One more "should we care"...

0 views
Skip to first unread message

Nicolas Alvarez

unread,
Aug 31, 2008, 8:19:30 PM8/31/08
to synecdo...@googlegroups.com
Do we have to keep compatibility with current state file format? Should we
support the scenario of migrating workunits in progress from BOINC to
Synecdoche?

Or would it be OK to, say, make client_state a SQLite database instead of a
XML file?


Der Meister

unread,
Sep 1, 2008, 2:30:37 AM9/1/08
to synecdo...@googlegroups.com
Nicolas Alvarez wrote:
> Or would it be OK to, say, make client_state a SQLite database instead of a
> XML file?
I think this would be OK. Maybe we can add something like a "state-file
importer" to help migration from boinc to synecdoche one day.

The only thing that concerns me: the client_state.xml is human readable
which is nice for debugging. What about SQLite? Are there any changes to
get the contents of the DB without running the application associated
with that DB? Is there some sort of "client" programm that can be used
for this? Or are SQLite's files human readable?

Nicolas42

unread,
Sep 1, 2008, 9:06:39 AM9/1/08
to synecdoche-dev
On 1 sep, 03:30, Der Meister <DerMeister...@gmx.net> wrote:
> Is there some sort of "client" programm that can be used for this?

There is a "sqlite3" command-line app you can use to do SQL queries on
a database file.

> Or are SQLite's files human readable?

SQLite files look like binary junk on a text editor. Just like a
MySQL .myd file would :)

Der Meister

unread,
Sep 1, 2008, 11:51:07 AM9/1/08
to synecdo...@googlegroups.com
Nicolas42 wrote:
> SQLite files look like binary junk on a text editor. Just like a
> MySQL .myd file would :) http://imagebin.ca/img/UefTmwk.png
As long as there is some utility (like the one you mentioned) to access
the contents of the database file this isn't a problem. Therefore I have
no objections against such change.

Reply all
Reply to author
Forward
0 new messages