Re: search for better regtest comparison algorithm

[Han-Wen Nienhuys]

On Mon, Jul 29, 2024 at 12:20 PM Luca Fascione <l.fascione@gmail.com> wrote:
>
>
>
> On Mon, Jul 29, 2024 at 11:09 AM Werner LEMBERG <wl@gnu.org> wrote:
>>
>> This means that it would be sufficient to make the Cairo backend also
>> create logging output.  In case I'm not missing something this
>> shouldn't be too hard to add.
>
> Can this be done?
> If one were to log all the calls and arguments to Cairo functions and then
> write a corresponding program to replay them later, would they get the
> same PDF as from a direct binding?

Possibly, but it's complicated:

* some Cairo calls take FT_Face instances as argument, for example. If
you log and replay the Cairo calls, how can you be sure you're passing
in the right Freetype font into the replay?

* Also, the current "production" backend is the PS + Ghostscript
route, which wouldn't be tested by a Cairo log/replay.

* Even if Cairo were the production code, then the log/replay is extra
code that is not used normally. How would you be sure it is free of
bugs?


-- 
Han-Wen Nienhuys - hanwenn@gmail.com - http://www.xs4all.nl/~hanwen