| 1 |
joko |
1.1 |
@title Industry Application Servers - do we really need them? |
| 2 |
|
|
@newsgroup nfo.caesar, nfo.dev.perl, nfo.dev.php |
| 3 |
|
|
|
| 4 |
|
|
@sequence meta |
| 5 |
|
|
@document-history |
| 6 |
|
|
@cvs-info |
| 7 |
|
|
$Id: joko_2003-03.twingle,v 1.6 2003/03/07 16:56:42 joko Exp $ |
| 8 |
|
|
$Log: joko_2003-03.twingle,v $ |
| 9 |
|
|
|
| 10 |
|
|
|
| 11 |
|
|
Hi there, |
| 12 |
|
|
|
| 13 |
|
|
[...] the text coming here got refactored from another post: |
| 14 |
|
|
"org.netfrag.xyz - ApiDoc available - powered by phpDocumentor" to "nfo.yakka" [...] |
| 15 |
|
|
|
| 16 |
|
|
Just the GUI people (Gnome and KDE) have their component architectures where |
| 17 |
|
|
they are struggling to get them more: a) stable b) responsive c) flexible which are |
| 18 |
|
|
things essential for a GUI. |
| 19 |
|
|
(Microsoft chose the other way: just d) integrate everything into a Win32API and |
| 20 |
|
|
sell this stuff - okay they have been faster with this approach....) |
| 21 |
|
|
|
| 22 |
|
|
Okay - back to the point: |
| 23 |
|
|
Not even to think about integration of anything where is nothing, hmmm.....? |
| 24 |
|
|
|
| 25 |
|
|
<b>In contrast to this, the Industry has even more component systems: </b> |
| 26 |
|
|
Microsoft's COM for its clunky OS, (okay - .NET is on top of the stack by now |
| 27 |
|
|
pushing COM/DCOM one level towards deprecation) - however: |
| 28 |
|
|
Apple's WebObjects or Macromedia's Xyz for Web-Application-Development |
| 29 |
|
|
and Java's J2EE Architecture for large-scale industry-wide software development. |
| 30 |
|
|
Not to mention all the other "Application-Server-Technologies" related to specific |
| 31 |
|
|
companies' products: Oracles iDatabase-stuff (q: they have a fs2db-mapper - |
| 32 |
|
|
do they have an orm?), IBM's WebSphere (or the classical well-known Groupware |
| 33 |
|
|
Application Server "Lotus Notes") or even SAP. |
| 34 |
|
|
|
| 35 |
|
|
Jonen and me talked yesterday about this point regarding a thread at the |
| 36 |
|
|
<b>Kroupware mailing-list</b> started by Thomas Zander: |
| 37 |
|
|
@thread-start http://mail.kde.org/pipermail/kroupware/2003-February/001114.html |
| 38 |
|
|
@thread-part http://mail.kde.org/pipermail/kroupware/2003-February/001117.html |
| 39 |
|
|
@thread-part http://mail.kde.org/pipermail/kroupware/2003-February/001118.html |
| 40 |
|
|
@thread-end http://mail.kde.org/pipermail/kroupware/2003-February/001139.html |
| 41 |
|
|
|
| 42 |
|
|
In contrast to this, for things like "mailrouting between kolab servers in the same domain" |
| 43 |
|
|
no additional solutions are developed: |
| 44 |
|
|
@link http://mail.kde.org/pipermail/kroupware/2003-February/000856.html |
| 45 |
|
|
|
| 46 |
|
|
And even things like "Using the kmail kolab client and outlook to access the same account" |
| 47 |
|
|
@link http://mail.kde.org/pipermail/kroupware/2003-February/000970.html |
| 48 |
|
|
@link http://mail.kde.org/pipermail/kroupware/2003-February/000973.html |
| 49 |
|
|
@link http://mail.kde.org/pipermail/kroupware/2003-February/000975.html |
| 50 |
|
|
are unfortunately completely unaddressed there. |
| 51 |
|
|
|
| 52 |
|
|
Mozilla is okay: |
| 53 |
|
|
@link http://mail.kde.org/pipermail/kroupware/2003-February/000977.html |
| 54 |
|
|
@link http://mail.kde.org/pipermail/kroupware/2003-February/000996.html |
| 55 |
|
|
|
| 56 |
|
|
Still not - but planned: "Synchronize the mozilla calendar against the kolab server" |
| 57 |
|
|
@link http://mail.kde.org/pipermail/kroupware/2003-February/001001.html |
| 58 |
|
|
|
| 59 |
|
|
|
| 60 |
|
|
|
| 61 |
|
|
What I want <b>to do</b> is creating a Groupware Application without a |
| 62 |
|
|
primary server required. |
| 63 |
|
|
|
| 64 |
|
|
Of course, one *can* be added and *will be* added to serve tasks which |
| 65 |
|
|
simply can't be solved using a passive-only system. |
| 66 |
|
|
|
| 67 |
|
|
|