|
Toady posted:the analysis of build's source confirms your suspicions about it that's decompiled object code though, isn't it? I'd expect something like that to be generated by (perhaps multiple layers of) macros that let you specify more semantically what you're trying to achieve
|
#
?
Oct 13, 2015 06:20
|
|
|
|
| # ? Jun 5, 2024 20:44 |
|
|
Barnyard Protein posted:His wide-ranging Harry Potter fan fiction story Harry Potter and the Methods of Rationality illustrates topics in cognitive science and rationality.[3]:37[14][15][16][17][18][19] he makes a living at that you know had you seriously not heard of the "
|
|
#
?
Oct 13, 2015 06:24
|
|
|
Barnyard Protein posted:people here are trying to create a programming language from scratch which expresses its semantics through XML syntax. so far they've basically recreated javascript but without methods, and their if statements don't support elses. the worst part is that the xml file is an intermediate output that they are generating based on the primary input source: an Excel file. the thought process is that it would let non-technical people be able to write and read tests non-technical people never wrote or read those tests writing and reading the tests also sucked
|
|
#
?
Oct 13, 2015 06:33
|
|
|
writing and reading tests is writing and reading code using a bad dynamically typed programming language with extreme amounts of required boilerplate is not a good idea unfortunately i've suffered the above twice so far i'll probably wind up with a hat trick at another job
|
|
#
?
Oct 13, 2015 06:35
|
|
|
also xml as a programming language is just atrociously re-invented lisp
|
|
#
?
Oct 13, 2015 06:36
|
|
|
i'm sure you will be not surprised to learn that the test project constantly fails to build and almost no tests have been written because reasons. the main reason is that i refuse to let the test team instrument the project under test with their magic strings that they need to get their dumb framework to work and they are actively developing the framework and dumb language instead of testing.
|
|
#
?
Oct 13, 2015 06:43
|
|
|
suggest they use cucumber or fitnesse or some other bullshit dumb testing framework for non-technicals someone has already invented for them at least they'll end up failing faster
|
|
#
?
Oct 13, 2015 06:45
|
|
|
yeah unless your company is a company whose product is programming languages you probably shouldn't be loving around with making programming languages at work
|
|
#
?
Oct 13, 2015 06:49
|
|
|
can your non-technical people write and read lisp code? no? then they certainly aren't going to be able to read an even more verbose homebrewn xml variant of it
|
|
#
?
Oct 13, 2015 06:53
|
|
|
MALE SHOEGAZE posted:yeah unless your company is a company whose product is programming languages you probably shouldn't be loving around with making programming languages at work yeah i said "now we have two problems" to this manager, "lets use groovy instead, it is already a language that meets all of our current and future usecases" but i got some talk back about being on a common toolchain and testers need an easy format and my diffidence grew and grew until i nodded and left the meeting room. i am going to do it. i am going to type a big long email and send it to the people backing this bad decision. a good buddy of mine who works at this company is experienced in these matters, he said its not as satisfying as i think it will be to point at an email and say "i told you so" when a project fails, but gosh.
|
|
#
?
Oct 13, 2015 07:00
|
|
|
comedyblissoption posted:can your non-technical people write and read lisp code? that is where the excel sheet comes in to play 
|
|
#
?
Oct 13, 2015 07:02
|
|
|
eschaton posted:that's decompiled object code though, isn't it? I'd expect something like that to be generated by (perhaps multiple layers of) macros that let you specify more semantically what you're trying to achieve
|
|
#
?
Oct 13, 2015 09:01
|
|
|
found this decompiled object code in my computercode:
|
|
#
?
Oct 13, 2015 11:42
|
|
|
pepito sanchez posted:found this decompiled object code in my computer I didn't know you worked in Haskell?
|
|
#
?
Oct 13, 2015 13:14
|
|
|
MALE SHOEGAZE posted:yeah unless your company is a company whose product is programming languages you probably shouldn't be loving around with making programming languages at work we had develop our own framework for unit tests, but as it's already established we work with bogus terrible languages from the 60s so we didn't really have a choice. Unit tests in C# are so beautiful and painless in comparison.
|
|
#
?
Oct 13, 2015 17:53
|
|
|
i don't understand how someone's first reaction to solving a problem is "well clearly no one else in the world has encountered this problem, much less provided a solution. time to create a solution from scratch using only the information i have in my head at this present moment".
|
|
#
?
Oct 13, 2015 18:15
|
|
|
that's not what it is. Its "I haven't solved this problem yet, so time for me to solve it!" they don't want the existing solutions because they don't understand that that's not the problem they're supposed to be solving.
|
|
#
?
Oct 13, 2015 18:17
|
|
|
its probably the #1 way to tell a novice programmer from an experienced one.
|
|
#
?
Oct 13, 2015 18:18
|
|
|
novice programmer: "oh boy, another fun problem for me to solve!" experienced programmer: "oh boy, another problem for me to learn the api of the tool which already solves it  "
|
|
#
?
Oct 13, 2015 18:21
|
|
|
it's NIH syndrome
|
|
#
?
Oct 13, 2015 18:23
|
|
|
Bloody posted:it's NIH syndrome my employer has a terminal case.
|
|
#
?
Oct 13, 2015 18:24
|
|
|
qntm posted:novice programmer: "oh boy, another fun problem for me to solve!" maybe a new language is easier to deal with than the tool's api and then the next guy will decide a new language is easier to deal with than your language
|
|
#
?
Oct 13, 2015 18:25
|
|
|
Barnyard Protein posted:i don't understand how someone's first reaction to solving a problem is "well clearly no one else in the world has encountered this problem, much less provided a solution. time to create a solution from scratch using only the information i have in my head at this present moment". This would be a fair criticism if we used a non-joke language. Please tell me where to find all these awesome libraries for ANSI MUMPS. We have to DIY it because they literally don't exist. And yes, it is painful, but not as painful as trying to port millions of lines of code to a new platform.
|
|
#
?
Oct 13, 2015 18:32
|
|
|
LeftistMuslimObama posted:This would be a fair criticism if we used a non-joke language. Please tell me where to find all these awesome libraries for ANSI MUMPS. We have to DIY it because they literally don't exist. And yes, it is painful, but not as painful as trying to port millions of lines of code to a new platform. sorry i was self-centered-posting about about these bozo's i was talking about yesterday with their artinsinal xml testing language and didn't provide context in that last post. yeah if you're using ANSI MUMPS you should be given a small bottle of fine scotch every morning and a voucher for scream therapy
|
|
#
?
Oct 13, 2015 19:11
|
|
|
my job right now is to turn a home grown spreadsheet system into a sql/c#/asp.net thing
|
|
#
?
Oct 13, 2015 20:08
|
|
|
should be pretty easy.
|
|
#
?
Oct 13, 2015 20:16
|
|
|
yeah, ive been given 3 months and i don't envision it should really take that long tbh
|
|
#
?
Oct 13, 2015 21:27
|
|
|
Awia posted:home grown mods
|
|
#
?
Oct 13, 2015 21:45
|
|
|
LeftistMuslimObama posted:This would be a fair criticism if we used a non-joke language. Please tell me where to find all these awesome libraries for ANSI MUMPS. We have to DIY it because they literally don't exist. And yes, it is painful, but not as painful as trying to port millions of lines of code to a new platform. is there no way to interface it with code written in better languages? I mean science and finance have managed to reduce the amount of new code they have to write in fortran and cobol that way
|
|
#
?
Oct 13, 2015 21:49
|
|
|
if you really want to implement things yourself, find a young open source programming language on github and have at it.
|
|
#
?
Oct 13, 2015 22:03
|
|
|
fart simpson posted:if you really want to implement things yourself, find a young open source programming language on github and have at it. or just learn to love common lisp.
|
|
#
?
Oct 13, 2015 22:05
|
|
|
jony neuemonic posted:my employer has a terminal case. A bunch of developers at Reuters came up with a new display format for time: 00:00:00:000:000:000, I cannot find a single instance on Google. Had to log a bug to change it to 00:00:00.000000000. I can understand NIH for coding, but for absolutely everything? I guess one could joke about vernacular spelling issues of English as a foreign language of being that they are not allowed to reference a dictionary for correct spellings 
|
|
#
?
Oct 13, 2015 22:05
|
|
|
Soricidus posted:is there no way to interface it with code written in better languages? I mean science and finance have managed to reduce the amount of new code they have to write in fortran and cobol that way Not easily, no. The popular MUMPS implementation, Cache, is basically it's own runtime environment. While there are some rudimentary ways to get out into the hosting OS zthey don't actually comply with the ANSI standard and we have to avoid using them to maintain vendor neutrality. Believe me, we've looked at a platform transition from many angles over the years, and there's just no feasible way to do it that wouldn't have a huge negative impact on our customers. We have organizations like Kaiser Permanente using our software, and it would cost them hundreds of millions to roll out a new infrastructure to support a platform migration. That's just how enterprise software is. You accept some pain on the developer side to make things easy and cost effective for customers. In fact, it would be a terrible programmer move to force a platform migration to get access to better dev tools if it wasn't motivated by a genuine need on the client's part that can't be solved on the current platform.
|
|
#
?
Oct 13, 2015 22:35
|
|
|
eschaton posted:that's decompiled object code though, isn't it? I'd expect something like that to be generated by (perhaps multiple layers of) macros that let you specify more semantically what you're trying to achieve it was open sourced in the early 2000s. the engine is one ENGINE.C file with nearly 9000 lines and few comments. granted, he was a teenager
|
|
#
?
Oct 14, 2015 00:47
|
|
|
LeftistMuslimObama posted:That's just how enterprise software is. You accept some pain on the developer side to make things easy and cost effective for customers. In fact, it would be a terrible programmer move to force a platform migration to get access to better dev tools if it wasn't motivated by a genuine need on the client's part that can't be solved on the current platform. i think it was Avenging Dentist who coined "Stackholm syndrome" he used to work at epic until he had a total breakdown and then got a much better job writting c++
|
|
#
?
Oct 14, 2015 01:53
|
|
|
LeftistMuslimObama posted:Not easily, no. The popular MUMPS implementation, Cache, is basically it's own runtime environment. While there are some rudimentary ways to get out into the hosting OS zthey don't actually comply with the ANSI standard and we have to avoid using them to maintain vendor neutrality. go the other way, then, and bring the good languages to mumps. it works for javascript! what could possibly go wrong! heck, give me just $10m and i'll do it for you, saving you time and your clients money where by "it" i mean "get on the next flight to a country with a nice climate and no awkward extradition treaties"
|
|
#
?
Oct 14, 2015 02:09
|
|
|
MUMPS actually seems like a cool language that had what sounds like ORM, in '66!quote:Perhaps the most unusual aspect of the M language is the notion that the database is accessed through variables, rather than queries or retrievals. Spiff up the syntax and spell the name in lower case and it could be the next web language.
|
|
#
?
Oct 14, 2015 02:21
|
|
|
Brain Candy posted:he used to work at epic until he had a total breakdown and then got a much better job writting c++ and now he's writing javascript
|
|
#
?
Oct 14, 2015 02:49
|
|
|
Brain Candy posted:i think it was Avenging Dentist who coined "Stackholm syndrome" Yeah, he posts in the Madison thread with the rest of us. It's possible Epic will eventually break my spirit, but considering that my high school days involved gang fights and an extensive sealed record and I've managed to claw my way up to 75k a year with a poetry composition degree to my name, I'm p happy with Epic atm. It pretty heavily depends on what area of the software you work on too. Some areas people just couldn't resist being "clever" with the language's features back in the day so you end up with garbage like a global with 17 subscript levels where the subscript names represent critical information about the data you're working with. Meanwhile other areas had developers who liked to keep things simple and it's all super readable and easy to maintain. I get the feeling AD didn't work on one of those. "Stringly typed" is definitely all over the place in our code though. You don't really have a choice in MUMPS though. There's good and bad examples of all of it. Unfortunately there's not really any way to anonymize any of it to the point where I could share it here. I do have to say, though, working in C in my intro to OSes class was pretty fun and I felt like a genius at the end of that course. I need to find ways to maintain my "real language" skills outside of work. At least we use C# for our newer frontend now.
|
|
#
?
Oct 14, 2015 03:15
|
|
|
|
| # ? Jun 5, 2024 20:44 |
|
|
LeftistMuslimObama posted:Not easily, no. The popular MUMPS implementation, Cache, is basically it's own runtime environment. While there are some rudimentary ways to get out into the hosting OS zthey don't actually comply with the ANSI standard and we have to avoid using them to maintain vendor neutrality. Serious question but why? You like, seem to enjoy doing this. Is it because you enjoy having to do terrible programming poo poo? Because yeah, I'm starting to feel like my future is maybe in fortran or something where inventing stupid DSLs is a good use of my time.
|
|
#
?
Oct 14, 2015 06:55
|
|