|
VOTE YES ON 69 posted:counterpoint: last job had like ~10% utilization, pretty much running a single service per VM, each VM scaled on whatever the resource axis it needed most, so you'd have some memory hungry thing sitting on top of 16 idle cores lmao use a smaller ec2 instance size op
|
# ? May 4, 2017 16:35 |
|
|
# ? May 26, 2024 21:55 |
|
VMs make nice fire walls for containing security breaches. I feel a lot better about the attack surface of qemu virtio than I do about the ~1000 system calls and api pseudo-filesystems in the Linux kernel. I'll go through the motions of setting up SELinux and such, but as far as I'm concerned the moment somebody gets local code execution you may as well consider the entire kernel to be compromised.
|
# ? May 4, 2017 16:49 |
|
Sapozhnik posted:VMs make nice fire walls for containing security breaches. we are most definitely not at the point where cgroups, namespaces, et al are a security boundary on the level of xen/kvm. however, that doesn't mean you should just disregard VMs running multitenant containers; you need to be aware of what kind of applications you are running and how they are externally accessed and gauge your risk accordingly.
|
# ? May 4, 2017 17:26 |
|
Shaggar posted:its not a syntactical error and SSMS is purposefully bad at SQL development work. you may be able to configure it as a compiler warning/error in a VS db project . yeah i know it isnt but i just wish ssms was as good as vs at highlighting my fuckups before i spend 30 minutes debugging them
|
# ? May 4, 2017 18:26 |
|
http://langserver.org/ this looks cool but i hate the programming paradigm where every language tool is written in its own language
|
# ? May 4, 2017 18:35 |
|
Powerful Two-Hander posted:yeah i know it isnt but i just wish ssms was as good as vs at highlighting my fuckups before i spend 30 minutes debugging them after talking to someone on the VS team about it they don't want SMSS to be used for anything more than basic maintenance. I had suggested creating a mechanism for handling schema updates and versioning directly between sql server and the source control system and he looked at me like I was crazy that the database might be the source of truth for schema.
|
# ? May 4, 2017 20:21 |
|
did you properly slap him in the mouth, shaggar?
|
# ? May 4, 2017 20:26 |
|
i just wish ssms would remember my unsaved tabs when i close and restart it like i probably dont care about this query enough to commit it to a file system but i don't just want to throw it away
|
# ? May 4, 2017 21:05 |
|
St Evan Echoes posted:i just wish ssms would remember my unsaved tabs when i close and restart it it remembers them for me, maybe it's a setting? i really had no idea you should be developing in vs for the database even, sounds bad
|
# ? May 4, 2017 21:15 |
|
St Evan Echoes posted:i just wish ssms would remember my unsaved tabs when i close and restart it ssms tools pack ads this functionality along w/ history for closed tabs
|
# ? May 4, 2017 21:35 |
|
also database projects in VS work pretty well and dacpacs are super super easy to deploy but i'd still prefer sql server based schema management
|
# ? May 4, 2017 21:36 |
|
good news: I made an Orca/C compiler disk for my GS bad news: the } key on my keyboard doesn't work
|
# ? May 4, 2017 21:38 |
|
lol
|
# ? May 4, 2017 21:39 |
|
Luigi Thirty posted:good news: I made an Orca/C compiler disk for my GS use %> edit: might need to use the trigraph instead ??> hobbesmaster fucked around with this message at 21:50 on May 4, 2017 |
# ? May 4, 2017 21:43 |
|
next coding interview I should use all alternative operators WHERES YOUR IDIOTIC TRIVIA NOW?!?!
|
# ? May 4, 2017 21:45 |
Luigi Thirty posted:good news: I made an Orca/C compiler disk for my GS Use trigraphs, they are there just for that use case!
|
|
# ? May 4, 2017 21:55 |
|
hobbesmaster posted:next coding interview I should use all alternative operators next time I have one and they say to use whatever language you're most comfortable in I should switch the editor to erlang
|
# ? May 4, 2017 21:59 |
|
Shaggar posted:after talking to someone on the VS team about it they don't want SMSS to be used for anything more than basic maintenance. I had suggested creating a mechanism for handling schema updates and versioning directly between sql server and the source control system and he looked at me like I was crazy that the database might be the source of truth for schema. gently caress this is like when i suggested that maybe we should actually version stored procs instead of relying on people to put comments in them and was told 'yeah that would cost money' so it never happened. instead i have to pedantically dig through svn history so i can tell who to blame and more practically, try and ensure that nobody is working on something that might touch the same proc at the same time edit: which works because i know this goddamn system like the back of my hand and if i was left alone for like 1 month i could refactor 50% of into mvc Powerful Two-Hander fucked around with this message at 22:04 on May 4, 2017 |
# ? May 4, 2017 22:01 |
|
Powerful Two-Hander posted:they also seem to be allergic to actually using relational data. from what i can tell what should be an entity model with lots of one to many relationships is frequently just one row in a table with values stuffed into it as strings. like, even storing the xml blobs and parsing at run time would be better than that. This with huge amounts of nulls is something I've seen in multiple Oracle databases. I'm not sure if it's an Oracle consultant thing Pointsman fucked around with this message at 00:06 on May 5, 2017 |
# ? May 5, 2017 00:02 |
|
Pointsman posted:This with huge amounts of nulls is something I've seen in multiple Oracle databases. I'm not sure if it's an Oracle consultant thing years ago when i was first starting out on the path of terrible programming i asked why an app had such a godawful database structure (1 to many in the same table to do a parent-child structure but where there could be thousands of children and no indexes) and was told "well, they were used to working on oracle so cut them some slack" from then on i dissed oracle at every possible opportunity so i guess it was a good education
|
# ? May 5, 2017 00:14 |
|
hifi posted:http://langserver.org/ this looks cool but i hate the programming paradigm where every language tool is written in its own language it's really convenient, because if you want Piss# support in your editor you probably already have a working Piss# environment
|
# ? May 5, 2017 00:21 |
|
we did it
|
# ? May 5, 2017 03:30 |
|
i would prefer a nullable column to a "one to zero or one" relation between two tables
|
# ? May 5, 2017 03:45 |
|
Luigi Thirty posted:good news: I made an Orca/C compiler disk for my GS orca! now there's a name I haven't heard for many many years I kinda want a IIgs and have a source, but I really want to stick to 68K Mac and LispM stuff for retrocomputing though maybe if I could find a proto Ethernet card...
|
# ? May 5, 2017 08:56 |
|
Luigi Thirty posted:we did it oh that white on blue, how I wanted it back before I committed to Mac... how's it feel, Luigi? how's it feel? (also, fix your keyboard, how hard could it be?) oh yeah, do you have the Apple IIgs Bose speakers? I have a former coworker that has his, they look slick and still sound great. eschaton fucked around with this message at 09:02 on May 5, 2017 |
# ? May 5, 2017 08:59 |
|
fleshweasel posted:i would prefer a nullable column to a "one to zero or one" relation between two tables so would i generally but not when its being used to store history of each row by inserting a 'child' row on every update into a 2 million+ row table with like 25 columns and no indexes. just adding a basic index improved performance massively compared to what the users had been left with by the original devs
|
# ? May 5, 2017 09:32 |
|
Luigi Thirty posted:we did it That looked like the worlds weirdest window air conditioner when I first looked at it.
|
# ? May 5, 2017 16:03 |
|
I feel like most people's DB schema falls apart in the face of "oh poo poo, the customer wants edit history"
|
# ? May 5, 2017 16:36 |
|
fleshweasel posted:I feel like most people's DB schema falls apart in the face of "oh poo poo, the customer wants edit history" I implemented event sourcing for an internal project here that needed edit history, and a pleasant side-effect was how much easier it was to handle read-model schema changes. replaying events ftw
|
# ? May 5, 2017 17:52 |
|
public announcement: if you are writing fancy multi-threaded code please check that it still works under process affinity settings. Online, configured, and available core counts can be different
|
# ? May 5, 2017 18:15 |
|
MrMoo posted:public announcement: if you are writing fancy multi-threaded code please use the OTP behaviors that erlang or elixir ship with
|
# ? May 5, 2017 18:51 |
|
MrMoo posted:public announcement: if you are writing fancy multi-threaded code please check that it still works under process affinity settings. lol if you manually specify a number of cores like a peasant
|
# ? May 5, 2017 20:09 |
|
eschaton posted:orca! now there's a name I haven't heard for many many years I don't have a 68k Mac set up for retro junk, just an Amiga and this GS. I even downloaded the GS Toolbox Reference so I can make mad Super Hi-Res Graphics eschaton posted:oh that white on blue, how I wanted it back before I committed to Mac... I swapped it for a different, working keyboard. nope, no Bose speakers
|
# ? May 5, 2017 20:51 |
|
Cocoa Crispies posted:use the OTP behaviors that erlang or elixir ship with this but just shove it down a gcd queue instead
|
# ? May 6, 2017 00:32 |
|
Potassium Problems posted:I implemented event sourcing for an internal project here that needed edit history, and a pleasant side-effect was how much easier it was to handle read-model schema changes. replaying events ftw how does event sourcing play with changing schema over time
|
# ? May 6, 2017 00:39 |
|
fleshweasel posted:I feel like most people's DB schema falls apart in the face of "oh poo poo, the customer wants edit history" create some triggers to log the old row versions to another table
|
# ? May 6, 2017 01:13 |
|
|
# ? May 6, 2017 01:15 |
|
some lengths to go there.
|
# ? May 6, 2017 01:20 |
|
at least they didn't use poop emojis
|
# ? May 6, 2017 01:21 |
|
|
# ? May 26, 2024 21:55 |
lmao
|
|
# ? May 6, 2017 01:36 |