 |
Tiger Snapshots: An experiment in openness
Posted by mreinhold on June 14, 2004 at 06:01 PM | Comments (13)
Last Friday we posted the first of the Tiger Snapshot releases to
the web. Build 55 is available right now, and the plan is to post each weekly
build until we reach the Release Candidate (RC) build. Unlike our milestone
releases these builds receive only limited testing, so they're not for the
feint of heart. If you'd like to live on the bleeding edge, though, then these
builds are for you.
Our main motivation for doing the snapshot releases is to ensure that the
FCS release of J2SE 1.5.0 is of the highest possible quality. The end-game of
a big software release, as anyone who's survived one knows, is a particularly
tense time during which every change is closely scrutinized, lots of hard
decisions must be made, and there's never enough testing. By posting each
weekly build on the web we're hoping to increase the level of external testing
and thereby improve the quality of the final result.
A second motivation for doing the snapshot releases is to gain some more
experience with the "release early, release often" development style that's
been so successful in the Open Source community. The Tiger Snapshot model
isn't exactly "release early", since these are post-Beta 2 builds, and
it's not quite as "often" (i.e., nightly) as some projects, but it's a
more aggressive one than we've ever tried before. If this experiment goes well
then we'll consider releasing even earlier and oftener in future releases.
Bookmark blog post: del.icio.us Digg DZone Furl Reddit
Comments
Comments are listed in date ascending order (oldest first) | Post Comment
-
any features missing
My only question is regards to the fullness of the release. Are there any missing features, or is this ready for prime time abuse? The reason I ask is that the previous 2 betas were missing Enums, or at least Enums as I thought they where going to be implemented. :-)
Posted by: malcolmdavis on June 14, 2004 at 09:21 PM
-
define what you want feedback on
I'm really pleased about increasing communication and improving the feedback loop.
It would seem to me that this feedback loop could serve two very useful purposes:
1. a bigger testing community will catch more bugs
2. the community will also report about partly implemented features that are missing small (but critical) bits.
#2 can be a biggie. Here's an example: pack200 is phenomenal. It shrinks ScheduleWorld from about 1.5MB to 350KB. It's easy to forget about folks who use modems, but there are a lot of them still out there and shrinkage like this is a big deal to them.
Unfortunately, pack200 is totally useless to me because it isn't supported with versioned Java Web Start jars. To me it would seem a major oversight like this should be accepted as part of the feedback cycle and the jdk should have this fixed before FCS. It would be disappointing to have to wait another 6+ months(?) for some future version to have this fixed.
Are (IMHO) critical issues like this up for consideration?
Posted by: markswanson on June 15, 2004 at 06:12 AM
-
unclosed feedback loop?
I'm not sure how the feedback loop you seem to want can work since I found no way where I can post any problems I find..
I had problems with the new Swing feature that allows a window to be managed by the native window manager.
The implementation checks if an option is set in the JVM, but it does so in a static{} block. Which means that no global changes can be made during runtime :(
Posted by: zander on June 15, 2004 at 07:12 AM
-
who do ya think you're fooling?
...feint of heart???
Posted by: rstinejr on June 15, 2004 at 11:56 AM
-
any features missing
Beta 2 was feature complete, so if you think something is missing please let us know ASAP.
Posted by: mreinhold on June 15, 2004 at 01:14 PM
-
unclosed feedback loop?
As mentioned on the snapshots page, as well as
in the release documentation, you can submit a bug or RFE via the usual
channels (http://bugs.sun.com/services/bugreport/index.jsp)
or you can send other comments or questions to tiger-snapshots@sun.com. You can
also contact specific engineering teams individually via the addresses and
forms listed on http://java.sun.com/mail.
Posted by: mreinhold on June 15, 2004 at 01:21 PM
-
Thanks Sun staff
I have installed your beta 3 build 55. It works well so far. I am glad to test it.
Thanks a lot
Posted by: ceo on June 15, 2004 at 09:38 PM
-
Re: unclosed feedback loop?
Hmm;
I could just admit I missed that and leave it at that, but the real reason I missed it is that I was looking for something different.
I was really looking for an OPEN discussion forum. Where I can read others having problems and the feedback from Sun employees.
A idea for a next iteration of Sun getting more open with Java?
Posted by: zander on June 16, 2004 at 02:32 AM
-
Feedback loop
I think a forum for discussing the releases would be good for some pre-bugreport queries like "have anyone seen this" or "can you guys test this".
Would probably filter some erratic bug reports as well, so it should be good for everyone.
I often find that posting a bug report is a bit too heavy, since I usually don't have the time to envestigat it enough.
Like now, when I found that GregorianCalendar jumps more than a year when I set the hour. Just happens in one place in my program and I can see it in the debugger.
I would like to have somewhere to ask the question: "Have anyone found some stange things with GregorianCalendar that wasn't there in 1.4?"
Cheers
Posted by: mgrev on June 16, 2004 at 05:31 AM
-
Re: unclosed feedback loop?
+1
Posted by: markswanson on June 17, 2004 at 06:12 AM
-
automating the fetch
The weekly snapshots are a great addition, but their usefulness is tempered by inability to automatically fetch the latest snapshot.
I'd like to use the latest snapshot with Apache Gump - the Java continuous integration tool, but it is a real hastle having to manually hit http://java.sun.com/, follow the links and agree to the license every time a new snapshot is available. Perhaps a "-Xupdate" flag to java and javac or something like the netbeans update centre could solve the problem.
--
Michael
Posted by: md84419 on June 17, 2004 at 07:31 AM
-
Feedback loop
wrt GregorianCalendar: I _heavily_ use and rely on it. I haven't seen your scenario, but I know the behaviour changed enough from 1.4.2_02 to completely break ScheduleWorld. I would have gladly shared this info and the fixes I made but there didn't exist a single place to share this info. The one-way communication mechanisms Sun provides are not enough. I want other Java developers to benefit from this knowledge as well.
Using google to search for problems and solutions enables me to leverage the experience and hard work of countless developers. I want to participate in and foster such a community as it enables me to do the work of 3 people.
Posted by: markswanson on June 17, 2004 at 09:56 AM
-
any features missing
While beta 2 might have been feature complete there were some new things added in Beta three that can have surprising side effects. One I noticed was the new spinning logo. Very nice to look at. Awful if you are doing any kind of extra work in the Applet init method! This was a feature that you guys tried to slip in in 1.4 but backed out due to licensee request for performance reasons. Here you go trying to sneak it back in! Note that a bug was filed for this...
Posted by: sasjaa on June 17, 2004 at 11:55 AM
|