Eric,

there is a program PDBcruncher (http://klyatskin.narod.ru/#PDBcruncher)
designed to pack any database before syncing it into device.
That is exactly what you should  look at to solve the problem with slow
synchronization.

-- 
Konstantin





"Erik Norvelle" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> To LionScribe & Others:
>
> I haven't yet written any code to implement your suggestion to pack more
> data (i.e. Latin words) into each PDB record.  I've been thinking through
> ways of doing this, and while it's certainly a doable thing, I'm wondering
> if this isn't a wheel that's been invented before.  Do any of you know of
> any (open source or reusable) code that I could borrow or examine that
> explains how to effectively pack multiple (real) records into a lesser
> number of PDB records.  I particularly have had problems with
word-alignment
> and endian issues which have caused me headaches in the past, so I would
> love to find some code that has already been debugged.
>
> TKA,
> Erik Norvelle
>
> --------------------- Quoted Message Below -------------------
>
> Subject: Re: Slooooowwwww syncronizations (.pdb file with 30000 plus
> records)
> From: "LionScribe" <[EMAIL PROTECTED]>
> Date: Tue, 15 Jun 2004 16:27:00 -0400
> X-Message-Number: 57
>
> Also, what a huge amount of wasted overhead, I believe around 18 bytes per
> record. (18 * 30000 = 540,000). Try using DmDatabaseSize and see the
> difference between the totalBytes and the dataBytes.
> I would suggest combining 200 words per record, and storing the offset.
> Also, with Databases that size, it would be more appropriate to use
> compression, try using the ZLib library to compress each record (of 200
> combined).
>
> LionScribe
>
>
> "Ben Combee" <[EMAIL PROTECTED]> wrote in message
> news:[EMAIL PROTECTED]
> >At 01:46 PM 6/15/2004, you wrote:
> > >Greetings,
> > >
> > >I just posted a notice about my latin dictionary program... I have
> another
> > >question related to synchronizing the database files that go with my
> > >program.  I use a method of creating a single record in my database for
> > >each word which is to be stored (and there are approximately 25000
> words),
> > >as well as an index into this database (which has about 30,000
> > >entries).  The index.pdb file is only about 640K, yet takes nearly
three
> > >hours to synchronize!  The other file (the Stems.pdb file) takes about
> the
> > >same time, and is 2.5M.
> > >
> > >The only thing I can think of in this context is that the creation of
so
> > >many individual records on the handheld side is holding up the
> > >synchronization... I am using a USB connection, so speed should not be
a
> > >problem.  Has anyone else had this problem, and if so, what's a good
> solution?
>
>
>



-- 
For information on using the Palm Developer Forums, or to unsubscribe, please see 
http://www.palmos.com/dev/support/forums/

Reply via email to