Demetra - code-name for what version?
What version is this the code-name for?
I've just been informed that some fairly critical bugs regarding J2EE deployment will be fixed in the final version of Demetra, and I am under the rather worrying impression that this was the code-name for version 6.0.
请先登录再写评论。
Update: IDEA-5470 reproduced, fixed.
-
Maxim Shafirov
JetBrains, Inc
http://www.jetbrains.com
"Develop with pleasure!"
Damn I need to list out the bugs that are affecting me at the moment, this has been a great response rate :)
First of all: Thanks a lot for being so responsive.
I really had the feeling that recently it became very hard to get noticed
at all.
So after all it were no false alarms at all.
I agree that many of the issues are with functionality that got extended
due to new features like Java 5 support.
Overall the impression that sticks with users is that functionality (like
formatting) got worse, though.
I can imagine that its is tough to keep up the very high quality you had,
esp. when stuff gets so complicated as formatting all the new JDK 5 constructs.
But even simpler things have quality issues, e.g. do a Jira search for
"subversion AND browser".
> Well, it took me 1.5 hours to get through these 9 with 2 false alarms.
> And we usually have 10 - 50 new JIRA items a day not to mention
> newsgroup posts. And the average true/false alarm is much worse most of
> the time.
That sounds like a confirmation of my opinion. If you get so many JIRA items,
isn't that an alarming sign that quality needs to be improved?
And with the rate you mention one experienced developer should be able to
handle all these items.
I agree though that false alarms are a problem.
For a start I propose to give more instructions and make these instructions
more visible: For example, if a bug is entered maybe you can fill in the
text area with something like this:
"Steps to reproduce:
-
-
Expected Behaviour:
-
Actual Behaviour:"
Another idea is to show a list of known, common issues before allowing
anybody to file new issues (that damned java font problem for example).
Then there is no prominent link to the FAQ on the way to jira.
Any maybe give some better instructions how to search in jira - if I
haven't been using jira before I would not have known that "svn AND timeout"
works, but "svn and timeout" does not...
If all of this does not help, maybe you really have to restrict the EAP
and make it accessible only to selected users. But IMHO that's a last
ressort as you will loose valuable feedback.
A good example that shows how user feedback can improve the product is
Idea's builtin thumbnail browser. I and some others gave lots of feedback
during the EAP, and Alexey was very responsive in dealing with issues.
As a result the browser is nicely integrated to the rest of idea, has a
slick UI and is more stable.
If there were more feedback on the subversion browser it probaby would have
been much better in final Idea 5.0 version.
> Annotation formatting: Are those annotations public? Can I get those to
> play with? Anyway that's not a bug but missing feature, which haven't
> been requested during 5.0 EAP.
Those are standard EJB3 persistence annotations. You can get them from
hibernate.org for example.
Maxim Shafirov (JetBrains) wrote:
Ok, I tried to reproduce and it's not that hard. At least here is one
behaviour that is wrong:
Set the options to "Tab limit: 4" and "Close less frquently used file".
Create five classes all in the same pakage: A, B, C, D, E
Close all editor tabs.
Open (in that sequence) by double clicking on the tree nodes in the project
view: A, B, C, D
Now click on the editor tab for A.
You can even edit A if you like.
Double click on E in the project tree.
=> Tab for A is closed.
Expected: B's tab gets closed.
Maxim Shafirov (JetBrains) wrote:
I went through my watchlist. 20 bugs triaged in < 1 hour. I didn't touch
the feature requests. Some of them have been open since 2003 and yet
there is not a single comment from JB. And all of these bugs had easily
reproducible test cases attached.
It would be nice to feel that the effort we put in to boiling these bugs
down to small test cases (I estimate at an average of 15 minutes per
bug) would be reciprocated. I can understand your frustration at having
to deal with bug reports like "Idea doesn't work properly" but can you
understand our frustration when clear, professional bug reports seem to
be ignored completely?
R
Fixed, thanks for the "steps to reproduce".
--
Best regards,
Eugene Zhuravlev
Software Developer
JetBrains Inc.
http://www.jetbrains.com
"Develop with pleasure!"
Robert,
I always appreciated your requests, and if not all of them have been
processed yet,
that is simply because I don't have enough time to look at.
Eugene.
"Robert Gibson" <robbie_usenet@yahoo.co.uk> wrote in message
news:djnfq9$7hs$1@is.intellij.net...
>> Well, it took me 1.5 hours to get through these 9 with 2 false alarms.
>
>
This demonstrates that an untold and likely huge number of identified bugs and features were never addressed and often dismissed. We are all happy that a handful of Stephen's issues were finally addressed, but it does nothing to change the process whereby the probably thousands of other quality issues filed in the past and thousands more quality issues to be filed in the future will languish and IDEA will suffer. If IDEA is to reverse the quality slide, process has to change in the EAP so that issues are given proper attention.
Jon
There was talk of having a select group of people (by area of expertise)
perform triage duties. Sounds like we still need something like that.
Jon Steelman wrote:
>
No problems, it's always nice to get feedback, even if it's "go away and
stop bothering me with twisted testcases"!
R
And it's still broken in 3536 and 3542. Why can't this get fixed?
And so I'll repeat. This is a concrete example of the problems with the 5.0 release. A bug (NullPointerException) has been in the product for many builds, and either the will to fix, or the ability to fix is NOT present.
Glenn McGregor wrote:
>> Quite strange thing, kinda "impossible"... What
>> plugins do you have installed?
>> What kind of file you've been editing in that CVS
>> diff window? xml, java?
>>
>> -
>> Maxim Shafirov
>> JetBrains, Inc
To be fair, the number of users who experience this seems to be fairly limited and it's
not a show-stopper with lost data or such. A bug that cannot be easily reproduced and that
should not be even possible with "normal" flow of execution is hard to track down. I
wouldn't attribute this to a lack of will, but I wold love to see some improved
communication on the issue as well.
Glenn, are you using the Customization Feature or maybe the "CVS bar" plugin? As I said I
have a placed the "diff" button in the main toolbar using the customization and I suspect
that this may be related to the problem, which would also explain why only few people
actually see the problem.
Sascha
I've assigned a key-stroke sequence to 'cvs commit'.
I'm not using the "CVS bar" plugin.
It (the NPE) might be more likely to occur if the file I am examining for diffs is long. (Not really sure, but seems that way).
It doesn't occur unless I double-click on a file in the commit dialog. So it would seem unlikely to be related to how I got there.
Other people are getting the NPE, as I see additional entries in the tracker entry every day.
Glenn
Glenn McGregor wrote:
Well, I've been able to reproduce both exceptions below even with my customization scheme
set to "default". I didn't bother to completely delete it though. Indeed, the size of the
files does matter, perhaps also the amount/kind of changes. I can't do it reliably, but in
about 5-10% of all attempts it throws the first exception when the file is modified and
the diff is invoked quickly after making the change.
The second exception usually happens when the code to be diff'ed is syntactically
incorrect and hasn't been fully reparsed yet when invoking the diff.
Yep, just noticed that. Unfortunately, it doesn't seem like anybody is handling exceptions
these days, probably because the old tracker has been abandoned.
http://www.intellij.net/tracker/idea/viewSCR?publicId=56872
http://www.intellij.net/tracker/idea/viewSCR?publicId=56854
I've created a JIRA issue about it:
http://www.jetbrains.net/jira/browse/IDEA-5823
Maybe this will get a bit more attention.
Sascha
Hello Sascha,
SW> Yep, just noticed that. Unfortunately, it doesn't seem like anybody
SW> is handling exceptions these days, probably because the old tracker
SW> has been abandoned.
It's not abandoned - it's still used as the back-end for the new exception
analyzer. You won't see any tracker requests getting fixed, though - all
the real activity is reported as JIRA requests created from exception reports
and then resolved.
Indeed, reviewing new exception reports is not performed on a regular basis
right now - hope this will resume soon...
--
Dmitry Jemerov
Software Developer
JetBrains, Inc.
http://www.jetbrains.com
"Develop with pleasure!"
Dmitry Jemerov (JetBrains) wrote:
Ha, I knew it, someone's out there :)
While this is good to know, please try to set this up in a way that it's possible to
receive feedback on user-reported exceptions. It's honestly the most frustrating thing
when reporting stuff that doesn't even seem to be looked at. It might not matter much for
the "casual" error/request submitter, but I like to see that the effort that I spend on
reproducing, describing and reporting an issue is at least acknowledged in a way and not
going to waste.
Cheers,
Sascha