powered by Jive Software

Next Release of Spark

I propose upping the next version number of Spark to 3.0.0 instead of 2.7.0 as it has been over a year since the last release and it would signify a sort of “re-release” and renewal of the project. This would be the first offical version of Spark to support/run on j7, and a lot of bugfixes and improvements have been made to justify it’s progress.

What do you all think?

Has there been any significant upgrades (especially visible to the user) that would justify a 3.0?

The use of j7 alone should be significant, as it gets rid of the dreaded “steal window focus” issue, and signifies the advancement of the project. Plus file transfer has been fixed (major for my users). I believe there was a copy/paste issue that was solved as well. It’s been a long time since a release, and a point release might be seen as just some minor stuff when in reality it’s more or less a re-vitilazation of the project, fixing some much needed issue and adding to stability/reliability…

Also, on a side note, not sure which version of JTattoo is in the current build of spark, but perhaps updating that would provide more fresh air to the project (The newer version of JTattoo look to have fixed a bunch of bugs and added some nice new LookAndFeel skins). That would definetely be a visible thing for users. I personally battled JTattoo issues at my office… sometimes just telling the user not to use any skin but default was the only solution as JTattoo would cause some weirdness to happen (2.6.3)

I would love to see this

The things you are mentioning are just bug fixes. Moving to j7 might be a big thing for a developer, but they are practically meaningless to an end user. You have a valid point with the revitalization, but I think it would be better to make the change, release a 2.7 to get the current bug fixes and then work towards some significant additions/changes for the next release, which would be major release.

If I upgrade from 2.6.3 to 3.0 and see no discernable difference in the look or functionality, I am going to simply wonder what the upgrade is for (with respect to a major release).

That being said, I don’t keep up with things in Spark, so maybe these significant changes/improvements already exist. I am simply basing my opinion on what has been brought up in this thread. I just don’t agree with the justification you are giving for a major release.

fair enough points. What would you consider to be major changes visually that would justify a 3.0 release? THe GUI is pretty well layed out in my opinion, not leaving much room for changes… however adding additional features sure could be enough i suppose… Also perhaps moving to the latest release of JTattoo which brings a lot of bugfixes and new skins? I’m just trying to get my sights on what would consitute a 3.0 release so we can work towards it. I think a 3.0 would be great for the project… we could promote Spark as being a revitalized project with lots of fresh air, etc… I think I can even get an open source / linux oriented podcast network (jupiterbroadcasting.com) to review it if we can a major release. That may get new users and possibly new contributors.

PS: anyone know what release of JTattoo we currently use?

I think your suggestions all point in the right direction. It is, of course, very relative, but I think as long as the user can see something reasonably significant then it makes sense.

2.7.0 is overdue and has significant bug fixes in the area of file transfer. We are rolling out 2.7.0 internally with a world wide base of 15000 named users. The key problem was and is the Java 7 build/packaging and the Mac client.

3.0 content depends on the features. I can certainly think about great ideas what is of interest next. IMHO the whole preferences area might deserve a rework. Managing privacy is also not necessarily easy or intuitive. Depending on server (and that is Openfire) features, I am sure that the management of external gateways is not flexible enough to deal with Spectrum.im (the Kraken replacement). File transfer of large files (more than 100 MB) could be done differently (e.g. via Bittorrent). And so on… This all depends on the amount of coding hours spend. It looks to me that Spark depends a lot on companies that push features into the trunk. If a student wants to spend his Bachelor/Master on it, we will certainly help.

I somewhat agree with Robin, that for the bug fixes and refactoring under the hood 2.7.0 is ok (it is already a big leap from the 2.6.3 version number). If the name it 3.0 it can be a disappointment for users if they won’t find anything “new” after they upgrade. Also if we use 3.0 now and then update the look and add some significant new features, then what number should we give it? 4.0? I don’t like fast major version number changes, though it is popular now (Firefox, Chrome). It is easy to get lost in those numbers, yet minor version number changes are easier to associate with fixes.

New skins and update to JTatoo, if it really looks different is a fine point for 3.0.0, but not enough, imho. Let’s take a look at this feature requests list and pick what would justify next major version bump http://issues.igniterealtime.org/secure/IssueNavigator.jspa?reset=true&jqlQuery= project+%3D+SPARK+AND+resolution+%3D+Unresolved+AND+issuetype+%3D+%22New+Feature %22+ORDER+BY+priority+DESC&mode=hide

Of course, someone would have to make those features and this is not an easy task Personally i would like to have separatable chat windows (a patch was in the works, but was never finished). Not for myself as i like one window approach of Spark, but for my users, who are using another client. SPARK-893 (additionally SPARK-935)

Other interesting features:

SPARK-916 Quarentine incoming messages when in DND mode

SPARK-634 Make drag and drop file transfer work in Linux

SPARK-996 Groupless mode

SPARK-1001 Add custom responses depending on the presence status

SPARK-1071 Add an option to automatically join ad-hoc group chat on invite

SPARK-1238 Add support for XEP-0184: Message Delivery Receipts

SPARK-1430 Create an installer for OpenSolaris/OpenIndiana

SPARK-1150 Allow the font face, size, colour & attributes (bold, italic, underline, strikethrough) to be changed

SPARK-1468 Make Spark compatible with WE4Java

Of course, most of these requests filed by me Or we can think of something new.

Agreed, save major version numbers for releases that users will feel like they’re gaining big benefits from, both visually and on the back end. I hate getting a major version release that doesn’t feel like anything has changed!

JTattoo was integrated in April 2011. We removed Substance as LaF because of bugs and lack of development. The JTattoo guys was very responsive regarding our issues. I am not aware of JTatto related bugs that are affecting day to day work. I think we deliberately removed some strange looking JTattoo skins due to issues with some ahrd coded colors in Spark. For example the black skin is ugly because of the light blue color that was used in the 2.5.8 Spark and is hard coded somewhere. That’s the reason why not all funky JTattoo skins are in Spark. At that point in time we wanted to clear the GUI and did not pursue a large scale refactoring/rewrite.

BTW: Looking into our Jira identifies the last JTattoo update http://issues.igniterealtime.org/browse/SPARK-1459 it is of November 2011.

Is there anything that would help from a resources perspective? My team (myself and one other person) have no additional time to spend on coding, but perhaps there is something else I could do to help? Is there any software that could be donated to assist with the project, for example?

On the Bamboo server, I see that the target dates for 2.7.0 are the beginning of May, how realistic are these dates?

Walter I think is on the right path: We should release what is in 2.7.0 and to me it does not matter what we call it. Waiting for the next beta relase of SMACK would liekly be a good idea. There are some DNS enhancements still pending that might be worth waiting for too. We should be able to commit invisibility capabilities by end of March. If everyone else wanting to contribute anything could line up with this, that would be ideal.

Hey all, any update on the release date for the new Spark client? I’m an admin, and the inability for Spark to use Java 7 is beginning to cause me a headache. I see in a previous post that it was scheduled to be released in May. How’s that coming?

We are missing out dev ressources for Spark. Two ways to fix the Java7 issue:

  1. Install der online version of Spark. It uses the OS JVM that can be kept current without Java integration in the next Spark release.

  2. Build your own installation package

Spark 2.6.3 runs with J7

Spark 2.6.3 runs with J7

Thanks, but that hasn’t been my experience. I haven’t gotten it to work once across WinXP/7 or x86/64. Still having to run J6 on every machine that requires it installed.

I am interested in running Spark online thouhg. I attempted SparkWeb in the past and had a whole bunch of problems. I’m running on Ubuntu 10.04. Can you link me to a recent how to article?


I’d like to weigh in, this has been my experience as well. I’ve been testing nightly builds of the online installer for a while, and not one of them has run on a machine with only Java 7 installed (XP, Win7 or Win7 x64).

Another option would be to run nightly builds of the offline installer. It has java 7 bundled and e.g. it has no focus intercepting issue which is present when running with java 6 (when you receive new message while doing something in another app). It also has some other issues fixed. I personally use it all the time and have installed for my collegues.

first link http://bamboo.igniterealtime.org/browse/SPARK-INSTALL4J-595/artifact/JOB1/Instal l4j/

1 Like

We generally don’t run the bundled java editions, since the bundled JRE won’t receive security and other bug patches in a timely manner. Much prefer to use the online edition that stays up to date with java as part of normal system maintenance.

Well, i understand there could be strict policies in place, but i don’t care much about older Java version as long as it’s not running in my browser, where most security holes exist. Other bug fixes. Hm. I think Oracle is mostly fixing security breaches and the only bug fix i cared about was that focus stealing issue.