Page 1 of 1

Bug - accented characters

PostPosted: Wed Aug 22, 2018 5:06 am
by vlapr
Dear developers, you have created a great product but it has a bug for years.

If you enter accented character (ěščřžýáíéůú, etc.) ALL TEXT is LOST on save. This make your great product unusable fro non english users.

There are two options:

1. Add unicode support that solve all similar problems
2. Add a filter that replace accented character with their non accented version.

Of course solution 1 is better but based on your answer many years ago you do not have support unicode and is not possible to rewrite the application.

Solution 2 is not so hard - it requires you to create one function that do the replacement and chain it into program before text data are saved.

I hope you will found time to solve it soon.

Re: Bug - accented characters

PostPosted: Fri Aug 24, 2018 12:21 pm
by ddhsoftwareadmin
Thanks for writing. The above is actually not a bug in the program- HanDBase does not support any characters that are not part of the Windows/Latin 1 character set. This is an integral part of the binary database format used by the application and we'd have to rewrite the format entirely and release updates for every version of HanDBase on every platform. I'd love to be able to do this but there is simply not enough interest in HanDBase from non Latin speaking countries to justify the great amount of work required for this purpose alone. I am the sole developer working on this project at this time and it is a part time job so I don't have the amount of hours required to get it done.

When and if I do update HanDBase to a new database format some day in the future I can promise you it will support Unicode!

Re: Bug - accented characters

PostPosted: Mon Sep 03, 2018 2:05 am
by vlapr
Thank you for your answer. Actually it is a bug, because the software allows you to enter such characters and then process them incorrectly.

Unfortunately I got same answer 5 years ago :-(

I'm developer as well so I understand your troubles with database limitation...

... but is it possible to include (maybe optional function) that convert accented characters to their non accented version?

1. Your actual version clear (lose) all text when it contains accented characters (probably due to an exception somewhere)
2. I enter text without accents but due to automatic correction there are some accented characters and I have to change them manually. If I forget I lost all text.

I estimate that write and test such conversion function (eg. filter out accented characters) should take just few hours and it helps a lot!

Thanks and regards.

Re: Bug - accented characters

PostPosted: Mon Sep 03, 2018 12:37 pm
by ddhsoftwareadmin
Thanks for your reply- we'll likely have to agree to disagree on the definition of a bug. To me a bug is an unintended behavior. Since HanDBase by its defined limitations does not support non-Latin languages, having to slow down input parsing to filter incoming text for characters that are not supported seems unnecessary.

I empathize with your concern and desire to support these characters, and I would like to some day support unicode within the database format. Your few hour estimate is not accurate for this change considering there are several separate screens for entering data (forms screen, edit record, edit popups, edit views, etc), and about half a dozen supported platforms (Windows, Mac, iPhone, iPad, Android, etc). So multiply your change accordingly and you'll see its not quite as simple!
If it were a single app with a single edit screen I'd agree, but then I'd also just change the format to support unicode since the work there would not be that exhaustive and the benefit would be greater!

I also have a list of higher priority items to work on during the few hours I have per month to write code for HanDBase. These items affect a much larger population of users (eg, iOS12 compatibility) so I'm just being completely honest when I say I don't have any plans to address this any time soon.


Re: Bug - accented characters

PostPosted: Wed Sep 05, 2018 1:29 am
by vlapr
It's pity that such problem can not be solved for more than five years :(

Do you really think that parsing input text for unsupported characters significantly decrease speed?

Can you at least check, why the whole text is lost when it contains some unsupported characters? It looks like wrong exception handling of the text processing.

Maybe you find out that the problem is somewhere deep in the system and instead of changing each input screen will be enough to put the filter at one place (eg. persist procedure, before post event, etc.)

I understand your opinion regarding priority but from my point of view it is high priority problem because it made the software unusable for me and also for others non latin language writers.

Thanks and regards

Re: Bug - accented characters

PostPosted: Sun Sep 09, 2018 10:51 am
by ddhsoftwareadmin

Thanks for your reply. I agree that it is a pity, I'd love to be able to support non latin languages with HanDBase, but the format itself is restricted in this way. So the problem is self-limiting in that the program itself doesn't support those who need to store data in non latin languages. So enhancing the handling of the use of non latin characters doesn't really make sense.

I would much rather spend my extremely limited dev time in a month adding features that benefit those who the software is targeted for, or in adding proper support for it in the first place!