it affected the page count of the thread but the last viewed calculator is either cached out of date or doesn't take total post count into account so we get an always wrong unread post count and in the case of the Microsoft thread the missing unread post count is enough to put the last unread page link onto a non-existent page. or something like that
|
# ? Oct 6, 2017 20:36 |
|
Shaggar posted:we may be hiring soon. I'm psyched. gonna get a qa manager i want a real qa team so bad currently it's a bunch of BAs that are supposed to do it in their down time. guess how well that works?
|
# ? Oct 6, 2017 20:37 |
Shaggar posted:when ZDRs posts got hosed it affected the page count of the thread but the last viewed calculator is either cached out of date or doesn't take total post count into account so we get an always wrong unread post count and in the case of the Microsoft thread the missing unread post count is enough to put the last unread page link onto a non-existent page. weirdly enough qcs tech thread works fine and zdr is ought to have dozens of pages of his posts alone there
|
|
# ? Oct 6, 2017 20:38 |
|
yeah must be something else in there too
|
# ? Oct 6, 2017 20:38 |
|
HoboMan posted:i want a real qa team so bad so you basically have a full time QA team, is what you're saying
|
# ? Oct 6, 2017 20:45 |
|
Flat Daddy posted:https://github.com/gaearon/redux-devtools + the browser extension for it. and the Facebook browser extension for react dev tools. tyvm question 2: how do i get my poo poo to build properly without me having to nuke my project from orbit and then reinstall everything every time?
|
# ? Oct 6, 2017 20:55 |
|
cinci zoo sniper posted:weirdly enough qcs tech thread works fine and zdr is ought to have dozens of pages of his posts alone there it only happens when there's some specific difference between the actual postcount and the bugged +zdr postcount. probably when it's enough that the inflated +zdr count would indicate a new page but there's not actually a new page.
|
# ? Oct 6, 2017 20:59 |
|
HoboMan posted:tyvm What do you mean? How does poo poo break if you haven't changed anything? Check out create-react-app if you're frustrated and want a simple setup that just works. Also use yarn or NPM >=5 for package lock and deterministic dependencies because npm is trash fire.
|
# ? Oct 6, 2017 21:10 |
|
necrotic posted:What do you mean? How does poo poo break if you haven't changed anything? i have no loving clue
|
# ? Oct 6, 2017 21:13 |
|
Shaggar posted:server side rendering is better than javascript any day of the year. i wish u would postback to wherever u came from
|
# ? Oct 6, 2017 21:19 |
|
HoboMan posted:i have no loving clue I meant how as in what breaks not how did it break suddenly. Build error? Runtime error? Something else entirely? Poor wording on my part
|
# ? Oct 6, 2017 21:24 |
|
cis autodrag posted:it only happens when there's some specific difference between the actual postcount and the bugged +zdr postcount. probably when it's enough that the inflated +zdr count would indicate a new page but there's not actually a new page. The Metal Gear Solid thread in Games is four posts into the next page and it's still broken.
|
# ? Oct 6, 2017 21:29 |
|
Elysiume posted:this thread has a permanent off-by-one error and if that one bridges a page, the buttons for the next page will be enabled even though the page doesn't exist yet that had been the case it's significantly worse now, with a permanent link on my control panel
|
# ? Oct 6, 2017 21:31 |
I wonder if the original extra post was a post that was deleted in the far-past and that's why the thread has been slightly broken And now it's really broken Trump and boston threads just closed and made new ones which is pretty hosed up
|
|
# ? Oct 6, 2017 21:33 |
|
necrotic posted:I meant how as in what breaks not how did it break suddenly. Build error? Runtime error? Something else entirely? oh. it builds ok, but throws a million runtimes about how poo poo is undefined or does not exist
|
# ? Oct 6, 2017 21:33 |
|
CPColin posted:The Metal Gear Solid thread in Games is four posts into the next page and it's still broken. the real worst part is people are going to be discovering this for the first time over the course of the next month, like with the broken chrome anchor links
|
# ? Oct 6, 2017 21:36 |
|
post some fuckin error messages dude
|
# ? Oct 6, 2017 21:36 |
|
https://forums.somethingawful.com/showthread.php?threadid=3648105&userid=72615&perpage=40&pagenumber=8 the ghost posts are still there
|
# ? Oct 6, 2017 21:42 |
Sapozhnik posted:https://forums.somethingawful.com/showthread.php?threadid=3648105&userid=72615&perpage=40&pagenumber=8 btw it died again
|
|
# ? Oct 6, 2017 21:44 |
|
that it did, keep screaming into the void i guess (seriously lol this poo poo is never going to get fixed just euthanize the threads that have database cancer and start again)
|
# ? Oct 6, 2017 21:45 |
|
The Bloodborne thread is up to 19 ghost posts, look quickly before the forums start eating them again.
Van Kraken fucked around with this message at 21:50 on Oct 6, 2017 |
# ? Oct 6, 2017 21:46 |
|
|
# ? May 25, 2024 11:06 |
Van Kraken posted:The Bloodborne thread is up to 19 ghost posts, look quickly before the forums start eating them again. 36 in microsoft thread
|
|
# ? Oct 6, 2017 21:58 |