Hang/crash in templates

[Michael Stahl]

On 25/04/13 13:29, Miklos Vajna wrote:

> TemplateContent::TemplateContent gets an INetURLObject, and the next
> assert fails, as GetProtocol() is INET_PROT_NOT_VALID. If you step up
> the backtrace in gdb, you see that the "URL" was ~memory garbage.
> 
> On the other hand, for now I would just change back the cache format.
> Does that make sense?
> 
> I mean if you are OK with pushing
> http://cgit.freedesktop.org/~vmiklos/lo-core/commit/?id=83544b27f3dcfb132b3eb46d387f62ebc9c3e160
> to master. And then this can be still discussed (if it's OK to change
> this format or not, if hundred/nano seconds matter for templates or not,
> etc) without being in a hurry.

argh... there is a binary format "template cache" in the user profile?

what would happen if the incompatibly changed format loading code were
deployed to users with a pre-existing cache in the old format?

ceterum censeo formatos binarios delendam sunt.