|
when you get there push them to rewrite in nodejs
|
# ? Jul 2, 2015 15:43 |
|
|
# ? May 25, 2024 21:39 |
|
bobbilljim posted:should i work with php y/n probably not.
|
# ? Jul 2, 2015 15:57 |
|
Bloody posted:always be exiting
|
# ? Jul 2, 2015 15:59 |
|
who cares what language you get paid to write bad code in, just stack that paper
|
# ? Jul 2, 2015 16:00 |
|
if you're hellbent on using php, at least do yourself a favour and use an ide. there's no compiler to catch anything early and debugging it at runtime is a huge pain.
|
# ? Jul 2, 2015 16:07 |
|
i wrote a thing in rust and haskell for work today
|
# ? Jul 2, 2015 16:13 |
|
fart simpson posted:i wrote a thing in rust and haskell for work today i'd love to hear more about this
|
# ? Jul 2, 2015 16:28 |
|
I started my professional work on a lovely old 10 year old PHP code base, but was paired with good engineers and some people from way back when the code base started. It proved to be an incredible experience in terms of knowing what tiny things grow up to be terrible down the road, things of what not to do, and how to organize things better. I have to think it would have taken a few more years to gain that experience when working in a nicer code base.
|
# ? Jul 2, 2015 16:29 |
|
MononcQc posted:things of what not to do, and how to organize things better. anything you can put in writing or is it just accumulated experience?
|
# ? Jul 2, 2015 16:30 |
|
coffeetable posted:anything you can put in writing or is it just accumulated experience? Be very careful
|
# ? Jul 2, 2015 16:34 |
|
coffeetable posted:anything you can put in writing or is it just accumulated experience? 90% of design patterns are "the original way we did this turned out to be unmaintainable poo poo, here's how you should do that sort of thing instead"
|
# ? Jul 2, 2015 16:52 |
|
coffeetable posted:anything you can put in writing or is it just accumulated experience? have somebody preternaturally talented at reducing complexity in charge of your codebase
|
# ? Jul 2, 2015 16:55 |
|
coffeetable posted:anything you can put in writing or is it just accumulated experience? It's hard to say. A lot of it is probably going to be a reaction in a code review where I go "noooo don't do this", but some of it has been materialized in blog posts, talks, and whatever:
there's probably more of it, and most of it would probably be easier to trigger during a conversation. Most of my time has also been spent working on server-side software, so it may not apply on any other area of the industry. MononcQc fucked around with this message at 17:05 on Jul 2, 2015 |
# ? Jul 2, 2015 17:00 |
|
MononcQc posted:It's hard to say. A lot of it is probably going to be a reaction in a code review where I go "noooo don't do this", but some of it has been materialized in blog posts, talks, and whatever: 5
|
# ? Jul 2, 2015 17:06 |
|
Stringent posted:have somebody preternaturally talented at reducing complexity in charge of your codebase
|
# ? Jul 2, 2015 17:10 |
|
The following paraphrased discussion happened in a meeting about providing alternative routers and stacks to customers a few months ago: coworker: "I want a single loving router to be installed there and nothing else" product: "Why don't we do X so we can provide customers with options of what to do with the router? Why just one that tries to do more?" coworker: "Because I can't have zero routers and that's what I'd ideally have." I still think that this is a great idea. It doesn't mean to go for monolithic apps that do everything in the world, but past a point, we're just creating problems for ourselves by trying to be flexible and super featureful. Also: predictability is loving great to have. I don't care for the fastest system if I can have an okay one that is reliable on these figures. There's nothing better for operations than well known operational boundaries. edit also: take a break. When you're tired and you can't see clear anymore, you'll spend 10x the amount of time you'd spent if you take a break and come back later. Knowing when to pause and recuperate can really help a whole drat lot for overall productivity. Know when you're out of your element, and go read books, papers, blog posts. Gain experience indirectly by people who solved these problems (or similar ones) before you did. It's invariably easier than rediscovering it all by yourself. MononcQc fucked around with this message at 17:13 on Jul 2, 2015 |
# ? Jul 2, 2015 17:10 |
|
MononcQc posted:It's hard to say. A lot of it is probably going to be a reaction in a code review where I go "noooo don't do this", but some of it has been materialized in blog posts, talks, and whatever: I identified with pretty much all of these points, but thinking back about how those mistakes were made, with the organization in place, none of them were avoidable.
|
# ? Jul 2, 2015 17:12 |
|
MononcQc posted:
batch processing is hell to start with, and cron is the worst batch processing system ever made. it has none of the features one would desire from batch processing. this drives open source weenies to move their batch processing into jenkins. this gets you some of the features you would expect from a commercial batch processor (chaining, reporting, triggering) but you're still loving sweating capacity planning and contention and/or aws magic this drives big data weenies to move their batch processing into yarn and map/reduce. this gets you the capacity planning and pretty good responses to aws issues but you just lost all the batch processing features you had on jenkins or your commercial batch processor, and now you are writing very big checks mononcqc is right. just say no. friends don't let friends batch process Notorious b.s.d. fucked around with this message at 17:21 on Jul 2, 2015 |
# ? Jul 2, 2015 17:19 |
|
Notorious b.s.d. posted:batch processing is hell to start with, and cron is the worst batch processing system ever made. it has none of the features one would desire from batch processing. Control-M is epic
|
# ? Jul 2, 2015 17:40 |
Hey what were this thread's recommendations for ML books again? I've got a physics background so bring on the math.
|
|
# ? Jul 2, 2015 19:32 |
|
I can't remember who posted it originally but the recommendation was for Machine Learning: a Probabilistic Perspective. Oh I found the post anyway:coffeetable posted:if you're comfortable with basic linear algebra and vector calculus, Machine Learning: A Probabilistic Perspective.
|
# ? Jul 2, 2015 19:44 |
|
VikingofRock posted:Hey what were this thread's recommendations for ML books again? I've got a physics background so bring on the math. i think it is Machine learning a probabilistic perspective, which i've started reading and found about right for a physics degree, but all the ones on .edu.cn domains are from the first printing and there's like 20 pages of errata not included in that one
|
# ? Jul 2, 2015 19:46 |
|
MononcQc posted:... this is the most important item in the list IMHO
|
# ? Jul 2, 2015 19:50 |
thanks guys
|
|
# ? Jul 2, 2015 19:57 |
|
just put in notice at my job today. i learned that we weren't really authorized to be using the version control we were using (TFS) and we weren't authorized to use the migration tool to the version control they want us using (svn) and that we basically were able to continue to work by escaping notice. too god drat much for me.
|
# ? Jul 2, 2015 20:05 |
|
Stringent posted:i'd love to hear more about this programming isn't actually my job so it's nothing big or substantial. i wrote a haskell program to parse through a few hundred megabytes of iOS app logs per day looking for errors that should manifest as an audible dropout in our speakers, and then i wrote a rust program to analyze audio files recorded from microphones in front of the same speakers also looking for dropouts, and then i wrote another haskell program to take the outputs of those previous two and assign them a score based on how many problems we found. i sometimes end up writing stuff like this to automate tasks people were doing by hand and i always like to use languages none of the software developers know because then i find it funny when management realized my stuff is actually useful and we should have the software team take over maintenance of whatever i wrote
|
# ? Jul 2, 2015 20:15 |
|
fleshweasel posted:just put in notice at my job today. i start my new job on monday
|
# ? Jul 2, 2015 20:18 |
|
triple sulk posted:i start my new job on monday how many jobs do you go through in an average year
|
# ? Jul 2, 2015 20:52 |
|
jobs are like toilet paper
|
# ? Jul 2, 2015 21:33 |
|
Notorious b.s.d. posted:god drat, sulk the company i'm leaving is based in canada so i was a contractor. i'm joining a company that i hope to actually stay with and has a very good name on paper.
|
# ? Jul 2, 2015 21:36 |
|
bobbilljim posted:should i work with php y/n eh. it's the worst plang by a much smaller margin these days the greater danger lies in who you're going to work with i'd want to ask some questions like why are you using php? acceptable reason imo is "we know it sucks but we'd rather pay you lots of money than rewrite a large codebase that works". "it's the only thing we know" is sensible but not a recommendation as a workplace what's your legacy code like and how are you handling it? how do you do testing? what other languages do you use when you have the opportunity? what experience do the people i'll be working with have? php has improved a lot as a language, but unless your team knows enough to make use of that and to actively mitigate the terrible parts it's still going to suck as much as its reputation
|
# ? Jul 2, 2015 22:55 |
|
gently caress! I've become the manager's "uhh I broke this and can't fix it, can you fix it" person I'm too good at database janitoring and figuring out the intricacies of our lovely product Luigi Thirty fucked around with this message at 23:00 on Jul 2, 2015 |
# ? Jul 2, 2015 22:57 |
|
Luigi Thirty posted:gently caress! I've become the manager's "uhh I broke this and can't fix it, can you fix it" person how does it compare to retail?
|
# ? Jul 2, 2015 23:07 |
|
i have one thing that i've learned over my medium term as a developer for actual money: write specs just write the drat spec. it is a hundred times easier to fix the code when it doesn't even exist yet.
|
# ? Jul 2, 2015 23:30 |
|
"but i can just code this up really fa-" WRITE THE loving SPEC FIRST
|
# ? Jul 2, 2015 23:30 |
|
you're going to spend enough time in the nitty gritty of "why doesn't this API call work the way I want" or "whoops i took the address of a temporary" as it is, you don't want to spend that time writing the wrong code
|
# ? Jul 2, 2015 23:31 |
|
that doesn't sound very agole
|
# ? Jul 2, 2015 23:31 |
|
just code
|
# ? Jul 2, 2015 23:32 |
|
Soricidus posted:that doesn't sound very agole it doesnt have to be a long spec, just write out "in order to do X i will do Y, Z and Q; in order to accomplish Q, B, C and D need to happen" and reason it all out then you can send it to your coworkers for comments while you start hacking, and they can point out issues long before you even implement the bugs
|
# ? Jul 2, 2015 23:33 |
|
|
# ? May 25, 2024 21:39 |
|
gently caress, does anyone know about MAPI integration? I have a c# console app that reads a set of custom properties from an msg file by invoking an outlook session and it works perfectly when run locally but on any other machine fails with an "error property http://schemas.microsoft.com/mapi/string/{00020386-0000-0000-C000-000000000046}/[PROPERTYNAME] is unknown" error. When I inspect the files with MFCMapi the properties are there and it is exactly the same file locally vs on the server yet it fails for no reason I can see. The properties are read using the PropertAccessor.GetProperty method on the message object and everything I can find says that this should work for custom properties as long as the namespace matches with the one set on the object. Also gently caress Outlook's "haha here's a generic error code that just means "failure" with no more information" behaviour
|
# ? Jul 2, 2015 23:36 |