[izpack-users] shortcutSpec.xml UTF-8 Byte Order Mark breaks IzPackoperation

Bartz, Klaus Klaus.Bartz at coi.de
Tue Oct 10 14:07:11 CEST 2006


Hi Hans-Georg,
byte order for UTF-8... it is matter of opinion whether a
silent add of a BOM is clever or not. May be you should
send a bug report to the developer of the editor.

IzPack is an OS independent installer which uses some third
party tools. Nobody of us have to delve into the problem to 
find the reason of the problem. If you look into the docu you
will see, that IzPack uses a really little XML parser called
nanoxml. This parser has some restriction e.g. it breaks on 
BOMs. Additional it cannot validate. And the error messages
are also restricted. But if you write no unreadable chars in
the file you can use UTF-8. We do it in some language files
(e.g. rus.xml).
We have discussed more than one time to switch to the  
xml parser which is in the VM since some releases.
But until jet nobody has made the needed work. 

If I look into wikipedia for BOMs
(http://en.wikipedia.org/wiki/Byte_Order_Mark) there is an
explicit warning to use BOMs on Unix-like systems. We use in
our sources unix like coding rules...

Klaus


>-----Original Message-----
>From: izpack-users-bounces at lists.berlios.de
>[mailto:izpack-users-bounces at lists.berlios.de]On Behalf Of Hans-Georg
>Michna
>Sent: Tuesday, October 10, 2006 12:32 PM
>To: izpack-users at lists.berlios.de
>Subject: [izpack-users] shortcutSpec.xml UTF-8 Byte Order Mark breaks
>IzPackoperation
>
>
>Just stumbled over a grotesque problem that took me hours to
>identify.
>
>I had mistakenly declared the Win_shortcutSpec.xml file as
>UTF-8, but actually stored it as ANSI code, i.e. it had no Byte
>Order Mark (BOM).
>
>For a long time I didn't notice my mistake, and everything
>worked, particularly since the file did not contain any
>characters that required UTF-8 multi-byte encoding, i.e. the
>file was actually correct, except for the missing BOM.
>
>Now I made a little change to the file and used an editor that
>was clever enough to recognize the UTF-8 encoding and add the
>missing BOM. From that moment IzPack no longer showed the
>shortcut pane. Imagine my frantic searching, because I could not
>see any error in the file or in any other. As I wrote, it took
>me hours to find the inconspicuous and correct BOM and the fact
>that its absence made IzPack work again.
>
>We have a problem here. Even if we tell users that they should
>use some other encoding, like ISO-8859-1, proper XML should be
>encoded in UTF-8, and many editors, including Windows Notepad,
>will add the BOM, rendering IzPack unworkable.
>
>Could any of the developers delve into this and remove this
>insidious stumbling block? It is no longer a problem for me, but
>it will surely cause endless grief to others.
>
>I'm actually surprised that this topic hasn't come up yet, or
>has it?
>
>Hans-Georg
>
>-- 
>No mail, please.
>
>_______________________________________________
>izpack-users mailing list
>izpack-users at lists.berlios.de
>https://lists.berlios.de/mailman/listinfo/izpack-users
>



More information about the izpack-users mailing list