awful programming: feature flags and suffering

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > YOSPOS > awful programming: feature flags and suffering

«‹›1114 »

Sapozhnik: Jan 2, 2005; Nap Ghost

boy i had a really lame and weird dream last night where the yospos programming threads had been taken over by unironic c++ likers, crazy poo poo huh

# ? Jan 16, 2019 18:59

Adbot: ADBOT LOVES YOU

# ? Apr 19, 2024 16:30

Powerful Two-Hander: Mar 10, 2004; Mods please change my name to "Tooter Skeleton" TIA.

i have just spent idk an hour trying to work out why our pre-production environment has no log entries in it after we adjusted some of our loggers to stop writing out garbage and oh boy is the answer a delight!

nobody ever runs andy loving maintenance so our log db had so much stuff in it that you couldn't query it, so seeing as its all test logs i truncated the log table, so far so good. except that now nothing else is getting written to it even when there's an error, that's weird. and wait when i do a manual insert there's nothing there either, huh???

turns out there's a trigger on the loving log table that deletes the oldest record following any insert, so if you have no logs in it at all it will insert, hit the trigger and delete the log entry again because your record is now the oldest.

I highly recommend this as a piece of troll code because of course, nobody ever thinks to look for loving triggers

edit: OH MY loving GOD THIS WAS ADDED BY SOMEONE *AS* MAINTENANCE WHAT THE FUCCKKKKK

VVV booooo_this_man.gif

Powerful Two-Hander fucked around with this message at 19:02 on Jan 16, 2019

# ? Jan 16, 2019 19:00

DONT THREAD ON ME: Oct 1, 2002; by Nyc_Tattoo; Floss Finder

Powerful Two-Hander posted:

I highly recommend this as a piece of troll code because of course, nobody ever thinks to look for loving triggers

well did you ignore the trigger warning?

# ? Jan 16, 2019 19:02

Shaggar: Apr 26, 2006

Powerful Two-Hander posted:

i have just spent idk an hour trying to work out why our pre-production environment has no log entries in it after we adjusted some of our loggers to stop writing out garbage and oh boy is the answer a delight!

nobody ever runs andy loving maintenance so our log db had so much stuff in it that you couldn't query it, so seeing as its all test logs i truncated the log table, so far so good. except that now nothing else is getting written to it even when there's an error, that's weird. and wait when i do a manual insert there's nothing there either, huh???

turns out there's a trigger on the loving log table that deletes the oldest record following any insert, so if you have no logs in it at all it will insert, hit the trigger and delete the log entry again because your record is now the oldest.

I highly recommend this as a piece of troll code because of course, nobody ever thinks to look for loving triggers

edit: OH MY loving GOD THIS WAS ADDED BY SOMEONE *AS* MAINTENANCE WHAT THE FUCCKKKKK

VVV booooo_this_man.gif

yeah don't ever use triggers cause they're a pain in the rear end for discovery. you should run a query of your schema for any existing triggers and get rid of all of them if you can.

# ? Jan 16, 2019 19:04

Powerful Two-Hander: Mar 10, 2004; Mods please change my name to "Tooter Skeleton" TIA.

Shaggar posted:

yeah don't ever use triggers cause they're a pain in the rear end for discovery. you should run a query of your schema for any existing triggers and get rid of all of them if you can.

agreed. there is precisely 1 valid use case for them imho which is for journalling data for audit history if you have to keep it, anything else is an abomination.

i have changed the trigger to only do anything if there are more than 100,000 logs in the table and written a sarcastic comment in it even though i know pefectly well nobody else will ever read it.

# ? Jan 16, 2019 19:08

redleader: Aug 18, 2005; Engage according to operational parameters

NihilCredo posted:

forget "every browser", just within msft it's gonna take a few more years for netcore to become a full replacement for netfx, and probably decades for it to become a full replacement for mono

you'd think that ms buying xamarin would have put the pedal to the metal on bringing netcore to mobile and deprecating mono, instead mono is apparently getting indipendently developed even further with stuff like aot compilation which would have been a real nice addition to netcore

yeah, i'm really wondering what the long-term strategy for netcore is. obviously part of it is "don't spook enterprises by deprecating netfx", but beyond that, ???
keeping both mono and netcore around seems like a big duplication of effort, given that they both seem so similar on the surface

# ? Jan 16, 2019 19:18

redleader: Aug 18, 2005; Engage according to operational parameters

Sapozhnik posted:

boy i had a really lame and weird dream last night where the yospos programming threads had been taken over by unironic c++ likers, crazy poo poo huh

please do not langshame in this thread. the heart wants what the heart wants

# ? Jan 16, 2019 19:19

DONT THREAD ON ME: Oct 1, 2002; by Nyc_Tattoo; Floss Finder

wow, did ya'll know starcraft 2 exposes a client protobuf protocol and c++ library that lets you program bots using all of the information and abilities the AI in starcraft 2 has??

so anyhow that's what i'm doing for a while.

# ? Jan 16, 2019 19:52

DONT THREAD ON ME: Oct 1, 2002; by Nyc_Tattoo; Floss Finder

i'm going to implement sorting algorithms with sc2 units

# ? Jan 16, 2019 19:53

CRIP EATIN BREAD: Jun 24, 2002; Hey stop worrying bout my acting bitch, and worry about your WACK ass music. In the mean time... Eat a hot bowl of Dicks! Ice T; Soiled Meat

please make gifs

# ? Jan 16, 2019 19:54

CRIP EATIN BREAD: Jun 24, 2002; Hey stop worrying bout my acting bitch, and worry about your WACK ass music. In the mean time... Eat a hot bowl of Dicks! Ice T; Soiled Meat

also make an AI that sends groups of enemies in a formation spelling out "YOSPOS", tia

# ? Jan 16, 2019 19:55

DONT THREAD ON ME: Oct 1, 2002; by Nyc_Tattoo; Floss Finder

CRIP EATIN BREAD posted:

also make an AI that sends groups of enemies in a formation spelling out "YOSPOS", tia

honestly, already planning on it

# ? Jan 16, 2019 20:11

Cold on a Cob: Feb 6, 2006; i've seen so much, i'm going blind
and i'm brain dead virtually; College Slice

DONT THREAD ON ME posted:

i'm going to implement sorting algorithms with sc2 units

nice

# ? Jan 16, 2019 20:15

NihilCredo: Jun 6, 2011; iram omni possibili modo preme:
plus una illa te diffamabit, quam multæ virtutes commendabunt

CRIP EATIN BREAD posted:

also make an AI that sends groups of enemies in a formation spelling out "YOSPOS", tia

*spinning newspaper headline*

TERRORISTS EVADE ECHELONSA USING SECRET "LOST TEMPLE" CHANNEL

# ? Jan 16, 2019 20:23

the talent deficit: Dec 20, 2003; self-deprecation is a very british trait, and problems can arise when the british attempt to do so with a foreign culture

Sapozhnik posted:

boy i had a really lame and weird dream last night where the yospos programming threads had been taken over by unironic c++ likers, crazy poo poo huh

I miss the old days where monad lovers ruled the threads

# ? Jan 16, 2019 22:05

The MUMPSorceress: Jan 6, 2012; ^SHTPSTS; Gary’s Answer

DONT THREAD ON ME posted:

wow, did ya'll know starcraft 2 exposes a client protobuf protocol and c++ library that lets you program bots using all of the information and abilities the AI in starcraft 2 has??

so anyhow that's what i'm doing for a while.

I yearn for this much free time.

# ? Jan 16, 2019 22:13

Plorkyeran: Mar 22, 2007; To Escape The Shackles Of The Old Forums, We Must Reject The Tribal Negativity He Endorsed

step one is quit your job and live off savings for a while

# ? Jan 16, 2019 22:28

VikingofRock: Aug 24, 2008

Sapozhnik posted:

boy i had a really lame and weird dream last night where the yospos programming threads had been taken over by unironic c++ likers, crazy poo poo huh

the talent deficit posted:

I miss the old days where monad lovers ruled the threads

why not both?

# ? Jan 16, 2019 22:30

gonadic io: Feb 16, 2011; >>=

the talent deficit posted:

I miss the old days where monad lovers ruled the threads

Anybody want to know what a monad is? I feel like I've done one post per person in the thread at this point

# ? Jan 16, 2019 22:33

CRIP EATIN BREAD: Jun 24, 2002; Hey stop worrying bout my acting bitch, and worry about your WACK ass music. In the mean time... Eat a hot bowl of Dicks! Ice T; Soiled Meat

monads are just monoids in the category of endofunctors

# ? Jan 16, 2019 22:34

gonadic io: Feb 16, 2011; >>=

I even got a new monad related av to spice things up a bit

Real talk though I stupidly decided to do a timed job interview programming thing in Haskell after having not touched it for like 4 years and boy was that a mistake. They did not call back.

# ? Jan 16, 2019 22:36

hackbunny: Jul 22, 2007; I haven't been on SA for years but the person who gave me my previous av as a joke felt guilty for doing so and decided to get me a non-shitty av

Illusive gently caress Man posted:

how can/should a programmer add this kind of detail in the exception world of that paper? A try-catch that just appends detail and re-throws sounds unwieldy to me, but that could be because I've been living in an exceptionless world for so long.

sorry:

Sutter-sensei posted:

It is not a goal to enable distantly handled errors to contain arbitrary programmatically usable information. Distantly handled error details primarily need to be human usable (e.g., debugging and trace logging), and a .what() string is sufficient.

with classic exceptions, you can easily wrap at the throw site, and keep the same catch sites, and gradually migrate the catch sites to extract the extra information. see how Boost.Exception does it: https://www.boost.org/doc/libs/1_69_0/libs/exception/doc/boost-exception.html

with value exceptions as currently proposed, you can wrap just as easily, but unwrapping becomes mandatory, because they don't have multiple inheritance: every error has one piece of type-erased data and a category, no more no less. no such thing as multiple views into an object, like you can do with classes

hackbunny fucked around with this message at 22:45 on Jan 16, 2019

# ? Jan 16, 2019 22:39

hackbunny: Jul 22, 2007; I haven't been on SA for years but the person who gave me my previous av as a joke felt guilty for doing so and decided to get me a non-shitty av

honestly I'm surprised that the proposal doesn't provide for user-defined error types, and a little disappointed because I was never a huge fan of funneling all error handling through errc comparisons/conversions

e: my bad, �4.6.5 says it may be possible in the future, but herb sounds skeptical of its usefulness. weird because I can think of a lot of things I'd like to have that his proposed std::error can't do

hackbunny fucked around with this message at 23:48 on Jan 16, 2019

# ? Jan 16, 2019 23:06

pokeyman: Nov 26, 2006; That elephant ate my entire platoon.

Powerful Two-Hander posted:

nobody ever runs andy loving maintenance

is this like norton disk doctor?

I�m andy loving maintenance! howdy howdy howdy

# ? Jan 16, 2019 23:32

qntm: Jun 17, 2009

what the hell is a monad

# ? Jan 17, 2019 00:46

Powerful Two-Hander: Mar 10, 2004; Mods please change my name to "Tooter Skeleton" TIA.

pokeyman posted:

is this like norton disk doctor?

I�m andy loving maintenance! howdy howdy howdy

he's like the blue collar dr watson

i spotted that typo and just left it

# ? Jan 17, 2019 00:58

pokeyman: Nov 26, 2006; That elephant ate my entire platoon.

Powerful Two-Hander posted:

i spotted that typo and just left it

ty for your service

# ? Jan 17, 2019 01:00

floatman: Mar 17, 2009

My favourite trigger is one that triggers on update/create of rows for table A. What it does is it creates a new record in another table, which has two columns: classname and args. Args is a text data type but it stores XML.
So what happens is there is a Cron job that would poll that table, use the language equivalent of "eval" to create a new "classname", deserialise the XML "args" and feed it into the class, then execute the class to do stuff.
So when you look at the project there are all these little class files that can't be connected with "find usages"

# ? Jan 17, 2019 02:06

Ellie Crabcakes: Feb 1, 2008; Stop emailing my boyfriend Gay Crungus

redleader posted:

please do not langshame in this thread. the heart wants what the heart wants

While Javascript exists, shaming is not always wrong.

qntm posted:

what's a monad

# ? Jan 17, 2019 03:56

CRIP EATIN BREAD: Jun 24, 2002; Hey stop worrying bout my acting bitch, and worry about your WACK ass music. In the mean time... Eat a hot bowl of Dicks! Ice T; Soiled Meat

floatman posted:

XML

could be worse... could be json

# ? Jan 17, 2019 06:52

floatman: Mar 17, 2009

I left out what the XML usually contains.

The XML usually contains one encoded string which decodes to json. Sometimes, they store CSV strings in the json too instead of creating a json array.

# ? Jan 17, 2019 07:57

Soricidus: Oct 21, 2010; freedom-hating statist shill

floatman posted:

I left out what the XML usually contains.

The XML usually contains one encoded string which decodes to json. Sometimes, they store CSV strings in the json too instead of creating a json array.

oh, I didn�t realise I had colleagues in this thread

# ? Jan 17, 2019 09:40

Powerful Two-Hander: Mar 10, 2004; Mods please change my name to "Tooter Skeleton" TIA.

floatman posted:

My favourite trigger is one that triggers on update/create of rows for table A. What it does is it creates a new record in another table, which has two columns: classname and args. Args is a text data type but it stores XML.
So what happens is there is a Cron job that would poll that table, use the language equivalent of "eval" to create a new "classname", deserialise the XML "args" and feed it into the class, then execute the class to do stuff.
So when you look at the project there are all these little class files that can't be connected with "find usages"

if you use json.net to de/serialise objects that implement an interface something like this happens, because the deserialiser doesn't know what implementation its dealing with, it looks for a type property that was added by the serialiser (if you called it with the appropriate flags) and uses it to instance the proper object type then populate it with the rest of the data

its actually really useful but I wouldn't want to do it in the db and involve triggers in it :barf:

# ? Jan 17, 2019 10:33

NihilCredo: Jun 6, 2011; iram omni possibili modo preme:
plus una illa te diffamabit, quam multæ virtutes commendabunt

is there an idiomatic name for a two-way lookup data structure (i.e. a bijective function)?

and/or a more clever design than a wrapper around two hashmaps, maybe

# ? Jan 17, 2019 11:56

Cold on a Cob: Feb 6, 2006; i've seen so much, i'm going blind
and i'm brain dead virtually; College Slice

floatman posted:

My favourite trigger is one that triggers on update/create of rows for table A. What it does is it creates a new record in another table, which has two columns: classname and args. Args is a text data type but it stores XML.
So what happens is there is a Cron job that would poll that table, use the language equivalent of "eval" to create a new "classname", deserialise the XML "args" and feed it into the class, then execute the class to do stuff.
So when you look at the project there are all these little class files that can't be connected with "find usages"

yeah but you can deploy code changes so easily now

lmao

# ? Jan 17, 2019 13:16

Cold on a Cob: Feb 6, 2006; i've seen so much, i'm going blind
and i'm brain dead virtually; College Slice

years and years ago i had a team lead bring in a business rule engine that worked like that via reflection, his justification was "it will let us deploy code changes into prod by going through the data update approval process instead of the release process, which is easier (!) and the business analysts will be able to change the rules directly (!!!!)"

it was slow as poo poo, hard to code (required a third party editor), and of course no business analysts ever touched it

as soon as he left i accepted the team lead position on the condition our first project was to replace that with native c# code written and maintained by developers

# ? Jan 17, 2019 13:19

Jabor: Jul 16, 2010; #1 Loser at SpaceChem

NihilCredo posted:

is there an idiomatic name for a two-way lookup data structure (i.e. a bijective function)?

and/or a more clever design than a wrapper around two hashmaps, maybe

it turns out that just having a map you can query in the forward direction, and another map you can query in the reverse direction, is actually a pretty convenient api for this sort of data structure

https://github.com/google/guava/wiki/NewCollectionTypesExplained#bimap

# ? Jan 17, 2019 13:39

NihilCredo: Jun 6, 2011; iram omni possibili modo preme:
plus una illa te diffamabit, quam multæ virtutes commendabunt

"bimap" is a pretty good name, thanks

# ? Jan 17, 2019 14:58

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

oh god converting bytestring py2 to unicode py3 just kill me

# ? Jan 17, 2019 15:14

Adbot: ADBOT LOVES YOU

# ? Apr 19, 2024 16:30

Carthag Tuek: Oct 15, 2005; Tider skal komme,
tider skal henrulle,
sl�gt skal f�lge sl�gters gang

also wtf is this poo poo, could it be any more impenetrable

code:

for t in range(2, len(word)):
    temp = [(x[0] + (j,), (x[1] * self.tran[x[0][-1]][j] * self.emis[j][word[t]]))
        for j in self.states for x in paths]

# ? Jan 17, 2019 15:16

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > YOSPOS > awful programming: feature flags and suffering

«‹›1114 »