Bruno Santos wrote in post #978012:
redcar is awesome
use to install
$ gem install redcar $ redcar install $ redcar
Thanks for the reminder. I'll check it out. As I've said elsewhere, I'll happily use a *good* IDE for Raila if I find one.
Best,
Oh, absolutely, but I try to discourage bad choices. Using an IDE for Rails is IMHO a bad choice.
You're wrong Mamen, not a bad choice at all. You may prefer a plain editor, that's fine, but that does not make other people's choices bad.
So what? It's not worth the clunky interface.
*For you*.
But RubyMine has a lot of knowledge about Ruby on Rails.
Such as?
Such as? RM has dozens of features specific to Rails, you don't even know them, and you have all these strong opinions about choosing an IDE as RM? Common!
Xavier Noria wrote in post #978074:
Oh, absolutely, but I try to discourage bad choices. Using an IDE for Rails is IMHO a bad choice.
You're wrong Mamen, not a bad choice at all. You may prefer a plain editor, that's fine, but that does not make other people's choices bad.
I see too many people addicted to IDEs as a crutch and unable to function without them. That is bad.
When not used as a crutch, a good IDE is a joy. It makes life easier and automates repetitive tasks in development. That is good.
However, Rails is not well suited to conventional IDEs. They make Rails development harder. That is bad. They attract people who would use them as crutches. That is bad. As far as I can see, they provide no particular advantage to compensate for the annoyance they introduce. That is very bad.
Therefore, using a conventional IDE for Rails development is a bad choice. I think the issues here are fundamental enough that this ceases to be a matter of taste.
So what? It's not worth the clunky interface.
*For you*.
Right. That *is* a matter of taste.
But RubyMine has a lot of knowledge about Ruby on Rails.
Such as?
Such as? RM has dozens of features specific to Rails, you don't even know them, and you have all these strong opinions about choosing an IDE as RM? Common!
I tried to use RubyMine. I really wanted to like it. I didn't see the point of it. What am I missing?
Best,
Michael Pavling wrote in post #977960:
I *can* code without a debugger integrated (or indeed, at all) - but I don't want to.
If you were my client, and you responded "because I want it" when I asked why a feature was of value to you, you wouldn't get the feature.
Now *that's* customer service! 
(but poor economics [1])
I try to discourage bad choices. Using an IDE for Rails is IMHO a bad choice.
I think you need to reconsider the difference between "discourage" and "deny"... denying your customers features you don't like is not the same as discouraging them.
So...have you used RDB at the command line? What do you not like about it?
Yes I have; before I started using Netbeans.... and you call IDEs "clunky"!
I was actually a little surprised when I started working with Rails to find that IDEs provided no benefit -- but that *is* the case.
You often say a variation on this statement, and I never really understand what you mean: how can an IDE not "suit" Rails? It would help me if instead of saying "Don't use IDEs or you're all cripples", you explained how you work front-to-end to highlight how features of IDEs [2] don't give you any productivity benefit, while using an IDE would hamper you.
[1] wants vs needs - Google Search http://www.businessknowhow.com/marketing/wants.htm
[2] For me it's the debugger and source control integration that boost
my productivity (even while having to wait for the rest of the bloated
app to limp along with me They're pretty much the only things I
want [3] in addition to the syntax highlighting that every beefy text
editor offers.
[3] "want" again, not "need". I want a black car over a pink one, I
don't *need* a black car 
Michael Pavling wrote in post #978091:
Michael Pavling wrote in post #977960:
I *can* code without a debugger integrated (or indeed, at all) - but I don't want to.
If you were my client, and you responded "because I want it" when I asked why a feature was of value to you, you wouldn't get the feature.
Now *that's* customer service!
It's excellent customer service. My customers get exactly what they want, and do not spend money on things they don't want.
The article you linked to is a red herring in this connection. If you want something, you should still be able to explain *why* you want it, not just "because I want it".
I try to discourage bad choices. Using an IDE for Rails is IMHO a bad choice.
I think you need to reconsider the difference between "discourage" and "deny"... denying your customers features you don't like is not the same as discouraging them.
It's not a question of features I don't like. It's a question of features that can't be justified. "Because I want it" is no justification.
So...have you used RDB at the command line? What do you not like about it?
Yes I have; before I started using Netbeans.... and you call IDEs "clunky"!
Some IDEs are clunky. Some aren't.
I'm addicted to GUI tools, but I love RDB's command-line interface. Editor integration would be pleasant, but I don't really notice the lack very strongly.
I was actually a little surprised when I started working with Rails to find that IDEs provided no benefit -- but that *is* the case.
You often say a variation on this statement, and I never really understand what you mean: how can an IDE not "suit" Rails?
Excellent question. Conventional IDEs tend to be designed with "heavy" languages like Java or Obj-C in mind, and it shows in the architecture. As I see it, the primary benefits of conventional IDEs include: * code completion (maybe -- non-IDE editors do this too) * code generation at a higher level of abstraction than the framework itself can provide * automation of builds and other repetitive tasks * visual GUI design * generally making up for usability shortcomings in the development environment itself
As I see it, none of these features provide any significant benefit with Rails: * Code completion simply doesn't work well in a dynamic language like Ruby,
Michael Pavling wrote in post #978091:
Michael Pavling wrote in post #977960:
I *can* code without a debugger integrated (or indeed, at all) - but I don't want to.
If you were my client, and you responded "because I want it" when I asked why a feature was of value to you, you wouldn't get the feature.
Now *that's* customer service!
It's excellent customer service. My customers get exactly what they want, and do not spend money on things they don't want.
The article you linked to is a red herring in this connection. If you want something, you should still be able to explain *why* you want it, not just "because I want it".
I try to discourage bad choices. Using an IDE for Rails is IMHO a bad choice.
I think you need to reconsider the difference between "discourage" and "deny"... denying your customers features you don't like is not the same as discouraging them.
It's not a question of features I don't like. It's a question of features that can't be justified. "Because I want it" is no justification.
So...have you used RDB at the command line? What do you not like about it?
Yes I have; before I started using Netbeans.... and you call IDEs "clunky"!
Some IDEs are clunky. Some aren't.
I'm addicted to GUI tools, but I love RDB's command-line interface. Editor integration would be pleasant, but I don't really notice the lack very strongly.
I was actually a little surprised when I started working with Rails to find that IDEs provided no benefit -- but that *is* the case.
You often say a variation on this statement, and I never really understand what you mean: how can an IDE not "suit" Rails?
Excellent question. Conventional IDEs tend to be designed with "heavy" languages like Java or Obj-C in mind, and it shows in the architecture. As I see it, the primary benefits of conventional IDEs include: * code completion (maybe -- non-IDE editors do this too) * code generation at a higher level of abstraction than the framework itself can provide * automation of builds and other repetitive tasks * visual GUI design * generally making up for usability shortcomings in the development environment itself
As I see it, none of these features provide any significant benefit with
Rails:
* Code completion simply doesn't work well in a dynamic language like
Ruby, particularly when you throw in Rails' heavy use of dynamic
metaprogramming.
* Rails doesn't rely on generated code in the sense that Java frameworks
do. It does have generator scripts, but those are meant to be run once
and the results customized, unlike (say) Struts where there are several
sets of generated files that must be kept in sync for each build.
* Build automation from the IDE isn't as critical in Rails either.
Rails isn't compiled and doesn't have a complex build process, and
anyway, there's Rake, which unlike (say) Ant has build scripts that
humans can actually write.
* Visual GUI design? Rails' GUI is HTML. Any Web developer who
primarily uses a visual HTML tool deserves what he gets.
* As for making up for shortcomings...well...Rails has its share, of
course. But the framework was designed for ease of developer use
without an IDE, and it shows. What shortcomings Rails has (from this
perspective) don't seem to be fixable by the current crop of IDEs.
So...Rails has no use for the big wins of IDEs. IDEs such as NetBeans are huge, resource-intensive programs. They cripple developers by encouraging dependence on them. They do not, as far as I can see, offer any advantage at all for Rails development. Why use them?
I'll turn the question around: what do you get out of using an IDE for Rails, in terms of features that a decent editor wouldn't provide?
It would help me if instead of saying "Don't use IDEs or you're all cripples", you explained how you work front-to-end to highlight how features of IDEs [2] don't give you any productivity benefit, while using an IDE would hamper you.
See above. I use KomodoEdit (a good project-aware editor) with Ruby
and Haml syntax highlighting modules, along with GitX and 6 consoles
open in iTerm. One runs script/server, one runs autotest, and the rest
are available for random command-line tasks.
[2] For me it's the debugger and source control integration that boost my productivity (even while having to wait for the rest of the bloated app to limp along with me
NetBeans' Git plugin is fantastic. I just don't see it as fantastic enough to saddle myself with the rest of the IDE.
BTW, even if NB is officially dumping Rails, couldn't you still use its Ruby support? And don't you think someone is likely to take up maintenance of the Rails tools?
[3] "want" again, not "need". I want a black car over a pink one, I don't *need* a black car
But you can explain how black adds value for you -- "I find pink ugly", "I live in a cold climate and so I want a color that absorbs heat", whatever.
Best,
If you were my client, and you responded "because I want it" when I asked why a feature was of value to you, you wouldn't get the feature.
Now *that's* customer service!
It's excellent customer service. My customers get exactly what they want, and do not spend money on things they don't want.
Again, I think you're confusing needs and wants (and blurring it more
with your interpretation of what they need). They can hardly "get what
they *want*" if you also decide that they "wouldn't get the feature".
Either way, it's academic, as how you manage your customer relations
is up to you (but if they *really want* those features that you and I
both know are pointless... if you don't want to do it, send them my
way and I'll give them a quote for the work
The article you linked to is a red herring in this connection. If you want something, you should still be able to explain *why* you want it, not just "because I want it".
Getting into psychology now... sometimes people *can't* justify their wants (or they just like something more for no particular reason).
You often say a variation on this statement, and I never really understand what you mean: how can an IDE not "suit" Rails?
<snip interesting answer>
Interesting.. but largely personal preference [1]... what you see as "the primary benefits of conventional IDEs" is probably different to what others may list. Or maybe the using the term "IDE" is a red herring (as it may not mean the same to everyone), and really what I mean when I say IDE is "the editing software I use to write my apps". (to me, any software with features more then editing and saving plain-text files is starting on the road to IDE... some just "integrate" more features than others)
I'll turn the question around: what do you get out of using an IDE for Rails, in terms of features that a decent editor wouldn't provide?
I've already said - I want VCS and debugging integration.
NetBeans' Git plugin is fantastic. I just don't see it as fantastic enough to saddle myself with the rest of the IDE.
Whereas I think that the Mercurial plugin *is* fantastic enough to saddle myself with the rest of it... personal preference again!
BTW, even if NB is officially dumping Rails, couldn't you still use its Ruby support? And don't you think someone is likely to take up maintenance of the Rails tools?
Probably, and probably. To a large degree, I don't really use the "Rails" features anyway. It just seems disappointing to be at the end of life for a product that's been getting better and better. To know it is never likely to get better still was just a spur to me to look around again (after using Netbeans for a couple of years, I've been using Rubymine for the last three days, and have some observations that I shall post separately)
[1] I think we should all agree that we will *always* _know_ for certain that the choice we made for ourselves is the best.
We should also know we'll *never* convince someone that made a different choice that we are right and that our choice is better for them than their choice.
I refer to this Dilbert cartoon: http://bit.ly/hP9Bk1
Michael Pavling wrote in post #978365:
If you were my client, and you responded "because I want it" when I asked why a feature was of value to you, you wouldn't get the feature.
Now *that's* customer service!
It's excellent customer service. My customers get exactly what they want, and do not spend money on things they don't want.
Again, I think you're confusing needs and wants (and blurring it more with your interpretation of what they need).
I do understand the difference. I also understand that people think they want things that, when questioned, they cannot explain why they want nor provide any justification for. That's the same as saying "I want a pony": sure, you may want it, but is it worth spending time and effort on?
They can hardly "get what they *want*" if you also decide that they "wouldn't get the feature".
You missed my point. I ask my clients why they want something, so I can prioritize and understand how things fit together in their minds. I wouldn't be satisfied with receiving an answer of "because I want it" -- I'd dig deeper to see what the real reason was.
You shouldn't be satisfied with *giving* an answer of "because I want
it" either. It's a cheap shot, and, frankly, a cop-out. It tells me
that you haven't even bothered to think about why you want a particular
feature, you just want it. Pony.
Either way, it's academic, as how you manage your customer relations is up to you (but if they *really want* those features that you and I both know are pointless... if you don't want to do it, send them my way and I'll give them a quote for the work
You waste your clients' time and money on feature bloat?
The article you linked to is a red herring in this connection. If you want something, you should still be able to explain *why* you want it, not just "because I want it".
Getting into psychology now... sometimes people *can't* justify their wants (or they just like something more for no particular reason).
Can't justify their wants? Example?
(Note that I think "because it's ugly" is in many cases a valid justification.)
You often say a variation on this statement, and I never really understand what you mean: how can an IDE not "suit" Rails?
<snip interesting answer>
Interesting.. but largely personal preference [1]... what you see as "the primary benefits of conventional IDEs" is probably different to what others may list.
OK, now I'm getting upset -- not that you disagree with me, but at how you're choosing to argue. You asked me about why I think conventional IDEs don't suit Rails. I answered at some length. So far as I can tell, instead of actually responding to my answer, you simply handwaved it. If you disagree with what I wrote, please tell me what you disagree with.
In other words, if you think others may list different key features, then list some!
Or maybe the using the term "IDE" is a red herring (as it may not mean the same to everyone), and really what I mean when I say IDE is "the editing software I use to write my apps". (to me, any software with features more then editing and saving plain-text files is starting on the road to IDE... some just "integrate" more features than others)
I consider an IDE to start somewhere above a project-aware editor -- perhaps with things like build tools, project management, shell integration, Rake task invocation, and so on. (KomodoEdit actually has a very nice module that, by that definition, makes it an IDE for developing its own extension modules.)
I'll turn the question around: what do you get out of using an IDE for Rails, in terms of features that a decent editor wouldn't provide?
I've already said - I want VCS and debugging integration.
So do I, other things being equal. But those two features alone are not enough for me to incur the overhead of something like NetBeans.
NetBeans' Git plugin is fantastic. I just don't see it as fantastic enough to saddle myself with the rest of the IDE.
Whereas I think that the Mercurial plugin *is* fantastic enough to saddle myself with the rest of it... personal preference again!
Or the fact that Mercurial is harder to use than Git and needs one more
crutch.
Seriously, I wonder if you'd feel this way if you were using Git. GitX is a marvelous standalone GUI tool; I am not sure that anything comparable exists for Mercurial.
BTW, even if NB is officially dumping Rails, couldn't you still use its Ruby support? And don't you think someone is likely to take up maintenance of the Rails tools?
Probably, and probably. To a large degree, I don't really use the "Rails" features anyway.
THEN WHY THE HELL ARE YOU USING NETBEANS?!?
Perhaps I shouldn't have shouted that; it just seems to me that you're saying you want to use a big heavy IDE that you don't use 90% of the features of. How can this be sane?
It just seems disappointing to be at the end of life for a product that's been getting better and better.
How would you know? As you've pointed out, you don't even really use it.
To know it is never likely to get better
WTF? NB is actively looking for someone to take it over. Frankly, I'd trust a group of Rails enthusiasts to make it better more than I'd trust the notoriously unresponsive NB core team to do so.
still was just a spur to me to look around again (after using Netbeans for a couple of years, I've been using Rubymine for the last three days, and have some observations that I shall post separately)
Good. Then also please consider which of your "showstoppers" really have to be showstoppers.
[1] I think we should all agree that we will *always* _know_ for certain that the choice we made for ourselves is the best.
No. There are times when I am not sure.
We should also know we'll *never* convince someone that made a different choice that we are right and that our choice is better for them than their choice.
Again no. I have done this many times.
I refer to this Dilbert cartoon: http://bit.ly/hP9Bk1
Best,
I do understand the difference. I also understand that people think they want things that, when questioned, they cannot explain why they want nor provide any justification for. That's the same as saying "I want a pony": sure, you may want it, but is it worth spending time and effort on?
Again, just because people *can't* explain, doesn't make them wrong in their desire. It may make it harder (or impossible) to agree with (as a "rational" developer/programmer, with all the borderline-autistic traits we possess), but it doesn't mean some fundamental psychological operation isn't happening in their brains.
You waste your clients' time and money on feature bloat?
Yup - if that's what they want... even if I don't like it (I have produced some websites that I think look *awful* and/or have music playing in the background - because that's what *they* want despite my protestations). It's their money... I'd prefer to produce something I'd like too, but at the end of the day...
sometimes people *can't* justify their wants (or they just like something more for no particular reason).
Can't justify their wants? Example?
I can't explain why I want a girlfriend with certain physical characteristics (wanted... I'm married now!) - hair colour, height, skin-tone, etc. I know what I like, but I don't know why... given a choice, I'll prefer one over another.
(Note that I think "because it's ugly" is in many cases a valid justification.)
...so you won't accept "because" as an answer, but you will accept
"because I like/dislike it" - I think we're saying almost the same
thing, and just boring anyone else reading this. I think your argument
would be more distinct from mine if you wanted people to give you
checklists of *why* they like/dislike (which seems to be what you
*are* asking me re: IDEs!
Interesting.. but largely personal preference [1]... what you see as "the primary benefits of conventional IDEs" is probably different to what others may list.
OK, now I'm getting upset
I'm sorry - I certainly don't want to upset you; and I didn't disagree with you. I read your answer, and it was long, and I will have to read through it a few more times to fully digest it.
instead of actually responding to my answer, you simply handwaved it. If you disagree with what I wrote, please tell me what you disagree with.
I didn't disagree - I agree that the list of things you think make an IDE, and the things that list is good/bad for is the list you believe is correct. All I've said (ever said) is that other people may have other opinions, and they can make their own minds up. I asked for the clarification, because everytime someone mentions using an IDE in their Rails development, you say "DON'T USE IDEs FOR RAILS!" and accuse people of having marginal handicaps because of the tools they use.
I consider an IDE to start somewhere above a project-aware editor
so that's pretty similar territory to my definition then...
I'll turn the question around: what do you get out of using an IDE for Rails, in terms of features that a decent editor wouldn't provide?
I've already said - I want VCS and debugging integration.
So do I, other things being equal. But those two features alone are not enough for me to incur the overhead of something like NetBeans.
But they are for *me*
However, if a "decent editor" got integrated VCS (specifically Mercurial, but I'd consider switching to Git or something else if it was what was integrated), debugging, and (much lower priority for me, but my third-most-used feature) test running (and through the other features, test debugging) then I'd drop the bloaty Netbeans like a shot and jump to it. All the other features, whether I use them or not, are just "gravy".
NetBeans' Git plugin is fantastic. I just don't see it as fantastic enough to saddle myself with the rest of the IDE.
Whereas I think that the Mercurial plugin *is* fantastic enough to saddle myself with the rest of it... personal preference again!
Or the fact that Mercurial is harder to use than Git and needs one more crutch.
Strange. I don't think Mercurial is harder than Git - I don't think either of them are hard... I'm happy to use command line hg... I just prefer to have it in my editor (seeing what changes I have made to a line at a glance in the editor is valuable to me).
Seriously, I wonder if you'd feel this way if you were using Git. GitX is a marvelous standalone GUI tool; I am not sure that anything comparable exists for Mercurial.
I don't want standalone tools... I have the command line for that. I want it in my editor pane.
BTW, even if NB is officially dumping Rails, couldn't you still use its Ruby support? And don't you think someone is likely to take up maintenance of the Rails tools?
Probably, and probably. To a large degree, I don't really use the "Rails" features anyway.
THEN WHY THE HELL ARE YOU USING NETBEANS?!?
Because it integrates Mercurial and Debugging very nicely (have I not
mentioned that?
Perhaps I shouldn't have shouted that; it just seems to me that you're saying you want to use a big heavy IDE that you don't use 90% of the features of. How can this be sane?
What if it isn't sane (for you)? So what? How does that affect anyone apart from me? *I* am more productive because things suit me better - so it's not insane from my point of view.
As an aside (and a not ideal analogy): I drive a big bloaty 4x4 for
the 2-days a year the British weather may demand it... possibly not
sane on an average day. But as fate would have it, shortly before
Christmas I had to get to the hospital with a very ill baby daughter
at 2am in a blizzard - the ambulance couldn't get to us, but with
chains on, I could get to them.
I'm happy to bear the extra expense and bloat of this machine for the
features it offers me that I *can't* get elsewhere. Being able to get
out of my village in almost any conditions is a feature I'm not
willing to give up, and neither is debugging
It just seems disappointing to be at the end of life for a product that's been getting better and better.
How would you know? As you've pointed out, you don't even really use it.
*sigh* The product, Netbeans, has been getting better - the Mercurial integration is quite new, and the debugging has gotten simpler in newer releases. So what if I don't use the "crutches" of Rake integration, etc? Does that obviate my opinion of the product? (if so, since you use *none* of them, how would you know any about it... (I'm not trying to be snide; just saying we're on the same ground - if you try to undermine me with "how would you know?", you're falling too)
To know it is never likely to get better
WTF? NB is actively looking for someone to take it over. Frankly, I'd trust a group of Rails enthusiasts to make it better
When they find them...
still was just a spur to me to look around again (after using Netbeans for a couple of years, I've been using Rubymine for the last three days, and have some observations that I shall post separately)
Good. Then also please consider which of your "showstoppers" really have to be showstoppers.
Please stop asking me to stop liking what I like... it's me getting
upset now 
[1] I think we should all agree that we will *always* _know_ for certain that the choice we made for ourselves is the best.
No. There are times when I am not sure.
I don't believe you... I don't think *you* ever make a choice for
yourself that you think is worse than the alternatives you could have
had (and don't think about starting to wave altruism around
We should also know we'll *never* convince someone that made a different choice that we are right and that our choice is better for them than their choice.
Again no. I have done this many times.
hmmm... or maybe you just brow-beat them into submission rather than "convincing" them - is it the same thing? :-p
On the subject of standalone GUI tools, I think both Tower (http://www.git-tower.com/) and Gitti (http://www.gittiapp.com/) beat GitX hands down. However, I have to add that neither of those have been able to lure me away from using Git in the terminal. Creature of habit I think.
Best regards
Peter De Berdt
Peter De Berdt wrote in post #978401:
comparable exists for Mercurial.
On the subject of standalone GUI tools, I think both Tower (http://www.git-tower.com/ ) and Gitti (http://www.gittiapp.com/) beat GitX hands down.
I didn't know about these. I'll check them out.
However, I have to add that neither of those have been able to lure me away from using Git in the terminal. Creature of habit I think.
GitX is a supplement to command-line Git, not a replacement for it.
IMHO Git absolutely needs both a GUI tool and a command line. There is no good way to visualize complex branching structures without a GUI, and no good way to perform many of Git's more complex tasks graphically.
Best regards
Peter De Berdt
Best,
In that regard I wholeheartedly agree. I will also fire up one of the Git GUIs out there (or sometimes just fire up “gitk” from the terminal).
Best regards
Peter De Berdt
On the subject of standalone GUI tools, I think both Tower (http://www.git-tower.com/
) and Gitti (http://www.gittiapp.com/) beat GitX hands down.
just went to check these out… only to be sorely dissapointed that they are both mac-only
anything comparable for linux or windoze?
*random note: is proud that i got an old-ish laptop running ubuntu server & cygwin playing nicely together… so now i can have my cake, eat it & share with only the people i want to share with!
hex
I'm trying it but when I run redcar I have:
Redcar 0.10 ( java ) Error loading plugin: <Plugin swt 1.0 depends:[dep(core >0)] 0 files> cannot load Java class org.eclipse.swt.SWT ........
I'm using linux amd64 and jruby 1.5.6.
I've solved
Now redcar is my ide, I've deleted netbeans, aptana, kate, gedit, vim, etc.
I hope that textmate have breakpoint and debugging like netbeans
Netbeans rails support is discontinued? Ok, I've destroyed netbeans from my pc.
I hope that textmate have breakpoint and debugging like netbeans
Netbeans rails support is discontinued? Ok, I’ve destroyed netbeans from my pc. Eh? Why? It’s not like it’s going to just stop working
Yes but they said that rails support is discontinued and I don't like it. Try redcar.
Do you know how to open redcar with the last session? Every time I open redcar I must browse directories and reopen all my last files.