908 again unusable
Hi,
You know, I've said, that Aurora EAP version's stability was much lower
than one or two years ago. They got better, so the 3 latest 8xx builds
were quite usable for daily work. I opened 908, tried to auto-import a
class, bang! Tried to auto-send the exception, bang! I tried to run
optimize imports, bang! Keep up the good work!
Tom
请先登录再写评论。
I agree...for a different reason. It wont read my 896 project. It doesn't
produce any errors and the 896 project works (under 896).
896 is pretty solid. I've been using it for about a week or more now and it
does the job. I'm going back to that version.
So while you and I are struggling to use 908 for basic java functionality, I
am sure Aspect J works though I suspect only (a very vocal) 5% of the
development community uses it.
"Thomas Singer" <idea@regnis.de> wrote in message
news:bihraq$rid$1@is.intellij.net...
>
>
>
Well, I've seen dozens of "908 is useless" messages. But, at least for
me, its working great. Not a single problem. I should add however, that
I /always/ install into a new directory, and always tell idea to use the
installation for the options, cache, and system files. (Yes, I have to
wait a bit to reoptimize. But then I don't end up with a "useless" build.)
Nor I, and this is what I find particularly puzzling. I know IntelliJ is trying to cater to the magazine reviewers with drivel like the UI designer, but when releasing software, particularly to developers, the largest market with release cycles is of course customers who already have the product and are looking to upgrade. I have been constantly dismayed by the feature list for Aurora and frankly have little interest in upgrading to Aurora when it's finally released. HotSwap and the jUnit integration are the only two new features that I have any interest in and by no means is that enough to sway me.
Kirk, happy in Ariadnaland...
Same with me. 908 is usable except for occasionally internal errors (rarely), which is normal for prerelease stuff. I install each new EAP build in a new directory, let Idea store the settings in the install directory and not under user's home and choose the option to import the settings from a previous installation. With all this, except for Build 859, I never had problems, which prevented me using EAP builds regularly. (My workstation is a Windows 2000 Pro machine).
With respect of new features, there are already a plenty of new ones, which let me prefer the new build over our Idea 3.0.5 version, most notably the improved CVS integration.
Thomas
Nope, even a completely fresh install with system- & config-dir in <idea-home>
and nothing but default plugins (tomcat, uiDesigner, etc.) does not work for me.
I get the error http://www.intellij.net/tracker/idea/viewSCR?publicId=15392
from the first character I type in any file.
Sascha
Erb wrote:
>
>
>
>
>
>
"Sascha Weinreuter" <sascha.weinreuter@NOSPAM-cit.de> a écrit dans le
message de news:bikcn1$pct$1@is.intellij.net...
<idea-home>
for me.
http://www.intellij.net/tracker/idea/viewSCR?publicId=15392
>
By the way, is it a zip install or an installer install ?
Maybe should we try the installer ?
Guillaume
Kirk Woll wrote:
>> The features themselves of course have almost no appeal to me.
Some of the features I like, just off the top of my head:
Generics is supported, and has helped me find a couple of potential
bugs in my current project.
The new inspection API paved the way for InspectionGadgets, which is
very useful
Improvements in the diff window
Ctrl-shift-F7 for try/catch/throws
Many new highlighting options
Line wrapping options for reformatting
Lots and lots of minor improvements here and there
I don't have any use for AspectJ or the GUI designer, but as an IDE
grows I suppose there will always be some things you don't have any use
for personally.
have any interest in...
Hmm. Have you looked over all 324 (for the moment)?
--
Best regards,
Maxim Shafirov
JetBrains, Inc / IntelliJ Software
http://www.intellij.com
"Develop with pleasure!"
And what would be the features you'd like to see in Aurora (beside
scripting macros, I do remember). Tracker ids please. How many votes are
there?
--
Best regards,
Maxim Shafirov
JetBrains, Inc / IntelliJ Software
http://www.intellij.com
"Develop with pleasure!"
"Maxim Shafirov (JetBrains)" <max@intellij.net> wrote in message
news:bikreo$kdb$1@is.intellij.net...
>
>
>
>
>
>
>
I installed the .zip version. Never used the installer for an eap version, never
had problems with the .zip install.
Surprisingly, the error while typing went away (for now) after I simply restarted
IDEA a few times. I even copied my previous configuration and plugins it still
works. I'm not heavily coding right now, so I'm not sure how it behaves when I am.
Guillaume Laforge wrote:
>
I am really fundamentally opposed to the voting procedure. And before I get some spiel about castrating the democratic process I'd like to point out that the voting is so easily prone to abuse that it's really nothing more than a way for people to anonymously express favor for an idea without qualifying why.
That said, the issues that I care about are the same that they've always been. Make the Open-API more open and also support it more. I'm pretty sure I've never once had a response from you guys in the Open-API topic despite a number of questions. Refactorings should be opened up. The debugger should be opened up. The enhanced debugging support you add for collections should be implemented via your own Open API so I can do something similar for my own data structures.
Other miscellaneous issues...
14277 (PSI API with no hedging)
3842 (User defined live template functions)
10355 (Ordering editor tabs)
But of course, this is really missing the point. My point was that in past versions, the forthcoming Idea release would be packed with so many new features that were useful to me (Coding by intention being my favorite new feature) I couldn't even entertain the notion of not upgrading. Maybe Idea has plateaued. I don't know. All I know is that my passion for the latest and greatest when it comes to Idea is no longer a burning desire.
I personally would be thrilled to have back the features that the 70x builds have. Multiple output paths (I whined and whined about wanting this in the last eap, and was happy for months when it went in, but now it's gone again!). Decent J2EE support (which is upcoming), and better javadocs for the openAPI (eg, when something is deprecated, give a hint as to what should be used instead!).
It'd also be nice if the memory leaks were adressed, currently when working with multiple projects, you'll have to restart IDEA after about 5-6 switches. It's not sexy or glamorous, but it's the small things like that that matter to me. The big stuff I'm happy putting my faith in whatever the IDEA folks come up with (vs what the idiot masses come up with, which is more often than not stupid shit like maven and aspectJ)
I have a completely different opionion. The new CVS integration alone is worth it for me. Granted, if you're not using CVS than that doesn't mean much to you but for me that is a very big issue. That's probably the biggest thing. There are a ton of small things that add up to one large thing but I won't go into them because I don't have that kind of time. I would like to see some WebLogic integration (at least from the perspective of updating weblogic specific ejb xml files) but other than that I'm quite happy with what I'm seeing in Aurora.
>>"The big stuff I'm happy putting my faith in whatever the IDEA folks come
up with (vs what the idiot masses come up with, which is more often than not
stupid shit like maven and aspectJ)"
You dared to say what I've been thinking for a while now. Every man hour
devoted to AspectJ takes away from time devoted to stuff most of us use.
For example, I have a issues with taglib resolution. Others do too.
Workarounds have been posted, but I would estimate that far more people use
taglibs than AspectJ.
Evgeny started a riot with one of his posts on Jetbrains testing, but I'll
come at it from a different and more politically correct angle...I wish
Jetbrains would focus on the basics first. Whether taglibs is considered
basic is up for opinion, but AspectJ is clearly not. Their decision to
devote resources to that is seemingly political. It doesn't stem from a
massive grass roots demand for it. A quick search on Amazon shows exactly 3
books on Aspect J -- all of them very recent -- meaning enough people
haven't had a chance to learn about and digest this new idea. So somebody
in the EAP is getting to play with join points (whatever that is) while I
have to deal with pain in the ass workarounds to get my taglibs (an idea
that's been around for a few years) to resolve.