How much JS have you had to tweak when used within TypeScript? Since you have types and objects and it's a first class citizen, I hear it gives you compile-type debugging. What's the syntax like? Has it changed the way you write FE code? Do you get any benefits of being able to use the CLR, like LINQ or what have you?
I'm looking at giving it a spin and am mostly looking at ease of use at this point.
How much JS have you had to tweak when used within TypeScript? Since you have types and objects and it's a first class citizen, I hear it gives you compile-type debugging. What's the syntax like? Has it changed the way you write FE code? Do you get any benefits of being able to use the CLR, like LINQ or what have you?
I'm looking at giving it a spin and am mostly looking at ease of use at this point.
The typing works well as long as you aren't interacting with any foreign JS code bases... at which point everything descends into over-usage of the "any" type and all the compile-time checking disintegrates
its like Actionscript that way, it's "strongly typed" until you decide you don't want to use types anymore
I haven't used the LINQ features but I know LINQ and if I had a reason to use it would definitely be awesome.
The "great" part about being a programmer is when something like the ICD10 transition, which you knew would cause problems, finds ways to annoy you that you never expected.
Working on a new report, which I'm basing off of an existing report. The report grabs a ton of different patient and encounter information.
For some reason, when I got to the part where I was getting all the primary and secondary diagnoses it started not returning any information.
The patients had already been qualified based on their diagnoses and procedures in an earlier query, this query was just taking the encounter identifiers I'd already gotten, and grabbing all the diagnoses, but not finding any.
After at least half an hour of digging around, and eventually commenting out qualifiers line by line, I found the problem.
Diagnoses are on one table, and joined to a nomenclature table which has among other things, the code vocabulary and description.
When I had qualified the patients, I merely checked to see if the diagnosis and the nomenclature items were both active, and the end effective datetime on the dx.
The part that was failing was also checking the end effective datetime on the nomenclature.
Which for all the ICD9 ones was apparently set to Sep 30 now. :rotate:
Which means that huge extract that runs monthly with like, 200 columns of data for 3k+ patients and takes a long time to run, hadn't returned anything for the diagnoses when it was run yesterday.
bonus semi-related fun:
When I was looking over the code for it yesterday, in the main query where it qualifies the patients, they'd written code to grow the record structure by 1000 records at a time.
Except in the line where the checked the count % 1000, they forgot to put the = 1, so it was growing by 1000 records 999/1000 times instead of 1/1000 times.
Fortunately it gets trimmed down to the actual size at the end, so it wasn't just sitting there taking memory.
0
NogsCrap, crap, mega crap.Crap, crap, mega crap.Registered Userregular
the worst part is that snomed only exists and is used by meaningful use because icd9 was going away and the mu certifying groups didn't want to settle on nomenclature that was going to change
It's pretty much worthless in the eyes of the physician
not a doctor, not a lawyer, examples I use may not be fully researched so don't take out of context plz, don't @ me
Does anyone know the name of the concept that MSOffice 2013+ uses for the 'file' menu.
I know the rest of the menus are a ribbon menu but the file seems to change the entire form.
Assuming it's like the other file menus that recent Microsoft programs use, I'm not sure if it's as much a "concept" as it was a menu that got converted to proto-hamburger form in the corner of the title bar, dodged everything else in the menu bar being converted to a ribbon by not being there at the time, and then got put back exactly where it was before mostly as it was before the general ribbonification.
it looks like they are using a similar approach to mine. i'm probably at or farther than their level of sophistication, and a requirement is the system be totally portable, so no "services"
I'm wondering why IDEs seem to take so long to load up. There's obviously a lot to them, but ultimately all the user is going to see when they first load it up is a menu or an editor window, which really shouldn't take anywhere near that long to generate. Is the idea just to load absolutely everything, so that once the IDE is running the user never has to wait for anything?
I'd be perfectly happy with an instant startup and a brief pause whenever I first use a component.
I'm wondering why IDEs seem to take so long to load up. There's obviously a lot to them, but ultimately all the user is going to see when they first load it up is a menu or an editor window, which really shouldn't take anywhere near that long to generate. Is the idea just to load absolutely everything, so that once the IDE is running the user never has to wait for anything?
I'd be perfectly happy with an instant startup and a brief pause whenever I first use a component.
I've wondered that myself. Netbeans takes forever to load, then has to rescan through the entire project to check for changes.
I'm wondering why IDEs seem to take so long to load up. There's obviously a lot to them, but ultimately all the user is going to see when they first load it up is a menu or an editor window, which really shouldn't take anywhere near that long to generate. Is the idea just to load absolutely everything, so that once the IDE is running the user never has to wait for anything?
I'd be perfectly happy with an instant startup and a brief pause whenever I first use a component.
I'm on board the with the other perspective -- I will sip my coffee while everything else loads, and then while I use Visual Studio my experience should be crisp AF.
TraceGNU Terry Pratchett; GNU Gus; GNU Carrie Fisher; GNU Adam WeRegistered Userregular
This is relevant to everyone's interests (except Bowen you javascript hater)
^js and also the json-like narsese macros should be familiar to many potential developers, since javascript currently is one of the most ubiquitous programming languages.
there is also a modified ^scheme interpreter which is probably a better choice since it wont involve strings that need to be interpreted back and forth, instead its symbols are homoiconic, expressed directly in NAL terms, which it can reprogram through reasoning.
NALObjects is a new toolset for binding plain-old java objects (POJOs) to a NAR. it uses dynamic class generation to produce proxy objects that generate NARS event beliefs and/or goals which contain type information, parameters, and return type of a Java class method ( if they can be successfully serialized/deserialized into terms, aka "termized").
it also reacts to goals that NARS can execute, attempting to invoke the specified methods of wrapped objects with any included parameters (these goals are input, for ex: training, or originate from NARS volition)
I, uh, might have to learn Ruby. What's a good place to start?
If you're already a programmer this is a good place to start, The ruby pickaxe book for further information, and ruby doc as a reference. If you google ruby and any class name (IE: "ruby array") the first result will probably be ruby doc's entry on array.
If you're new to programming, something more like Why's guide is a less formal way to get into the language.
Posts
Fixed a few errors, 300 more! I love whack-a-mole
99 bugs to patch in the code, 99 bugs to patch...
only with a butt load of configuration
every single day for the last six months
ask
Oh yeah, for sure.
They could use a lot of room for improvements.
How much JS have you had to tweak when used within TypeScript? Since you have types and objects and it's a first class citizen, I hear it gives you compile-type debugging. What's the syntax like? Has it changed the way you write FE code? Do you get any benefits of being able to use the CLR, like LINQ or what have you?
I'm looking at giving it a spin and am mostly looking at ease of use at this point.
The typing works well as long as you aren't interacting with any foreign JS code bases... at which point everything descends into over-usage of the "any" type and all the compile-time checking disintegrates
its like Actionscript that way, it's "strongly typed" until you decide you don't want to use types anymore
I haven't used the LINQ features but I know LINQ and if I had a reason to use it would definitely be awesome.
Working on a new report, which I'm basing off of an existing report. The report grabs a ton of different patient and encounter information.
For some reason, when I got to the part where I was getting all the primary and secondary diagnoses it started not returning any information.
The patients had already been qualified based on their diagnoses and procedures in an earlier query, this query was just taking the encounter identifiers I'd already gotten, and grabbing all the diagnoses, but not finding any.
After at least half an hour of digging around, and eventually commenting out qualifiers line by line, I found the problem.
Diagnoses are on one table, and joined to a nomenclature table which has among other things, the code vocabulary and description.
When I had qualified the patients, I merely checked to see if the diagnosis and the nomenclature items were both active, and the end effective datetime on the dx.
The part that was failing was also checking the end effective datetime on the nomenclature.
Which for all the ICD9 ones was apparently set to Sep 30 now. :rotate:
Which means that huge extract that runs monthly with like, 200 columns of data for 3k+ patients and takes a long time to run, hadn't returned anything for the diagnoses when it was run yesterday.
bonus semi-related fun:
When I was looking over the code for it yesterday, in the main query where it qualifies the patients, they'd written code to grow the record structure by 1000 records at a time.
Except in the line where the checked the count % 1000, they forgot to put the = 1, so it was growing by 1000 records 999/1000 times instead of 1/1000 times.
Fortunately it gets trimmed down to the actual size at the end, so it wasn't just sitting there taking memory.
depending on what you mean by profiling, have you looked at Strongloop's suite of tools?
PARKER, YOU'RE FIRED! <-- My comic book podcast! Satan look here!
the standard webkit tools dont quite get deep enough to be useful...
also there is Kojak
and then, ya webstorm has spyjs
PARKER, YOU'RE FIRED! <-- My comic book podcast! Satan look here!
this might interest you
http://dtab.io/
PARKER, YOU'RE FIRED! <-- My comic book podcast! Satan look here!
Also, snomed can go fuck itself too
They're too busy fucking each other.
It's pretty much worthless in the eyes of the physician
I know the rest of the menus are a ribbon menu but the file seems to change the entire form.
Assuming it's like the other file menus that recent Microsoft programs use, I'm not sure if it's as much a "concept" as it was a menu that got converted to proto-hamburger form in the corner of the title bar, dodged everything else in the menu bar being converted to a ribbon by not being there at the time, and then got put back exactly where it was before mostly as it was before the general ribbonification.
it looks like they are using a similar approach to mine. i'm probably at or farther than their level of sophistication, and a requirement is the system be totally portable, so no "services"
they should be paying me for consulting!
git commit -m "FIRE"
You don't want to be the one who breaks the build....
So truly, it is a game of reverse check-in chicken, where the last person to commit goes and diaf. :P
git checkout -b ticket#-omgFire
i.e. push to branch so that you can worry about merges later but still have code saved
PARKER, YOU'RE FIRED! <-- My comic book podcast! Satan look here!
I'd be perfectly happy with an instant startup and a brief pause whenever I first use a component.
I've wondered that myself. Netbeans takes forever to load, then has to rescan through the entire project to check for changes.
I'm on board the with the other perspective -- I will sip my coffee while everything else loads, and then while I use Visual Studio my experience should be crisp AF.
SE++ Forum Battle Archive
If you're already a programmer this is a good place to start, The ruby pickaxe book for further information, and ruby doc as a reference. If you google ruby and any class name (IE: "ruby array") the first result will probably be ruby doc's entry on array.
If you're new to programming, something more like Why's guide is a less formal way to get into the language.
...Ew I now own an Apple product.
Shout proudly whenever you have it in the presence of others that you didn't pay for it, so that's different.
Annoy the fuck out of everyone repeatedly!
Yay free iPad though, they are handy around the house I find.
they have always been a pleasure to use compared to the android alternatives
Their operating system is aces too.
Yesss, be the insufferable smug asshole from both sides.
I may install a bunch of games for my daughter to learn from on it and use it as a reward if she's been good.
The first time I used an iPad it hard crashed starting up an app.
I made a game, it has penguins in it. It's pay what you like on Gumroad.
Currently Ebaying Nothing at all but I might do in the future.
What app?
I've never had this issue except with third party apps, even then, it's super rare.
Business apps tend to be the worst because business is all about cost cutting, and, surprise, offshored workers are not known for QC.
No no no.
The only steps should be:
./fire.sh
Exit building.
Why are you not automating this? You're a programmer with an algorithm for god's sake!
3DS: 0473-8507-2652
Switch: SW-5185-4991-5118
PSN: AbEntropy