Pom Grapher generates a graph from  multi-module and parent-child relationships of Maven projects (POMs).

My use-case was to understand the POMs relations in order to understand the build.
It produces a graphic representation of the modules (nodes) and parent project and sub-module relations (edges) using GraphViz.
Here is a (tiny) excerpt from Apache Geronimo:

black edges indicate module relation, orange edges indicate parent project relations.

This is intended for understanding POM structures, not artifact dependencies.

It can be found over at source.valtech.com.

I have to say that I do not agree.

As Jean Louis Gassee points out in “The Adobe – Apple Flame War” permitting cross-platform development would imply that Apple would accept a level playing field defined by the capabilities of some 3rd party cross platform development tool. If that tool does not support a specific feature it won’t be in the application. It is very unlikely that “that tool” will support platform specifics unless they are widely available as this is against the very nature of a portable toolkit.

In reality we are of course talking about Flash. Flash is – IMHO – the antithesis of an innovative platform. It has been a pain in the neck on some many aspects (usability, stability, resource usage to name a few) for a so long time that I would see a widespread adoption of Flash as a mobile development platform as stifling innovation.

A simple example: Assume for a second that “pinch zoom” is an Apple-only UI feature (it is to some extend).
As it’s not portable there won’t be “pinch zoom” in a portable toolkit thus no pinch zoom in portable applications. Who looses? The users. If it’s not on all platforms you don’t get it. Most – especially corporate – developers may not care that much as they will deliver the same suboptimal user experience on all platforms. Some will go the extra mile and develop native applications as they differentiate against other Apps based on UI experience. Apple, as a hardware vendor looses as they lost differentiating UI features in some applications. Apple is simply “protecting their investment” which is their right to do.
Although I have an opinion about Apple’s handling of the App Store and the iPhone ecosystem in general, I won’t join the flame wars here.

It is simply my opinion that – as Martin wrote “we are only in the beginning of the new mobile age” we need room to innovate and we need competition which is exactly what a portable application landscape won’t give us. “Mobile cross platform died” – so what.

I assume that – once the mobile application space has matured and evolved  and we are looking at more stable platforms I will reconsider my opinion but for now I think Apple’s move – which I also instinctively  dismissed as being ‘evil’ –  will eventually help the whole market.

Unless of course you are assuming that the mobile application market has already been won over by Apple once and for all – an opinion which I haven’t heard so far.

Although it offers a really nice, effective and lightweight way of _structuring_ concepts, my problem right now is that I can not use it because I can not force the concepts in a simplistic tree.
In my head the concepts pop up in different places depending on the context  or point-of-view one is currently looking at. What I’d like to have is something where I can mesh concepts (and actual text) in 1..n qualified graphs – depending on the point of view / context / focus, without re-creating relations (simple aggregations or level of details) in various documents.
All this with an UI that is not getting in the way (aka I am looking neither for  a Wiki not a  Microsoft product)

Help!

Looking at NOKIA’s current product offerings, this move may make sense. NOKIA currently has no competitive offering in the iPhone/Android space and it’s “Business E-Series” does not stack up well against the Blackberry.

Looking at my crystal ball last year I figured that NOKIA was simply taken by surprise. The market-leader – who has not been leading the market on a conceptual level for years – was simply out-innovated. However in 2009 it seems that NOKIA is also having a hard time to catch-up and one of the reasons for this may be it’s platform strategy. The FT article speculates “Symbian is far to clunky to keep up with modern operating systems”. This is consistent with the NOKIA user experience.
My user experience with Nokie E60, E90 in this case: Integration of higher-level functionaity is generally poor and highly inconsistent. Overall device performance ranges from “somewhat sluggish” to “apathic”. Ovi seems to be mostly “missing in action”. Symbian applications are scarce and overpriced.

Why can’t NOKIA deliver on Symbian?
Again the FTD article: “Symbian bases on the EPOC OS from the 90s [..] it is very difficult to add new features in such a complex structure”.

Intresting. Symbian (and it’s predecessor EPOC) is a purpose-build system for small devices with limited memory, CPU and I/O resources. It’s design is from the 80s it was first put on a phone in 2000. So it should be a rather modern (modular, flexible) operating system.

Mameo on the other hand is basically Linux + some extras. It’s monolithic kernel design and it’s primary APIs date back to the original Unix form 1969. Same is true for Android. This is also mostly true for the iPhone OS being Mac OSX which in turn is Mach+FreeBSD.  All of these are older designs. And the Linux kernel is not exactly small either, it clocks in at slightly over 10 million lines of code. The statements above only apply to the kernel, the UI and large parts of the application APIs are  a very different story.

I can only speculate why can’t NOKIA compete on a more recent, purpose build OS against something that was conceived as a (not very good) multi-user timesharing system.

My 0.02€: The system (and especially S60) got screwed up badly and a rewrite on top of the Symbian kernel would be more time-consuming than switching platform. The really bad news is that Mameo (as a platform) is just another run-off-the-mill Linux for small devices. It’s UI concept is far from being innovative. I don’t even know whether it’s up to date (I quit using Mameo when I sold my N800 quite a while ago).
If NOKIA is viewing this as their best play on the smartphone market I assume that we are not going to see much innovation from NOKIA in that market in the future, unless they can finally breathe some life in Ovi.
Sell, Sell!

Opinions?

UPDATE: See also Eike’s analysis on Mameo as a phone platform.

If you are intrested in improving your coding skills and meet fellow developers – please drop me a line.
Once we found a handful of people,  I will set-up a wiki/mailing list from which we can work out details.

I am personally interested in  Randori Katas with various subjects in “traditional” and  emerging functional languages (pretty much as practiced at agile finland)

Optional: Buy a small country of your choice.

Seriously:  The  MobileMe iPhone integration has risen the stakes considerably. I just hope that remote wipe is protected by a user-defined password on the device (and not just by the MobileMe account). They can’t be that stupid

Unfortunately the client uses a JIRA professional with the JIRA internal user management and JIRA currently doesn’t support import of the project between two different versions (Ent – Pro).  I assume, there are more people out there having the same “problem”. It was a manual process but did work in the end.

Prepare the local JIRA installations:

• Install a local JIRA Enterprise – not using the inmemory database
• First create a backup from your JIRA installation and restore it on a local JIRA enterprise installation. Unfortunately you can not connect to the local jira enterprise as you most likely don’t the users and passwords.
  Note the users stored in crowd will not be part of the backup/restore.
• If you can connect to the local JIRA enterprise skip the next steps.
• Next install a local JIRA Professional and create an admin user. Connect to the local database and export the data of the table userbase. You need to know the password as it is stored in the database, in the encrypted format.
• Connect to the database of the local JIRA Enterprise and check if there are any users in the table userbase.
  
• Update the password of the admin user in the database or if the user is missing – create the user in the database table userbase using plain SQL.
• Now you should be able to connect to the local JIRA Enterprise installation

Cleanup the local JIRA Enterprise

• Switch off external user management
• Delete all other projects
• Delete users and groups which are not needed any more
• Delete any customizations or set it as default. Customizations are i.e.
  • workflows
  • permission schemes
  • issue types, screens and screen schemes
  • …
• Change the project configuration (of the only project left over)
  • Issue Security Scheme is set to None.
  • Field Configuration Scheme is set to System Default Field Configuration.
  • Issue Type Screen Scheme is set to Default Issue Type Screen Scheme.
  • Workflow Scheme is set to None.
  • Project Category is set to None.
• Create a backup of the JIRA Enterprise Installation
• Restore the backup into the local JIRA Professional
• Now configure the local JIRA Professional according to your needs

And you’re done with it.
Create a new Backup or project export and Import it into the JIRA Professional Version.

This process did work for us and I hope it will help you resolving this issue if it will arise.