Hacker News Reader

Comment by cromulent | Hacker News Reader

...according to the marketing manager for the product ;)

With all respect, I don't I understand your point. FTR, I'm also a developer in past and in present. If you have an argument in regard to actual points shared in my article, please share it with others.

I don't think it's improper for it to be pointed out that someone who works as the marketing manager for a product may have a slight bias towards talking up said product's benefits and downplaying its limitations.

It's just a fact, not a slight.

(And I say this as someone that happens to love just about everything this company puts out!)

I think Andrey took it not as "he is a marketing manager for JetBrains, so of course there is a bias", but rather: "he is just a marketing manager, and not a developer, so what does he know?" And with that line of thinking, the follow up question makes perfect sense. He's not just a marketing manager.

When there's two possible ways to read something, and one of them makes all kinds of sense and one of them doesn't, I think parsimony suggests we should believe the one that makes sense.

Sure, but that assumes you read it two possible ways.

And my interpretation makes perfect sense, especially if you consider the individual in question might not be a native English speaker. Seriously, I think people complaining about this are reading far too much into this. Chill.

He's a marketing manager who responds with an argumentative tone to someone pointing out he is a marketing manager. There's a thing you don't see every day.

Argumentative tone? I didn't read that. Couple that with the fact that he might not be a native English speaker, I think it's a bit much to imply a tone.

An argument against the article was specifically solicited, even though it was orthogonal to the OP's point. That's not an implication, it's a statement. I'm not sure how else to read such a thing, frankly, native speaker or otherwise.

Edit: and not that I'm telling the jetbrains folks how to do their job, but a marketing manager who has communication difficulties in a given language generally isn't given the responsibility to make public company statements in that language.

Basically your objections aren't holding much water.

Considering your other comments, I see why you would think ill intent of others.

Reasons I prefer eclipse:

The git "commit and push" workflow is less ergonomic than eclipse's - you have to wait for it to perform the commit and then ok another dialogue to do the push, rather than just hitting "commit and push" and letting it happen in the background.

There's no obvious way to search within a directory. I don't want to search everywhere which is the only option I see in ctrl-A. I want to be able to right click or similar on a directory in my project and search just in that directory.

I can't tell how to reorganize the UI. In Eclipse I have the Window > Views menu. In IntelliJ panels seem to appear and disappear arbitrarily as I invoke particular actions and I have no idea what rules it follows or how to control it. Edit: aha, there's a weird square in the very bottom left corner: having clicked that the tabs for the panels have appeared which makes it much easier.

No save button means I can't tell the IDE when not to try to compile. This is important when working on complex code (particularly Scala) and I know compiling it is going to take a few seconds of sluggishness.

The Scala presentation compiler (admittedly not really the subject of this article) has too many spurious errors (most prominently one I reported as SCL-9700).

Most importantly of all, continuous building simply isn't reliable enough. In eclipse, if I've hit save I can tell when the build has finished, and if the build has finished and there are no problems in the problems view then I am 100% confident that my code actually compiles. This simply isn't the case in IntelliJ (e.g. after renaming a class it won't necessarily complete the rebuild and so errors don't always show up until after you deliberately make).

> There's no obvious way to search within a directory. I don't want to search everywhere which is the only option I see in ctrl-A. I want to be able to right click or similar on a directory in my project and search just in that directory.

Right-click on folder, "Find in path..." does exactly that, doesn't it?

It's not in the right-click menu (at least on the bar at the top, and I don't know how to make the panel on the left appear), but that does it, thanks! I assumed it would be called "search" since the startup background talks about "search", didn't think to look for "find".

On OS X it's cmd + 1, I'd guess ctrl + 1 for Windows? Or click the box in the bottom left which brings up panes for everything. You can also bind the keymap however you like, to make these panes accessible in commands you can remember

Alt-1 on Windows/Mac.

I switch between Mac (work) and Linux (home) and trying to remember which Linux shortcuts move onto Cmd is turning out to be a struggle.

> There's no obvious way to search within a directory.

On OSX at least: CMD+SHIFT+F brings up the Find dialog. Scope defaults to Whole Project (ALT+H) but you can change it to Directory (ALT+O) which seems to default to the directory of the currently selected file. (Or maybe Scope actually defaults to the last one you used.)

I'm wondering what version of IDEA you tried that wasn't this way?

I don't run into compiler issues too often, but I also don't use shapeless or scalaz so I wouldn't be surprised if more complex type signatures give IDEA trouble. I have noticed a drastic improvement with the Scala support since IDEA 13 or so.

I use JetBrains Webstorm, and on that hitting CMD+SHFT+F creates a search local to the selected folder in the directory tree. Seems pretty intuitive to me, but finding files themselves, as opposed to string constants and the like, is actually the more difficult thing Ive yet to figure out.

Ctrl-Shift-N gets you an "open file" dialogue that does slightly fancy filtering/searching (e.g. you can put a substring of the name and it'll find it).

This was 15. I didn't think to look for "find" because "search" is advertised on the default background screen and I expected it to use the same word - per the cousin comment it's working now.

Sadly the Scala support is still quite unreliable at least at the presentation level - lots of spurious red in the editor - maybe better than it was. I use both shapeless and scalaz extensively.

You can also disable auto save...

Thank you for jumping in with this, because I completely agree that a list of advantages AND disadvantages, collected by people with both sets of preferences is more useful than a list of advantages written by the marketing manager of the product (although that is still quite useful).

That being said, as someone who uses IntelliJ frequently, I wanted to respond to a couple of your points:

> No obvious way to search within a directory

In the "project" pane (the place where it shows the folder structure of your files) just right-click on any folder and select "Find in Path" (Ctrl-Shift-F). Definitely an important feature, I use this all the time.

> I can't tell how to reorganize the UI.

You solved this yourself. ("aha, there's a weird square in the very bottom left corner...")

> No save button means I can't tell the IDE when not to try to compile [...] compiling it is going to take a few seconds of sluggishness

Interesting. My experience has been that the compiling just takes up background, low priority threads and doesn't interfere with my work -- I don't even notice it getting behind. I have never tried it with Scala code (which does a LOT more during the compile phase) so maybe it's more noticeable when doing that.

> continuous building simply isn't reliable enough [...] (e.g. after renaming a class it won't necessarily complete the rebuild...)

Curious. I never seem to encounter problems like this. I wonder why. Does the ability to trigger a make when you want it mitigate this somewhat?

> You solved this yourself. ("aha, there's a weird square in the very bottom left corner...")

Indeed. I think this whole thread is showing that the UI has a lot of surprising/nondiscoverable aspects though. It would be better if all the functionality was accessible via the main menus as a fallback.

> Does the ability to trigger a make when you want it mitigate this somewhat?

Ironically it would probably be better to get myself into the right habit if the continuous building were less reliable - it goes wrong rarely enough that I feel dumb insisting on a make every time, but if I start trusting it I get tripped up.

The key combination for make is relatively fiddly. Maybe I should try binding it to ctrl-S.

> you have to wait for it to perform the commit and then ok another dialogue to do the push, rather than just hitting "commit and push" and letting it happen in the background.

There is a "commit and push" button. You have to hover your mouse over the Commit button in the Commit Changes window to see it.

Yeah but even if you do that it opens a dialogue for the push.

I think that the point is that it would be a good practice to disclose that. In no way does it invalidate what you have to say (and by the way, I really liked having the nicely-written presentation of these features; this is something I'm bookmarking to show to people), but disclosing the potential conflict of interest makes it clear that you have nothing to hide.

> I think that the point is that it would be a good practice to disclose that

Did you check the articles' URL? It's blog.jetbrains.com (i.e. an official JetBrains blog). A disclosure would be redundant.

cromulent's point is that you made an editorial claim of "best" in a submission title, against the policies of HN: https://news.ycombinator.com/newsguidelines.html

Then here is something I am very happy with Netbeans but bought a license a couple of years ago.

What really really turned me off was when I found and reported a really annoying bug only to be told to buy the not yet released(IIRC) upgrade.

Reply on news.ycombinator.com