First useable version of the portable scribble merger released.

[thomega]

Originally Posted by hanche

I discovered how to make it work: If you cd into the directory where the pdf file resides then ism finds the scribbles and merges them just fine. I am not at all sure why this is so, but now that you know a workaround, maybe you can figure out what is going on.

I had forgotten to strip the directory from the filename before querying the database. It's fixed in the SVN repository.

[hanche]

Okay, next bug … If I try to run “irext rm filename.pdf”, I get the response

Fatal error: exception ThoSQLite3.Unexpected("single_step", "CONSTRAINT")

Any idea what could cause this?

[thomega]

Originally Posted by hanche

Any idea what could cause this?

The SQLite database complains that a request to delete a record would violate one of its internal constraints. I have checked some more verbose error messages into the svn. Could I ask you to try the reproduce the behaviour with them?

Also, could you send me the metadata.db file, so that I could investigate? Did the file contain annotations?

A temporary workaround should be rm filename.pdf followed by irext update. Does it work?

[hanche]

Originally Posted by thomega

The SQLite database complains that a request to delete a record would violate one of its internal constraints. I have checked some more verbose error messages into the svn. Could I ask you to try the reproduce the behaviour with them?

Also, could you send me the metadata.db file, so that I could investigate? Did the file contain annotations?

A temporary workaround should be rm filename.pdf followed by irext update. Does it work?

I got the same result with the new version, no more verbose error messages. irext update seemed to work, and it removed the file, but when I looked in metadata.db with sqlite3, the metadata was still there (but it appears the reader itself removed it after it regained control). Yes, there were annotations in the file. Now I fiddled around trying to reproduce the problem (I had removed the file, remember), and failed to do so. There must have been some odd circumstance that made it fail, and we'll have to wait for the problem to reappear before we can test it again.

Now to a slightly different topic: make opt failed for me because ld on the Mac doesn't like -rpath=…, it wants the pathname argument separately, as in -rpath …. I believe other versions of ld also understand this format, so I think you can just replace -Wld,-rpath=… in Makefile.in by -Wld,-rpath,…. (Or else, it will have to another thing for configure to work out.) This worked on the Mac, at least. Also, there are some occurences of make that really ought to be $(MAKE) (I see you have it right where it counts the most, though.)

[rkboyce]

Thanks a bunch to the author of this tool. It was key to helping me decide to purchase the Irex 1000S. I just wanted to report that the issues that I have had with 'test' target of the Makefile for irextools-0.3.

I successfully installed the suite on my laptop running Debian Lenny after installing the 'ocaml' and 'sqlite3' packages. However, running 'make test' from the base folder of the project produces the errors that I have listed below. I do not have my reader yet (coming next week) so I do not know if these errors indicate a problem with irextools. The commands 'ism' and 'irext' appear to execute with no errors but I did not see a way to manually test the programs since I don't have the reader in my possession.

thanks

-------------------

richnix% pwd
/home/boycerd/Downloads/irextools-0.3
richnix% make test
make -C testing bin
make[1]: Entering directory `/home/boycerd/Downloads/irextools-0.3/testing'
make[1]: Nothing to be done for `bin'.
make[1]: Leaving directory `/home/boycerd/Downloads/irextools-0.3/testing'
make -C testing run
make[1]: Entering directory `/home/boycerd/Downloads/irextools-0.3/testing'
./test_suite
....EEEEEE
================================================== ====================
Error: iRex Tools:2:Metadata_File:1:not empty db

Invalid_argument("database file ./metadata.db exists!")
----------------------------------------------------------------------
================================================== ====================
Error: iRex Tools:2:Metadata_File:0:empty db

ThoSQLite3.Unexpected("get_integer", "NULL")
----------------------------------------------------------------------
================================================== ====================
Error: iRex Tools:1:Cp & Mv:3:mv (no db)

ThoSQLite3.Unexpected("get_integer", "NULL")
----------------------------------------------------------------------
================================================== ====================
Error: iRex Tools:1:Cp & Mv:2:mv

ThoSQLite3.Unexpected("get_integer", "NULL")
----------------------------------------------------------------------
================================================== ====================
Error: iRex Tools:1:Cp & Mv:1:cp (no db)

ThoSQLite3.Unexpected("get_integer", "NULL")
----------------------------------------------------------------------
================================================== ====================
Error: iRex Tools:1:Cp & Mv:0:cp

ThoSQLite3.Unexpected("get_integer", "NULL")
----------------------------------------------------------------------
Ran: 10 tests in: 0.03 seconds.
FAILED: Cases: 10 Tried: 10 Errors: 6 Failures: 0 Skip:0 Todo:0
make[1]: *** [run] Error 1
make[1]: Leaving directory `/home/boycerd/Downloads/irextools-0.3/testing'
make: *** [test] Error 2

[thomega]

Originally Posted by rkboyce

I just wanted to report that the issues that I have had with 'test' target of the Makefile for irextools-0.3.

Something changed in the SQLite that comes with Debian Lenny. There's a working workaround in the current svn version, but I'm not happy with it because I don't fully understand what's going on. (Note to self: make a new release anyway)

[rkboyce]

Originally Posted by thomega

Something changed in the SQLite that comes with Debian Lenny. There's a working workaround in the current svn version, but I'm not happy with it because I don't fully understand what's going on. (Note to self: make a new release anyway)

Thanks for the quick reply. I downloaded and built the svn version and 'make test' runs with no errors on my laptop. I look forwarded to giving the code a work out within a week or so.

-rdb

[godel10]

Originally Posted by rkboyce

Thanks for the quick reply. I downloaded and built the svn version and 'make test' runs with no errors on my laptop. I look forwarded to giving the code a work out within a week or so.

-rdb

I would like to try the svn version (because I have the same "test" problem), but what is the download link for the svn version?

[thomega]

Originally Posted by godel10

what is the download link for the svn version?

Go to http://forge.ocamlcore.org/scm/?group_id=60 and ye shall find (including a link to download a tarball of the current svn version).

[godel10]

Originally Posted by thomega

Go to http://forge.ocamlcore.org/scm/?group_id=60 and ye shall find (including a link to download a tarball of the current svn version).

Thanks for the link. I have succeeded to download this last version, but now I do not know how to compile it (I guess the source is in the "trunk" folder); this time there is no "./configure" script to launch.

Could anyone tell me how to compile it?

[thomega]

Originally Posted by godel10

this time there is no "./configure" script to launch.

Could anyone tell me how to compile it?

1. (preferred) run the commnd autoconf in the directory (trunk) where the file configure.ac resides, this will create configure
2. (last resort) the old configure script should work

[godel10]

Originally Posted by thomega

1. (preferred) run the commnd autoconf in the directory (trunk) where the file configure.ac resides, this will create configure
2. (last resort) the old configure script should work

Thanks, this time I succeeded using autoconf.

[Myself]

Thank you! This is a marvelous tool.

One question: is there an easy way to change the color of the scribbles?

[thomega]

Originally Posted by Myself

is there an easy way to change the color of the scribbles?

ism's options -color and -opacity come to mind, as relvealed by the option -help . The former accepts all the color names that showrgb produces.

[Myself]

Originally Posted by thomega

ism's options -color and -opacity come to mind, as relvealed by the option -help . The former accepts all the color names that showrgb produces.

Ah, well. I'm using the Mac binary. I've tried playing around with calling it from the command line with various arguments but that doesn't seem to work. Let me poke around and see if I can figure out how it works.