@thecreativeone91 said:
@tonyshowoff said:
When I was in Russia there wasn't a Taco Bell for thousands of kilometres, it's actually the first thing I got when I got back to a country with real food.
Taco Bell is Real food? I like Mexican but I wouldn't hardly call Taco Bell Mexican.
Yeah well compared to boiled cabbage covered in sour cream it's practically gourmet
When I was in Russia there wasn't a Taco Bell for thousands of kilometres, it's actually the first thing I got when I got back to a country with real food.
@scottalanmiller said:
No "splitting" functionality, yet. Hopefully the developers of NodeBB take note.
NodeBB is clearly a clone of Discourse, I figured it'd be on their to do list sin ce Discourse allows that.
@creayt said:
Ok I lied. It'd be too detrimental to productivity to hop in and try to wield WebStorm for a normal day of productivity, so I'll play around with it more after hours.
I used to use mostly plain text editors, I didn't even like highlighting before I started using PhpStorm. After I got into the groove it's pretty crazy how much time I save and I don't have to take the same sort of short cuts which lead to bad practices/breaking standards.
At this point it's looking like it might be helpful for big, legacy, unwieldy PHP or Java projects, but can't really handle what I'd need it to do as a code authoring platform. I hope I'm wrong, but even things like my additioanl mouse button mapping shortcuts aren't working, and those are standardized across most modern apps ( like Ctrl + W to close a file, I mean yikes JetBrains, get with the times ). I'll write more when I have a chance to really dig in.
I actually had to remap a few of the preset short cut keys as well, some of the more useful features had some of the most obscure ones, but they're easy to change in the settings.
@creayt said:
So far I've had to right click the file tabs and chose split vert/horizontally, which feels cumbersome and costs more time, especially when you want to go from looking at a few, to looking at some extra related code for a few moments, and then back to your original position. Time is money!
That's one usability issue, but you have to view it in perspective of everything else it does. Plus also there are short cuts to jump to the last file you were in and in the position, and you can keep going back historically or move forward as well. There's probably a plugin thought that does what you're talking about.
Well isn't that presumptuous.
Well, yes, I mean having to defend the use of an IDE is sort of unusual and I can only presume that someone who needed justification for standardised and IDE enforced code organisation hadn't dealt with a large project.
The project I'm working on right now is massive, and doing a quick check it shows that it has 15,737 code files, not including front-end resources like images or video.
And you edit that only in Sublime Text know what ever file does, and where every class is, etc?
From what you were saying I was right, and you were talking about something totally different, code navigation.
That's part of it, yes. But the organisation is important because it helps with refactoring, naming, maintenance, etc. Your perspective seems to be based on much smaller projects, even if you seem to be dealing with an obscenely large one. Our ERP is up to about 2 million lines and it's only 4,500 files, including HTML templates.
But in reality I probably navigate my code as quickly or more quickly than you do
I doubt that, I can ctrl + click any class or anything and find it's origin, or right click and find all usages of it anywhere, and the context, or do the same with alt + b if I don't want to leave my keyboard, and since all the code is indexed, it barely takes a second despite 2 million lines of code in this ERP.
It sounds like you're making these giant assumptions about Sublime Text based on your perception of its class of application, throwing it into a lump with "text editors" in your mind, when in reality it can be extended to do endless things in the same way that any "formal IDE" ( like Eclipse ) can.
I never said it wasn't a great editor, I said it was better than Notepad++ earlier. Eclipse is a piece of garbage, it's like a busted version of ... I don't even know, it's just bad.
I personally see having to rely on your IDE to help you figure out where other code lives as the digital equivalent of a training wheel.
This is from the perspective of someone who hasn't worked on a team with probably more than two people and really doesn't know what a large project is, at least that's my perception. This is sort of silly, honestly, the idea you could know hundreds or thousands of classes, especially if you work on multiple large projects. Not only that, why spend the extra effort to do that in the first place if you don't have to?
It's better to deeply understand your code, strategize its organization in a way that that problem is automatically solved and that you know the exact location of every piece of code by convention, and could see it being useful when you take over a large project from someone else and have to learn it quickly by exploration using that method, although in all honesty that'd be just as easy to do using Sublime's keystroke navigation features.
IRL enterprise doesn't work this way, sorry, it just doesn't, and it can't, otherwise large projects would never get done if everyone on the team had to create a special bond with all of the code. Personal projects that does happen though.
Right, which is why your IDE has nothing to do with your code organization, which was my point.
Yes, it does, because it keeps you on track and also the organisation of the code can tell your IDE about it as well, such as testing, class autoloading/including (for use in code), etc. You seem to believe that structure and names are meaningless beyond what you feel like, but that's not the case; certainly you can approach it that way, but then it becomes unmaintainable if the directory file structure and naming does not match name spaces and classes in some meaningful way.
It's just sort of weird to defend using an IDE and have it called training wheels, that's certainly not a professional perspective on development, and is basically out of line with probably much of the professional programming world; especially in team environments.
Is WebStorm written in Java? The choppy and awkward interface kind of feels like it was. If so that's very unfortunate.
Yes, I also consider that a problem, and it being a memory hog is a pretty big give away that it's Java. However, it's not that a problem if you've got enough system resources; I'd certainly rather see it written in C++.
Also missing the granular font-rendering and aliasing controls off the bat, and it's sad that you can't shrink and magnify the font size w/ Ctrl +/- like in a browser.
You can probably create a short cut for that, you can for damn near everything else. I suggest using it longer than a day, the problem is it takes time to really understand how to use it properly to get everything out of it, because you can approach it like a regular text editor and get no benefits at all; I've seen this happen.
Wow, there's a tangible delay between when you open a file and when the colorization and text-modification renders. That feels gross! And I'm on a 2015 3.9 Ghz quad w/ a PCIe SSD and 16GB. I'm not sure I could live with performance like this, it's slowing me down within a few minutes of firing it up.
That's really weird, my specs are almost the same and I don't experience that sort of latency. Sometimes on larger files, first view, there's a lag as it renders in the background, but that's substantially large files, and that's why the single responsibility principle is your friend.
@scottalanmiller said:
Or, far better, that ReactOS and other projects like Wine could stop trying to rewrite the Windows code and could focus on replacing the components that Microsoft can't release. Piece by piece they could leverage the MS code to make the whole thing free. Much like the BSD projects did long ago to free UNIX from AT&T.
Well, I think you're just asking for the moon when it comes to a lot of people interested in Windows.
@scottalanmiller said:
I don't think it was PBS I would have seen it.
http://en.wikipedia.org/wiki/Computer_Chronicles
Yeah, PBS
@scottalanmiller Yeah, it was called Computer something or other and it was on PBS I think, he was a host for a while. They used to re-air that a long time ago.
@scottalanmiller I know they're unrelated, I was merely pointing out command and how it was controlled in the same manner, though you can do a lot of the same operations. And CP/M, now there's something from a time machine. The APIs were originally so close it often didn't take much editing, sometimes none, to port a piece of software.
@thecreativeone91 in NT4 (NT 3.51 and before had NewShell only I think), because MSDOS was not a part of NT, but I know what you're talking about, the little MSDOS icon. I think until XP an emulated "command" (MSDOS) was available from run like cmd, with limitations like no copy and pasting, irritating crap like that. The great thing was that it often ran outside the bounds of group policy, so if you were locked out of cmd, certainly you could open command.
@thecreativeone91 said:
@tonyshowoff said:
ReactOS would work out
Would there even be a point to ReactOS if Windows was open source? Seems about the only reason would be so they could say they finally did something besides an alpha.
Licensing, not just because I am certain Windows would not be released under GPL or anything close to that, but also because things within it are held patent by other people and may not even be available or carry other licensing schemes, and it'd be easier to just rewrite it into one thing, with one licensing scheme.
Even the most ignorant computer users are vastly more educated in computers and software differences than they were 10, 15, 20 years ago, so I don't see that happening. Normal people already know there's a difference between an iPhone and a Windows phone, even if they can't articulate what it is, and since DOS no longer has any ubiquity at all, I can't even imagine people calling Linux that, and I know you were referring to calling other things Linux, well I'm not so sure, it's certainly possible, but ideally most normals won't even see the CLI at all, regardless of OS.
The best outcome I could see is that maybe more of the Windows API would be implemented in things like Wine and/or perhaps things like ReactOS would work out much better and/or one could add the API right into the OS, similar to how FreeBSD has Linux binary support, there could be Windows binary support as well.
I won't say it'll never happen, in the past Microsoft has released source code to very old products like MSDOS to MSDN members, so it could happen one day, but I imagine there are tons of patent and copy right issues involved, and likely licensing with other companies too that may not want their code distributed freely.
@creayt said:
How your code is organized depends on how you organize it, into files and folders. What am I missing here?
I think you're missing having worked on extremely large projects, for example a large enterprise project can be thousands of files. Being able to ctrl + click a class, function, method and it take me to exactly where it is, or easily search for part of a file name, class name, or something else, is a powerful thing that allows me to better and more easily organise my code is functionality products like PhpStorm offer that text editors don't, aside from obvious things like proper hinting, type recognition, etc.
The organisation of the code itself depends on what, if any, standard you're following, and so on, and it's obviously still easy to also badly organise your projects with something like PhpStorm, but being able to find things in a second (or less) in an indexed system and also properly name and organise them (things like PhpStorm will warn you if the naming and/or directory structure is off scheme) are vastly superior to having to search all files for my class, method, etc or go manually find the file with the class I'm looking for, and also cull results which are similar matches, and do it all manually.
So yes, it's organised depending on how you organise it, but that doesn't mean you did a good job. In the same way just because one writes code doesn't make it good, just because one wrote it.
I don't think anyone uses XML for almost anything in 2015. JSON won for most intents and purposes.
For transport over the web, yes JSON has won, but XML is still widely in use, especially amongst people who write in .NET or Java, and not just for protocols, but storage, metadata, etc. Many insanely verbose "standards" exist for XML eDocs out there, lest we forget all the insane things that use XML like many types of configuration files, Microsoft Office documents/spreadsheets/etc, just tons of things.
@creayt said:
@tonyshowoff said:
This isn't the worst thing I've ever seen, but again feels too XML-like for my taste, and in fact aside from return x, this whole thing is XML parse-able; though I imagine less strict XML engines would allow an argument without value.
Right, I think most people would probably opt for the script-based syntax for a class, but the point was it doesn't encourage one thing or another, you can do everything you want to in either of the two ways, the art becomes prudently judging when to use which:
class { function doSomething( a ){ return x; } }
That's much better 
