Old RFEs - Java SE (Archived)

What about having a look at the most popular RFEs? I think, that this can be a very useful source for finding, what developers want. Look at http://bugs.sun.com:80/bugdatabase/view_bug.do?bug_id=4691932 or  http://bugs.sun.com:80/bugdatabase/view_bug.do?bug_id=4093687 - there is a request for a small enhancement  , that would solve some problems. This enhancement caused no feedback from Sun since 1997 (!). I think, that the best place to begin are this RFEs.

The RFEs and Bugs from top-25 lists are considered first for the new major release. They all are being considered for Mustang.
You must also understand, that there is another thing that is very important - the feature or bug must be really useful for millions of developers. So votes without justifications are not enough - post your justifications on Bug Parade or forums, participate in JCP.

Related

Would be good with some talk back from the Mustang developers

Some "what do you mean", "this is crap", "this would be very hard to implement", "this sounds good", "Concider it added" (my favourite one) - comments by actual Sun Mustang developers would make us posters feel more "interacted with". :)
Cheers,
Mikael
Sure, the forum just started so it didn't attract significant amount of developers yet. I can't promise for other than desktop areas, but you may at least be sure that your ideas will be heard and won't live in the void.
As for "Consider it added" - there is a big distance between an idea and a feature, and you know the process - file an RFE, vote for it, justify why it is important for millions of developers - and you may "consider it added" :)
Hi, are you with Sun?  It's hard to tell who are and who are not with Sun. They (you) should have a _Sun in your nick name or something.
Keep up the good work anyway. :)
Cheers,
Mikael Grev
ps. I'm in a very boring phase in my project a.t.m so I might just post a little too much, as an escape... ;)
I am "with Sun"
Denismo:  I always appreaciate when new blood starts new initiatives.  But doubtfull as ever, I have to ask how much of a corporate backing you have?  Are the suggestions here
a) going to be read by actual developers
b) their managers as well
c) when both like it how likely is it they can convince management to actually follow the asked path?
See;  java.net did not bring around much more interaction between developers and the community; I surely doubt any technological directions were changed.  Why should we feel this forum will?
I have asked several managers, tech leads and engineers to watch this forum for ideas and comment on them.
I am AWT Tech Lead, so you may at least expect your ideas to be heard in AWT/Swing teams.
> I have asked several managers, tech leads and
> engineers to watch this forum for ideas and comment
> on them.
> I am AWT Tech Lead, so you may at least expect your
> ideas to be heard in AWT/Swing teams.
Thank you for responding to my question in a fair and open manner.  I do notice that you only answered part of my question;  perhaps you missed it (I should not ask that many questions in one post..)
Management obviously thinks open communication is in order; the question that remains is how much the questions and ideas asked here _can_ be followed up.  Are issues brought up here a priority for management?
I feel very unsatisfied with the amount of public-relations on javadesktop.org in contrast to the actual results in the latest Java release.
I am sure you understand that no single source of information for us can be a priority. We consider them all together - bugs, rfes, votes, JDC comments, forums, feedback  messages, JavaOne interactions, customer problems, personal relations, JSRs, etc. - this all builds up a huge source of objective information for consideration.

Are the JSRs working for the 'public'?

Part of the JCP/JSR review process is a 'public' draft, where anyone can comment on it. Comments on this are then supposed to lead to possible changes in the reference implementation followed up by a final draft and a final acceptance.
1.5 (tiger) is supposed to be released this summer. And JSR 201 which includes autoboxing and a lot of other stuff ( http://jcp.org/en/jsr/detail?id=201 ) is in that. There is a beta (or alpha) program right now for 1.5. Yet 201 is not in the review stage yet. That isn't too unexpected given that they have to work the kinks out. I suspect it just isn't done yet.
However, it does make me wonder what point the 'public review' serves. I doubt that by the time it is released that any comment, other than perhaps documentation changes, will be incorporated. It simply will be too late.
I have reason to believe that at least one other JSR will end up being handled in the same way.
Does anyone else have reason to believe that comments received during a public review of any JSR in 1.5 would result in either a major or even minor change?
I've never been sure about whether the JCP's are there to keep the punters happy or to actually take
comments.
However it is possible to contribute to the 201. There are email addresses for comments. Many of the
issues have been discussed to death on specialist forums and interest groups. (I have no idea if this
made any difference, but then the JCP does not say that they will pay attention to what you say)
I know a lot of the tecnicalities of enhancement proposals are worked out by private experts groups.
(Generaly consisting of a few invited experts and a larger number of business subscribers (who also
have interests in the outcome of the JCP/JSR)).
Basically, as far as I can tell, it is NOT a democracy. And perhaps that is a good thing (It doesn't seem
to have done much harm yet although I have some doubts about a few of the 1.5 enhancements).
matfud 
Yes, and I could join as a member if I wanted. I am sure I could have some impact earlier in the process if I did that.
But I am specifically looking at the 'public review' part of the process.
Basically, as far as I can tell, it is NOT a democracy.And actually it does look pretty much like a democracy right now - a representative democracy but still a democracy. And looking at the process I suspect it is more democratic than say - ANSI (which requires membership fees and quite a bit of money to 'participate' in one of the standards committees.)
There are several individuals (rather than company members) on the executive committee which suggests that all it costs is your time.
To be honest, I am not sure. I considered joining the JAI experts group (A long while back). It did not
appear to be free. At the very least it requires you have funding available to attend meetings and work on
the proposals (which tends to limit contributors to those working for supportive companies). That may
have just been JAI though.
JSR's are evaluated by SUN as to whether they will provide market share/utility/cost/detrement/whatever for SUN. If
it does not satisfy thier criteria it does not happen. (Many JSR's are additional/optional packages so they don't
generally have these problems). Represenatives and proposals from larger companies (who
are heavily invested in java) do tend to get more attention (which is not surprising) especially when it comes
to changes to critical parts of java.
All in all I've found that if you have a good point then it will be considered. However, it is often the case
that it is a good point which has already been made.
I don't know why 1.5 additions have not been opened for public comment. (Are you sure they haven't
already?)
matfud 
To be honest, I am not sure. I considered joining the
JAI experts group (A long while back). It did not
appear to be free. At the very least it requires you
have funding available to attend meetings and work on
the proposals (which tends to limit contributors to
those working for supportive companies). That may
have just been JAI though.In the past it cost - I looked into joining as well.
What I looked at a couple of days ago says there is no cost to join.
I don't know if they have on-site meetings. I figured most of it was electronic.
>
JSR's are evaluated by SUN as to whether they will
provide market share/utility/cost/detrement/whatever
for SUN. If
it does not satisfy thier criteria it does not happen.
(Many JSR's are additional/optional packages so they
don't
generally have these problems). Represenatives and
proposals from larger companies (who
are heavily invested in java) do tend to get more
attention (which is not surprising) especially when it
comes
to changes to critical parts of java.Interesting point.
The JSR process does not include that (or at least it specifically seems to not be that way) but that doesn't mean that Sun has to implement any JSR. However, as far as I know they have implemented everyone that impacted the JVM/API directly.
>
All in all I've found that if you have a good point
then it will be considered. However, it is often the
case
that it is a good point which has already been made.
I don't know why 1.5 additions have not been opened
for public comment. (Are you sure they haven't
already?)I am sure that the one I listed has not.

public static void main(String[] args) { /* again */ }

My feedback, sent via the link on the left. I'd be interested to see the responses of others. If you have the time and inclination, perhaps you could send me a copy by email so that it survives the inevitable ThreadDeathException
---
To whoever it may concern,
"Upgrades" to the Sun Java forums are a byword for incompetence amongst the regular contributors. If you look at the latest fracas, you will find that the loudest complaints were made by those who contribute the most to the technical discussions at hand. This is no coincidence. Those who invest most in the community are the most affronted when it is abused by the powers that be.
The problem has several parts, but I think it would be fair to group them into liaison problems, and technical problems. In view of this, I would appreciate it if someone in a position to speak for Sun on these issues would take the time to respond, publicly (i.e. in the Forum Maintenance section) on the following points:
Liaison:
Why do Sun not communicate more fully with community members before making sweeping changes? Why is feedback from the community so often ignored? If feedback from the community is not ignored, why do you not make the aggregate feedback public?
A case in point, natuarally, is the recent upgrade. The community were "polled" for the prefered cosmetic re-design. No option was given for NO cosmetic re-design. Many forum users responded via the Feedback link to complain that no design was wanted. And yet the design was changed anyway. Were most votes in favour of a re-design? Is this still true if feedback against is counted as a "vote" against? Was the most "popular" design the one ultimately selected?
Why do Sun not appoint moderators from the user community? At one point I received feedback from Sun that there were "legal" difficulties that were being looked into. This was several years ago, and yet nothing has apparently been done about it. Sun is a large company with many lawyers; surely they can achieve something so trivial? If the Javaranch forums can manage a community moderation scheme, one would think that a technology company as large as Sun could solve the same problem.
Why will Sun not accept that the natural place for feedback on the Sun Java Forums is in the forums. The feedback link has its place for raising some delicate issues. The notable reasons not to use this approach are:
(a) Your personnel do not, can not, or will not, monitor a feedback forum on even a daily or weekly basis to monitor for problems. This is simply not acceptable. If it is not worth allocating resources to this task, then the forums are not worth maintaining.
(b) You do not wish the issues raised to be made public. This is not acceptable. The forums are used by the public, and they will raise and discuss issues in public. If you prevent them from doing so in an environment in which you have editorial control, then they will do so elsewhere. Besides, there are advantages to having technically astute developers discuss the possible causes of technical problems. And surely it helps if the bugs are catalogued once, instead of being reported numerous independent times?
Technical
Why do Sun not block persistent offenders at the network level? The user commonly known as "Goldie" and currently using the "Stoic" account has been trolling the forums for more than a year. He has had tens, if not hundreds of accounts deleted, and hundreds if not thousands of posts deleted. Clearly deleting forum entities has no effect - even the signup email address is disregarded, so it is trivial to create a new account. Please, consider instating confirmation emails, OR take advantage of Jive's built in IP blocking facility, OR explain why you won't take one of these approaches.
Why do Sun not manage releases better? Releases should be managed on three tiers of hardware: Test, for establishing that the application is capable of working, Staging for replicating the production environment, and Production which is NEVER touched except as an application is rolled out from staging to production. Failures on a release should ALWAYS be rolled back IMMEDIATELY to the last known good version. Bug fixing should ONLY be carried out on the Test environments; never on Staging or Production. Regression suites should be used to ensure that fixed bugs do not reoccur. Usability testing should be used to ensure that enhancements really are enhancements. Sun Microsystems are the leader in the field of large software management and roll out, and yet the Sun Forums are managed less professionally than their amateur competition.
Conclusion
The above are the questions we would all like to see answered. The Sun Java Forums are a magnificent attractive force towards the Java technologies. Hundreds of new Java developers post questions every day. Please, don't waste the good will that the forums generate by alienating the very users who make them a worthwhile resource. Or if you intend not to manage them, please terminate them and let us concentrate our efforts where they will be appreciated. 
While this thread may not be long for this world, it's certainly well-written, and likely to be more well-received than the more strongly-worded flames.
~ 
likely to be more well-received than the
more strongly-worded flames.Oh I sent them too ;-)
You know, I actually kind of doubt it. The cycle goes:
1. Mismanaged change
2. Developers complain vociferously
3. Sun promise to communicate more/better/etc.
4. Initial promising signs
5. Sun's commitment to (3) reduces to a murmer
6. Repeat.
One possibility is that the people who matter (management) never see the problems here. Last time I tried to get the attention of anyone who I thought might be able to help us.
I sent appropriately mildly phrased emails to a dozen or so prominent Sun Java personalities. A few replied with encouragement and promised to do what they could. There seems to have been some reaction from the Sun people (I forget what specifically), and then it all tailed off and we're back where we started.
I really would prefer them to shut down the forums, rather than keep them going like this. I think maybe it would be better to make a contribution somewhere where that was valued. But as I say, at the moment I still marginally prefer the Sun forums to the competition - probably just because I'm more used to them. Another mess up like this, though...
Dave. 
Considering all the forum software out there that is far superior to this you'd think they'd learn something from it. It's not like they're sailing into uncharted territory here. 
> Considering all the forum software out there that is
far superior to this you'd think they'd learn
something from it. It's not like they're sailing
into uncharted territory here.
As I understand it, this is powered by Jive forum software, and I don't believe it's written by Sun. 
As I understand it, this is powered by Jive forum
software, and I don't believe it's written by Sun.I know and while I don't particularly like Jive that doesn't change the fact that it's (presumably) Sun employees butchering it and doing stupid things. 
Well said Dave.
As I understand it, this is powered by Jive forum
software, and I don't believe it's written by Sun.If the logo in the corner is to be believed. ;-)
I wonder if they are using a very old release, or highly customized version?
[url #" style="display: block; background-image: url('http://img76.imageshack.us/img76/7452/42uh.gif'); height: 20px; width: 20px]
Ohh, that cat is out the box.
Message was edited by:
mlk :( 
> Ohh, that cat is out the box.
Yep; I reported it a while back. Hopefully it'll do some good... 
> Ohh, that cat is out the box.
Yep; I reported it a while back. Hopefully it'll do
some good...Mm, can't edit posts after a reply. That is better, less fun but better.
Maybe the JS bugs will go when they change the colour scheme to include pink and a javascript based mouse trails thingie?

Proposal: Rehaul of BugParade

Hi,
I think Tiger is a great release, if not for technical reasons then for social reasons. Sun has finally begun deformalizing some of its communication channels and that is a welcomed changed.
That said, I think it has become clear by now that Sun is unable to address the overwealming number of RFEs and bug reports thrown at it in a manner that is satisfactory to its customers.
All I am proposing is that in the hopes of further deformalizing communications between end-users and Sun engineers and reducing turn-around time, end-users should be capable (and encouraged!) to contribute patches against BugParade RFEs and bug reports. The average turn-around for bug fixes for Java bugs is 2+ years. The average turn-around time for patches contributed by users is under a month. The AWT/Swing API, for one, has a large number of user-contributed code that enhancer it one way or another and fixes bugs that were never fixed in the JRE.
It is far easier to review code contributions than figuring out how to code a solution in the first place. I, for one, would love to see BugParade transform into the issue-tracking system used by Netbeans. Their development process is extremely community friendly and turn-around time on fixes is very satisfactory.
Who can get this ball rolling?
Gili
Hi Gill,
Great feedback. I hate to say 'stay tuned', but... uh... stay tuned.
And I promise to send your message to the Java lead architect. Indeed, it's now been sent to him.
jeff
Jeff,
Your help is *very* appreciated!
Thank you :)
Gili

VOTE: discontinue use of the mailing list

Hi everyone
(amusing--I need to post this to both the mailing list and forum; case in point!)
I think that for now, we have to declare victory, and retreat. The gateway is simply not reliable, period.
The fact is, if you post to either the mailing list or the forum, there is no guarantee your message will be received on the other end.
Because this is pretty high-volume list, I'm suggest (sadly) that we discontinue the use of the mailing lists, and go so far as to disable them. Archives will still be available, but one should no longer be able to post to the list.
My reasoning is that, since our main goal is to simplify and enhance Swing programming, we have a secondary goal of providing support to all developers interested in the SwingLabs projects. We can only do this if we can reliably track and respond to questions. If questions, concerns or bugs are "lost" due to the failing gateway, we'll leave users out in the cold. Moreover, valuable information (responses, ideas, suggestions) may be lost in the process. We have few enough resources as it is and we can't afford to lose input and discussion in the community.
So, I'm calling for a vote to discontinue the mailing list until such time as it can be *proven* that we have a reliable gateway between forums and mailing list. This may be *never*. Although I much prefer the lists (I have great search functionality on Google mail), forums, being web-based, are accessible to the greatest number of people who may be behind corporate firewalls or restricted by corporate email policy.
It may be that in the future, we can set up forum and email software on the project servers, hosting swinglabs.org, but we shouldn't count on that.
If you vote against this, please explain how we can continue to operate as a community when we can't reliably communicate with each other.
Best regards
Patrick
+1
I only use the forum, but its disconcerting to have a suggestion completely ignored.
You never know if it was because it was completely stupid, or that it was never seen.
For those of us on the fringe I feel feedback is essential if we are to stay involved.
> IMO, this _has_ to be solved - I honestly can't
> believe that a giant like Sun is unable to build up
> pressure to have it fixed (it seems to be known that
> the culprit is a very outdated jive version on
> collab.net).
Ding, ding, ding. We have a winner! :)

Categories

Resources