Re: [Bug 60848] SUMPRODUCT fails when first arg is of form --(...)

[Jakub Vojtíšek]

Thanks PJ Fanning and Stanton Fisque for your comments.

I must say that for me it would be a sad news if POI withdraws from
supporting excel formulas evaluations. The company I work for uses POI for
more than 10 years to process excel files. Being able to evaluate excel
workbooks in java and have a full control over it gives our clients many
benefits. We can evaluate workbooks automatically, provide audtabilty and
integrate with our system as well as 3rd party systems (using user
functions for that). We can deploy our system everywhere JVM runs (e.g. on
Linux) without problems. We use POI to evaluate large financial sheets and
actually we had only very few problems with POI so far. Due to the POI
being open source we were able to fix them locally. As I have some spare
time now I'm trying to contribute back to the community and trying to have
those issues fixed in POI. Also looking at other existing issues, the Bug
62271  is one of
them. It is not a problem for us, but it seems to be an interesting issue
with potential to resolve multiple bugs.

So I hope the evaluation stays in POI as I believe it is valuable for many
people. If possible I'll be happy to help with it.

Thanks and regards,
Jakub



On Fri, 13 Oct 2023 at 16:06, stanton fisque  wrote:

> there is a use case, and i have used it, so i will raise a counter comment
> for POI to retain running calculations:
>
> i had a project where i built a harness to load an excel spreadsheet and
> then exercise it as a system of record.  if POI were to lose the ability to
> perform that function, it would be a loss for that use case.  i do accept
> that it is an exotic case, but i'd wager it is not the only one.
>
> as a side note, some thought could be given to (if it were deemed
> necessary to go down this route) providing a contract layer so that
> existing functionality could be retained as an extension library.  the
> current implementation allows for runtime patching of the execution engine
> (i  had to use this for a missing function before PJF accepted  a PR from
> me sometime ago), so there is a minimal contract in place already.
>
>
> Stanton Fisque
> principal technologist
> latticeware.com
> portland, oregon
>
> > On Oct 13, 2023, at 05:27 AM, bugzi...@apache.org wrote:
> >
> > https://bz.apache.org/bugzilla/show_bug.cgi?id=60848
> >
> > --- Comment #15 from PJ Fanning  ---
> > I think it was a big mistake for POI to start evaluating formulas. It is
> too
> > complicated and there are not enough devs interested. It doesn't just
> take
> > someone to submit a change. It takes multiple people to review it.
> >
> >
> > Releases are irregular.
> >
> > Does anyone need POI to calculate formulas? The answer is no. Excel can
> be made
> > to recalculate the formulas when you load the xlsx into Excel.
> >
> > See
> >
> https://poi.apache.org/apidocs/dev/org/apache/poi/xssf/usermodel/XSSFWorkbook.html#setForceFormulaRecalculation-boolean-
> >
> > --
> > You are receiving this mail because:
> > You are the assignee for the bug.
> > -
> > To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
> > For additional commands, e-mail: dev-h...@poi.apache.org
> >
>
>
> -
> To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
> For additional commands, e-mail: dev-h...@poi.apache.org
>
>

Re: [Bug 60848] SUMPRODUCT fails when first arg is of form --(...)

[stanton fisque]

there is a use case, and i have used it, so i will raise a counter comment for 
POI to retain running calculations:

i had a project where i built a harness to load an excel spreadsheet and then 
exercise it as a system of record.  if POI were to lose the ability to perform 
that function, it would be a loss for that use case.  i do accept that it is an 
exotic case, but i'd wager it is not the only one.

as a side note, some thought could be given to (if it were deemed necessary to 
go down this route) providing a contract layer so that existing functionality 
could be retained as an extension library.  the current implementation allows 
for runtime patching of the execution engine (i  had to use this for a missing 
function before PJF accepted  a PR from me sometime ago), so there is a minimal 
contract in place already.


Stanton Fisque
principal technologist
latticeware.com
portland, oregon

> On Oct 13, 2023, at 05:27 AM, bugzi...@apache.org wrote:
> 
> https://bz.apache.org/bugzilla/show_bug.cgi?id=60848
> 
> --- Comment #15 from PJ Fanning  ---
> I think it was a big mistake for POI to start evaluating formulas. It is too
> complicated and there are not enough devs interested. It doesn't just take
> someone to submit a change. It takes multiple people to review it.
> 
> 
> Releases are irregular.
> 
> Does anyone need POI to calculate formulas? The answer is no. Excel can be 
> made
> to recalculate the formulas when you load the xlsx into Excel.
> 
> See
> https://poi.apache.org/apidocs/dev/org/apache/poi/xssf/usermodel/XSSFWorkbook.html#setForceFormulaRecalculation-boolean-
> 
> -- 
> You are receiving this mail because:
> You are the assignee for the bug.
> -
> To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
> For additional commands, e-mail: dev-h...@poi.apache.org
> 


-
To unsubscribe, e-mail: dev-unsubscr...@poi.apache.org
For additional commands, e-mail: dev-h...@poi.apache.org