![[TEST] Document Foundation Community](https://communitytest.documentfoundation.org/images/discourse-logo-sketch.png){kind=logo}
NoOp wrote
...
What database doe you think works better than OpenOffice or LibreOffice?
Pick.
I don't have any views on this matter.
I vaguely remember that Lotus Approach allowed far more flexibility in
allowing updates to multiple tables, but that was a long time ago and is no
longer available (AFAIK).
Access allows update queries, which are sadly missing from LO/OO.
Am 17.04.2012 10:52, ptoye wrote:
Access allows update queries, which are sadly missing from LO/OO.
Why can't you run SQL update queries? This is trivial to do in SQL. The SQL is trivial to execute by macro.
If you desparately want a graphical query designer for update/insert/delete queries, then you should write one and store your queries somewhere in the .odb container.
After using OO/LO Base for over 5 years now I gave up last month after the latest "crash". It has been an ongoing battle to keep my DB operational and I was really fed-up with all the crashes, loss of data and time spent fixing the DB etc etc over the years.
I'm NOT running a fancy DB here - just a Membership List with about 2500 records, each record with three photos and approx 70 fields. A bunch of Queries and a few Reports so it's not exactly "Rocket Science"!!
I came to the very sad conclusion, (after my last "crash" when Base managed to randomly drop about a quarter of my Data (not complete records but fields in the records), and scramble about another third!!!!!!! AND it would no longer even run my 'User Form'!!), that Base is NOT fit for purpose. This is MY opinion - and I think 5 years is a long enough time to get a good "feel" for it - but no doubt I will get "flamed" for this comment.
LONG story short........ I have now moved my (re-built) Database on to Kexi - and what a difference!!!! It's like getting out of an old broken-down VW (or other make) of car and getting into a brand new Jaguar and driving down the highway at 100mph!! And I know what that feels like as I did just that last year in the UK!! Yes there is a lot to learn and work out but the overall effect is like Chalk and Cheese!!
For anyone interested I can recommend this DB - some of the more fancy options are only due out in future releases but for a basic DB job it "just works"!!
I'm NOT missing Base!!
Ian Whitfield
Pretoria
Andreas Säger wrote
Why can't you run SQL update queries? This is trivial to do in SQL. The
SQL is trivial to execute by macro.
If you desparately want a graphical query designer for
update/insert/delete queries, then you should write one and store your
queries somewhere in the .odb container.
Of course I can. If I can find out how to do it.
But this isn't the point, is it? The real point is that database front-ends
and macro languages are meant to make life easy, not difficult. You seem to
dislike "sugar" - worried about your diameter? The computer was made for
man, not man for the computer. Why not program in assembler - gives you much
more control and isn't really that hard (I speak from years of assembler
experience and with my tongue firmly in my cheek)?
Hi 
I guess you tried using various external back-ends? Did you consider using one of the back-ends without having Base as the front-end?
People have already suggested that Kexi is not even as 'advanced' as Base but that was about a year ago so it's quite possible they were wrong or that things have changed. It's good to have something i can recommend to people that are unhappy with Base because Base is clearly not ready and doesn't have enough devs working on it.
Regards from
Tom
Hi
Dohhh, that last post was meant to be off-list!!
Regards from
Tom
Am 17.04.2012 11:34, ptoye wrote:
Andreas Säger wrote
Why can't you run SQL update queries? This is trivial to do in SQL. The
SQL is trivial to execute by macro.
If you desparately want a graphical query designer for
update/insert/delete queries, then you should write one and store your
queries somewhere in the .odb container.Of course I can. If I can find out how to do it.
But this isn't the point, is it? The real point is that database front-ends
and macro languages are meant to make life easy, not difficult. You seem to
dislike "sugar" - worried about your diameter? The computer was made for
man, not man for the computer. Why not program in assembler - gives you much
more control and isn't really that hard (I speak from years of assembler
experience and with my tongue firmly in my cheek)?-----
You get what you pay for. The rest is about community. Unfortunately, this office suite has more consumers and complainers than contributors.
Regarding your remark on assembler code: A very high software stack imposes a new quality of problems that would be unknown on hardware level.
Fortunately, there is still the OpenOffice.org community:
http://extensions.services.openoffice.org/en/project/access2base
(which could be even more useful if someone would rewrite these Basic routines as a true extension of UNO-components)
Am 17.04.2012 11:28, Ian Whitfield wrote:
>> >> What database doe you think works better than OpenOffice or
LibreOffice?
>>I don't have any views on this matter.
After using OO/LO Base for over 5 years now I gave up last month after
the latest "crash". It has been an ongoing battle to keep my DB
operational and I was really fed-up with all the crashes, loss of data
and time spent fixing the DB etc etc over the years.
Beware of the troll!
Hi
Not really a troll. The chap has worked harder than most to try to get Base more widely accepted both within TDF and outside and has tried to be positive about it. However it's good to hear he has found success with a different OpenSource program. It's also good to hear that Kexi appears to be better than it was.
Regards from
Tom
Well said Peter!
I totally and completely agree with you and Ian Whitfield.
And thanks to Ian for the tips about Kexi – I am almost ready for anything but LibO-Base.
Since the beginning of this year I have been trying to create a similar member register about a boy scout organization’s senior leaders also collecting memorable info from gone days – before these leaders are gone. The most I got from the efforts is waste of time and frustration.
A week ago Dan Lewis kindly sent to me drafts on documentation about creating a Base-db. It is very well written, clear, logic, easy to follow – but it is a draft, written in 2010 for OpenO. As far as I could see, there are nothing similar for LibO-Base yet.
Regarding LibO-Base’s situation today, very much should be helped if that Dan’s documentation should be completed and finished and included in the LibO’s (Base) introduction.
(When following Dan’s documentation I also noticed that in LibO-Base you can create a form /_based on a query_/ only using the Form Wizard – Design View allows you to create a form /_only based on a table = one table!_/)
To the LibO-folks and developers I repeat:
Stop developing LibO-Base;go back to the latest stable version and start to complete it and repair all discrepancies in the program and between the program and it’s documentation!
There is so much good done in LibO-Base so why not get the rest OK? So it at least should meet the promises in advertising.! Without problems!
And:
What (in @!?*@!??) can be the clever idea with having in LibO-base a embedded HSQL v.1.8 that is not supported any more when the latest version is 2.2.x? What about e.g. SQLite?
As Peter said:thus making the life easier -- and if not easier but not more difficult!
Regards
Pertti Rönnberg
Hi
Almost no-one is developing Base or at least developments are not
reaching the releases. So LO Base is almost no different from OOo's
Base except that it might have some regressions from the developments
of other components of LO.
I get the impression that there is no official guide for OOo's Base and
so i am hoping that the documentation being written for LO/OOo Base
will help the devs work out what needs to be worked on or at least how
Base should be working.
The BoD are against the idea of pulling in expertise from the various
back-ends that could work well with Base. Someone did approach
Postgresql and they seemed positive but they need to break through the
reluctance of the BoD.
I figured that we might get some feedback if people here started using
MariaDb. Obviously people from MySql are unlikely to help as MySql is
owned by Oracle. However the database experts here would rather
promote MySql which is only ever going to be a one-way street.
I had hoped that other people would approach other projects such as (crucially) HSqlDb.
Collaborating about Base is difficult because the experts and required
skills are scattered across several different mailing list and none are
willing to join another high-traffic list that hardly ever deals with
Base. The BoD say they have more important issues to deal with and
can't spare the resources of setting up a mailing list dedicated to
Base. - After all the other apps/modules of LO don't have their own
mailing lists and don't need them so why treat Base any different?
The BoD's big plan for getting Base up to scratch is to sit&wait to
see if someone gets annoyed enough with it to start working at it but
the problems are more extensive than 1 or 2 people could handle so
every-time someone starts they get demoralised and leave. Ian is just
one more example of this.
Dan is a star for making such a good start on the documentation.
Regards from
Tom
Am 17.04.2012 16:03, Pertti Rönnberg wrote:
To the LibO-folks and developers I repeat:
Stop developing LibO-Base;go back to the latest stable version and start
to complete it and repair all discrepancies in the program and between
the program and it’s documentation!There is so much good done in LibO-Base so why not get the rest OK? So
it at least should meet the promises in advertising.! Without problems!
As far as I can see, the Base component of LibO 3.5 is identical to the one in OOo 3.3.
And:
What (in @!?*@!??) can be the clever idea with having in LibO-base a
embedded HSQL v.1.8 that is not supported any more when the latest
version is 2.2.x? What about e.g. SQLite?
Within a few minutes you can export your embedded HSQL 1.8 to an external HSQL 2.8. This is really a piece of cake.
SQLite would not solve a single problem with Base, even if somebody would manage to "integrate" this thing. It would be rather different from all the other supported databases.
HSQL is a "vanilla database". There is nothing special with its features and data types. Being a very well done Java application it is platform independent byte by byte.
Hi
Would it be easy to upgrade the HSqlDb embedded in Base or would all the tweaks need to be applied? Are the tweaks vital or could they left out?
Regards from
Tom
Am 17.04.2012 17:38, Tom Davies wrote:
Hi
Would it be easy to upgrade the HSqlDb embedded in Base or would all the tweaks need to be applied? Are the tweaks vital or could they left out?
Regards from
Tom
The easiest improvements in my honest opinion:
- Wipe out the .odb container all together and get back to the configuration of database access in OOo 1. Back to external databases with configuration data in configuration files, forms and reports in stand-alone office documents. More speed more safety, security, accessibility and transparency. People who do not understand how to connect separated software tools do not understand the .odb container neither (as we can read in all the mail merge topics).
- Remove all the half done wizards. Do not improve them. Remove! No database developer nor interface designer needs all this stinky rubbish. There are graphical tools to compose forms and reports within office documents. There is SQL for all the rest, including all the things we can not do in the current graphical interface. There are plenty of SQL editors to produce valid SQL for various databases. SQL text can be pasted into any database configuration.
- Add native database queries, so the current "direct SQL mode" returns editable row sets and the (useless or even harmful) graphical query designer can be removed as well.
- Having removed all the wizards (they are Java components) without losing any functionality at all, extensions could substitute .odb packages. When you open the extension, the database gets installed _permanently_ (rather than _temporarily_ like the .odb) into the configuration tree together with the forms and reports documents. The database will be registered and all the tables, queries, forms and reports are accessible form the data source window, hyperlinks and desktop links just like it used to be in OOo 1.
Hi
It looks like most of this can be done without touching Base itself? Is LO able to use databases in this way already? Is it something that could be passed to the devs list an/or the BoD discussion list? Something like this could be phrased to capture the imagination, Saying it's the old way might put them off.
Regards from
Tom
+ 1000
Pertti Rönnberg wrote
Well said Peter!
I totally and completely agree with you and Ian Whitfield.
Well, someone likes me! My point is that I'm really a hacker when it comes
to databases, not a professional coder, and life really needs to be made
easier.
I'm almost thinking of installing something like MySQL and writing my own
front-end in Java.
Is there any point in suggesting that LibO and OO should continue to
diverge? I'd have thought convergence would be more productive, both in
terms of the development which seems too thinly spread, and for users who
might even see a serious contender emerging.
Pertti Rönnberg wrote
Well said Peter!
I totally and completely agree with you and Ian Whitfield.
Well, someone likes me! My point is that I'm really a hacker when it comes
to databases, not a professional coder, and life really needs to be made
easier.I'm almost thinking of installing something like MySQL and writing my own
front-end in Java.
That's what I did, but front end in Apache/PHP. Take it anywhere, easily migrate, update. Data security (from fall over and future proofing. Divergence not an issue.)
Steve
Pertti Rönnberg wrote
Well said Peter!
I totally and completely agree with you and Ian Whitfield.
Well, someone likes me! My point is that I'm really a hacker when it comes
to databases, not a professional coder, and life really needs to be made
easier.I'm almost thinking of installing something like MySQL and writing my own
front-end in Java.
Try MySQL Workbench as MySQL/MariaDB front-end. It has a nice GUI interface and is FOSS. There is Ubuntu ppa for Natty and Oneiric if the version found on Oracle does not work.
There is a browser/php based GUI front-end phpMyAdmin I believe is the name.
Another front end is HeidiSQL, it is included with the Windows version of MariaDB. There is no Linux version of this front-end.
Am 17.04.2012 19:23, Tom Davies wrote:
Hi
It looks like most of this can be done without touching Base itself? Is LO able to use databases in this way already? Is it something that could be passed to the devs list an/or the BoD discussion list? Something like this could be phrased to capture the imagination, Saying it's the old way might put them off.
Regards from
Tom
I use to use Base like this. I connect a Base document to HSQL 2.8 databases with forms and reports on the desktop (spreadsheet reports are sufficient for us). The Base container is nothing but a configuration file storing the connection URL, SQL query strings and the default log-in (and I left some rarely used admin forms in it).
My users do not know anything about the database document.
The content of the database document could be some XML and query strings in the configuration tree (just like it used to be in OOo 1).