Choosing an office suite

H

Hilary Bailey

I am trying to decide which office suite to choose from. The only
exposure I have had is Microsoft's, but now I have the choice between
OpenOffice.org and LibreOffice. I have had lengthy discussions with
great individuals from the community regarding writing an education
software having a database component as its main focus. Briefly, as a
novice to programming (despite having read a book on Ruby, which
offered the "sky", but ended-up confusing reality to that of
programmer's
dream), I have decided, based on various discussions with the community,
to familiarize myself with: a) HTTP, b) CSS, c) vim, d)already created
office suites by the community,and then e) Rails or JRails - as a way of
deploying such software.

I know that there are many surprises and my writing shows an innocence,
which will eventually receive a "reality check". But in the meantime,
based on my line of thinking, can anyone advice me on 1) the difference
between the suites mentioned above and 2) my plan of action in deploying
such product.

Thank you Open Source community for your past and continued help.

Hilary
 
H

Hassan Schroeder

...., I have decided, based on various discussions with the community,
to familiarize myself with: a) HTTP, b) CSS, c) vim, d)already created
office suites by the community,and then e) Rails or JRails - as a way of
deploying such software.

Perhaps you can explain what you intend to do with this "office suite"?
It certainly has no relationship whatsoever to deploying a Rails app.
 
M

Mike Stephens

Hi Hilary

If you are starting from square one, I would honestly recommend you look
at Mendix (see http://www.mendix.com/ ).

You don't need to code anything. You draw 'models'. Draw your database
as an entity-relationship diagram and you have a working database
(choose Oracle, MySQl etc) with professional CSS update etc screens -
much more advanced than Rails. Drag and drop your controls. Your
micro-logic you do with graphics.

It's free to develop but then you would incur costs to deploy - but
against that you don't have to cross the even bigger divide ie setting
up and configuring your infrastructure. You just get them to host it
probably at the press of a button. I don't know what the cost per
transaction would be. I'll try and get an idea from a contact I have in
the UK, if you're at all interested.

Office packages are wonderfully powerful but complex in their own way
and not easy to deploy as web apps.
 
P

Phillip Gawlowski

I know that there are many surprises and my writing shows an innocence,
which will eventually receive a "reality check". But in the meantime,
based on my line of thinking, can anyone advice me on 1) the difference
between the suites mentioned above and 2) my plan of action in deploying
such product.

There is, for now, little difference between OpenOffice and
LibreOffice, seeing as LibreOffice forked only recently off of
Oracle's code base.

Considering, however, that pretty much all of the former OpenOffice
developers went over to LibreOffice, I'd stick with that. Especially
since Oracle hasn't shown so far to be trustworthy (3 OSS projects
have "split" from Oracle already: MySQL -> MariaDB, Hudson CI ->
Jenkins CI, OpenOffice -> LibreOffice; and Oracle isn't one to honor
promises, either made by their acquisitions (OpenSolaris isn't Free
anymore), nor promises they made themselves (see the Java Community
Process hubub)).

2) You can't really use Rails to deploy software to a client. It would
be possible, but it'd require providing some sort of client software,
and at that point you can use something simpler than use a web
framework (like password protected directories accessed via a GUI that
enables easy installation; similar to MS's Web Platform Installer, or
Ubuntu's Aptitude).


The question is: What is it you want to achieve?

--
Phillip Gawlowski

Though the folk I have met,
(Ah, how soon!) they forget
When I've moved on to some other place,
There may be one or two,
When I've played and passed through,
Who'll remember my song or my face.
 
H

Hilary Bailey

Phillip Gawlowski wrote in post #980112:
There is, for now, little difference between OpenOffice and
LibreOffice, seeing as LibreOffice forked only recently off of
Oracle's code base.

Considering, however, that pretty much all of the former OpenOffice
developers went over to LibreOffice, I'd stick with that. Especially
since Oracle hasn't shown so far to be trustworthy (3 OSS projects
have "split" from Oracle already: MySQL -> MariaDB, Hudson CI ->
Jenkins CI, OpenOffice -> LibreOffice; and Oracle isn't one to honor
promises, either made by their acquisitions (OpenSolaris isn't Free
anymore), nor promises they made themselves (see the Java Community
Process hubub)).

2) You can't really use Rails to deploy software to a client. It would
be possible, but it'd require providing some sort of client software,
and at that point you can use something simpler than use a web
framework (like password protected directories accessed via a GUI that
enables easy installation; similar to MS's Web Platform Installer, or
Ubuntu's Aptitude).


The question is: What is it you want to achieve?

--
Phillip Gawlowski

Though the folk I have met,
(Ah, how soon!) they forget
When I've moved on to some other place,
There may be one or two,
When I've played and passed through,
Who'll remember my song or my face.

Hi Phillip,
What I want to create is a database that can measure the performance of
all entities in a school district. The closest software that exhibits
some semblance is that of Microsoft Access. Where, as I understand it,
the input entry of a single data can be housed and then derived, through
a set of queries, then further analyzed through/by Microsoft Solver
software.

The difference with my proposal would be that based on selected
indicators [which will be dynamically influenced by changed event(s) and
policy(ies], which would be able to measure success. I have been exposed
to a statistical software named SPSS and having worked as an economist,
has influenced my outlook on creating an approach/database/software
which would indicate in real time, measured results.

As you can tell, there is an element of nervousness regarding saying
too much. But on the other hand, if not much is said, not much help can
be given. So it's a "catch 24", where since the last 20 years I have
been improving on a systems that would be able to measure defined
academic output, vis-a-vis, financial constraints etc..

Mike Stephens recommended Mendix as a possible solution to my woes. Do
you know of such arena?

Therefore, I figured that, doing it all by myself may be the best
solution. However, some of my concerns are: "Why reinvent the wheel?',
How can I create a sustainable system that does not compromise quality?,
What curriculum structure should I follow that will meet my needs
without, straying from my goals?

Therefore, this is my dilemma, which seems to be going in circles. Any
suggestions.
 
S

Sebastián Rotta Seletti

[Note: parts of this message were removed to make it a legal post.]

2011/2/7 Hilary Bailey said:
Phillip Gawlowski wrote in post #980112:
There is, for now, little difference between OpenOffice and
LibreOffice, seeing as LibreOffice forked only recently off of
Oracle's code base.

Considering, however, that pretty much all of the former OpenOffice
developers went over to LibreOffice, I'd stick with that. Especially
since Oracle hasn't shown so far to be trustworthy (3 OSS projects
have "split" from Oracle already: MySQL -> MariaDB, Hudson CI ->
Jenkins CI, OpenOffice -> LibreOffice; and Oracle isn't one to honor
promises, either made by their acquisitions (OpenSolaris isn't Free
anymore), nor promises they made themselves (see the Java Community
Process hubub)).

2) You can't really use Rails to deploy software to a client. It would
be possible, but it'd require providing some sort of client software,
and at that point you can use something simpler than use a web
framework (like password protected directories accessed via a GUI that
enables easy installation; similar to MS's Web Platform Installer, or
Ubuntu's Aptitude).


The question is: What is it you want to achieve?

--
Phillip Gawlowski

Though the folk I have met,
(Ah, how soon!) they forget
When I've moved on to some other place,
There may be one or two,
When I've played and passed through,
Who'll remember my song or my face.

Hi Phillip,
What I want to create is a database that can measure the performance of
all entities in a school district. The closest software that exhibits
some semblance is that of Microsoft Access. Where, as I understand it,
the input entry of a single data can be housed and then derived, through
a set of queries, then further analyzed through/by Microsoft Solver
software.

The difference with my proposal would be that based on selected
indicators [which will be dynamically influenced by changed event(s) and
policy(ies], which would be able to measure success. I have been exposed
to a statistical software named SPSS


Hi Hilary:
Semi-off topic but.... Exists PSPP (free, libre) for users of proprietary
program SPSS

Take a look:

http://www.gnu.org/software/pspp/


and having worked as an economist,
has influenced my outlook on creating an approach/database/software
which would indicate in real time, measured results.

As you can tell, there is an element of nervousness regarding saying
too much. But on the other hand, if not much is said, not much help can
be given. So it's a "catch 24", where since the last 20 years I have
been improving on a systems that would be able to measure defined
academic output, vis-a-vis, financial constraints etc..

Mike Stephens recommended Mendix as a possible solution to my woes. Do
you know of such arena?

Therefore, I figured that, doing it all by myself may be the best
solution. However, some of my concerns are: "Why reinvent the wheel?',
How can I create a sustainable system that does not compromise quality?,
What curriculum structure should I follow that will meet my needs
without, straying from my goals?

Therefore, this is my dilemma, which seems to be going in circles. Any
suggestions.


--
==============================
SEBASTIAN ROTTA SELETTI
(e-mail address removed)
MSN:[email protected]
==============================
 
P

Phillip Gawlowski

Hi Phillip,
What I want to create is a database that can measure the performance of
all entities in a school district. The closest software that exhibits
some semblance is that of Microsoft Access. Where, as I understand it,
the input entry of a single data can be housed and then derived, through
a set of queries, then further analyzed =A0through/by Microsoft Solver
software.

That's a feature all relational databases share: You store data, query
the data, and do something with the results.

However, you don't *need* Access (which is an expensive toy to deploy
software on for a school) for that, but a plain' ol' database.

Investigate SQLite (excellent for "embedded" databases, since it's
light-weight and can be used from pretty much every programming
language, and is Public Domain) if you want to write software that
ends up on a PC, or whatever is popular on the web: MySQL/MariaDB or
PostgreSQL.
The difference with my proposal would be that based on selected
indicators [which will be dynamically influenced by changed event(s) and
policy(ies], which would be able to measure success. I have been exposed
to a statistical software named SPSS and having worked as an economist,
has influenced my outlook on creating an approach/database/software
which would indicate in real time, measured results.

Sounds like a standard use-case for OLAP:
http://en.wikipedia.org/wiki/Online_analytical_processing
Mike Stephens recommended Mendix as a possible solution to my woes. Do
you know of such arena?

I don't trust tools that claim to write software for you. ;)

Mind, a visual tool can be very helpful (I like to easily visualize
SQL databases and their key relations, for example), but I'm quite
sure that Mendix itself won't help you in your case: There's no
business process per se in what you want to roll out, and all those
"UML to Software" tools have fallen flat, requiring manual
"optimization".
Therefore, I figured that, doing it all by myself may be the best
solution. =A0However, some of my concerns are: "Why reinvent the wheel?',
How can I create a sustainable system that does not compromise quality?,
What curriculum structure should I follow that will meet my needs
without, straying from my goals?

Therefore, this is my dilemma, which seems to be going in circles. Any
suggestions.

Yes. Incorporate a business before too long, and start looking for a
technical co-founder now. What you want to do isn't impossible for a
single person. Said person needs a bit of experience in a lot of
technologies. While you can grab database engines, libraries, etc. off
the shelf, you have to check licensing (Anything GPL licensed will
*force* you to publish your own source code, for example, while the
LGPL doesn't have this problem), learn and deal with a lot of
technologies (databases, client / server computing or web programming,
OLAP, reporting, data entry), which is... well, just a tad much to get
started with a business of any sort.

And you have a good litmus test for a technical person: Can they and
do they want to teach you to code. :)

Mind, you should find a technical person you can, push comes to shove,
bind with an NDA either way, simply to buy expertise in rather tough
areas that you shouldn't deal with as a beginner, like application
security (doubly so if you want to launch a website!).

--=20
Phillip Gawlowski

Though the folk I have met,
(Ah, how soon!) they forget
When I've moved on to some other place,
There may be one or two,
When I've played and passed through,
Who'll remember my song or my face.
 
S

Stu

I would choose LibreOffice due to recent events since sun's buyout.

As it's been mentioned an office suite is not really the correct tool
for creating a ruby on rails site.

Here is a short list of tool/software to look into:

1) Ubuntu Linux (the laymans linux)
2) rvm Ruby Version Manager
3) vim (my preferred text editor)
4) shell terminal ( once again preference. I like zsh but ubuntu defaults bash)
5) browser (firefox, opera, chrome... take one or use them all)
6) database (sqlite3 for development, postgres for deployment ...
mysql falls into the same area of sun/oracle debacle but it's still a
choice)
7) installing ruby on rails gem

If you need help getting something like this going I would be more
than happy to walk you through the steps to getting the proper
environment for you to experiment and learn on.

Also look in your neighborhood for ruby sigs and meetings. Definitely
the place to network with others and have someone show you directly.

As others have mentioned avoid tools that lock you in, don't run on
everything, are closed and/or attempt to charge you to learn/use.
 
M

Mike Stephens

Hilary Bailey wrote in post #980168:
What I want to create is a database that can measure the performance of
all entities in a school district.

Your question is interesting because unlike almost all others on this
forum, you have no real committment to Ruby. You're just looking for the
easiest route to solving your problem.

As a result it rather depends precisely what your problem is.

A database cannot measure performance. You might mean you would like to
capture something like grade performance and then analyse that by
teacher, course, school, curriculum, student entry grading - various
things like that.

The issue is do you want to do that yourself, sell a package that other
groups could use, or set up a web site that offers a central service?

These involve very different challenges and skill requirements.

By the way Access is brilliant - I've never seen another database come
close. I understand Bill Gates had a soft spot for it. It's a half
decent relational database (Jet) but more than that it is a Query By
Example environment. After that it does reports and forms etc.

Access (like Excel) is a functional programming environment. Personally
I think it is easier to pick these up than OO platforms like Ruby.

However, Access (like Excel) is no good for Web deployment simply
because
Microsoft went off down the geeky .NET and Sharepoint routes and left
simple folk high-and-dry.

Bear in mind Phillip's remarks about Mendix are without any knowledge of
it.
 
P

Peter Hickman

technologies. While you can grab database engines, libraries, etc. off
the shelf, you have to check licensing (Anything GPL licensed will
*force* you to publish your own source code, for example, while the
LGPL doesn't have this problem), learn and deal with a lot of

Ok you are clearly channeling Steve Balmer here. Stop with the bullshit.

The only requirement to redistribute your source is when you modify
the source and then distribute the resulting product.

If you use GPLd code (say to parse XML or access a database) as part
of your product then you have no requirement under the GPL to
redistribute the your source code unless YOU HAVE MODIFIED THE GPLd
CODE. And only then you only have to release the source of the
modified code and not the whole product.

It is in fact very rare for anyone to need to redistribute the source
code of GPLd code. It usually only happens when you take a GPLd
product and then modify it and release that as your own product.

If your product is an online service then you are not distributing the
product and there is no requirement to release your source code.

A more pertinent issue is escrow, customers may want your code placed
in escrow to ensure that should you go under they can continue to use
and develop the product that they depend on.

See a lawyer about that.
 
E

Everett L Williams II

[Note: parts of this message were removed to make it a legal post.]

Peter said:
Ok you are clearly channeling Steve Balmer here. Stop with the bullshit.



A more pertinent issue is escrow, customers may want your code placed
in escrow to ensure that should you go under they can continue to use
and develop the product that they depend on.

See a lawyer about that
Peter,

What you have said about GPL is largely true, but don't count on
escrowed code to be available to you or anyone but the successor owner
of the patents and copyrights. Bankruptcy courts have consistently ruled
that the source code is a capital asset that must be sold to satisfy the
various people who apply to the court for redress. This overrides all
such escrow contracts. Most companies specifically state that you do not
own the product that you are paying for, only a right to use it. If you
owned it, you might have some redress, but that opens up another whole
can of worms for the company actually owning the rights. There have been
many attempts to get around this, but I know of no specific technique
that has consistently worked. If the source is considered abandoned or
has been legally published into the public domain, then anyone can use
it, but that is unlikely in any commercial situation that is not
deliberately GPL or some similar situation. It is difficult but not
impossible for a public corporation to release commercially valuable
code into the public domain or to GPL, because it dilutes shareholder
value, so the company must show that they receive value for what they
release, which can be done.

The real crimes and difficulties arise in the area of software patents,
which should be banned as pernicious. Though IANAL, I dealt with
copyright issues for a billion dollar company in my past, coaching the
stupid lawyers on the issues. I will say that the Millennium Copyright
Act is an almighty screwup, designed to keep Disney from losing Mickey,
while screwing up almost everything else.

Everett L(Rett) Williams II
 
E

Everett L Williams II

[Note: parts of this message were removed to make it a legal post.]

Mike said:
Hilary Bailey wrote in post #980168:

Your question is interesting because unlike almost all others on this
forum, you have no real committment to Ruby. You're just looking for the
easiest route to solving your problem.

As a result it rather depends precisely what your problem is.

A database cannot measure performance. You might mean you would like to
capture something like grade performance and then analyse that by
teacher, course, school, curriculum, student entry grading - various
things like that.

The issue is do you want to do that yourself, sell a package that other
groups could use, or set up a web site that offers a central service?

These involve very different challenges and skill requirements.

By the way Access is brilliant - I've never seen another database come
close. I understand Bill Gates had a soft spot for it. It's a half
decent relational database (Jet) but more than that it is a Query By
Example environment. After that it does reports and forms etc.

Access (like Excel) is a functional programming environment. Personally
I think it is easier to pick these up than OO platforms like Ruby.

However, Access (like Excel) is no good for Web deployment simply
because
Microsoft went off down the geeky .NET and Sharepoint routes and left
simple folk high-and-dry.

Bear in mind Phillip's remarks about Mendix are without any knowledge of
it.
You must be kidding...Access has holes in it that large trucks can be
driven through, and the "Jet" engine must have flamed out in a
rainstrom, because performance outside of single user setups is dismal.
It also does not have referential integrity nor secure record locking.
The problem is that if you use the integrated tools to build a product
or service, when you try to scale it, you basically have to start over.
If you use external methods to use it, the problem is less, but with
what is available for free out there in the SQL world, not to mention a
few other relational DB's that can scale, why would you use Access. Ugh.

If you just must spend money on a package deal, then go buy a copy of
Alpha 5, which can create both local and web apps and has a really nice
IDE. If you like rolling your own, then Ruby of one form or another plus
a free DB can be yours for nothing plus however much you want to spend
on doc, but avoid Access for anything but one-off, single user apps for
personal use.

Everett L(Rett) Williams II
 
M

Mike Stephens

Everett L Williams II wrote in post #980275:
driven through, and the "Jet" engine must have flamed out in a
rainstrom, because performance outside of single user setups is dismal.
It also does not have referential integrity nor secure record locking.
We're trying to help a person who doesn't really want to grapple with
all sorts of technolgies. Access is much more... accessible.

I think there's a bit of snobbishness here. SQlite is trendy whereas
Access is a rank loser. Well Jet has sophisticated view mechanisms,
supports complex sub-queries, does transactions, referential integrity,
security, replication etc etc. I've worked in some of the largest
financial companies in the World and you'd be surprised how many big
multi-site many-user Access systems are in use everyday.

The attraction of tools like Access (and Excel) for non-professionals is
you can experiment by seeing all the time what you are programming and
what effects it is having.

It's not Ruby but don't dismiss this style of IT.
 
H

Hilary Bailey

What I want to create is a product that is computer accessible, that is
similar to a teacher's Grade Keeper, but goes farther by adding defined
info from principals, guidance counselors, lunch provision for students,
etc.. The reason for this approach is that this will now allow more
complete view of a what impacts a child.

Therefore, the daily entry of data from all participants (teachers,
principals, security, janitors, etc...) will give an analyzer a wider
set of defined parameter inputed data to access, then analyze. The
problem is where to start. I read a book on Ruby, some say that my next
step is to play with scripts, alter some commands and then test such
adjustments. The problems is to follow a logical sequence of learning.
For example, since i use Windows 7, have installed Ruby 1.9.2 p.136 and
Rails, Vim7.2, and LibreOffice 3.3. and saved info to htmldog.com from
which HTML & CSS can be learned.

I know very well that it will take me some time, however, now where do I
start? Should I star with htmldog tutorials, then open Rails along with
vim7.2, then the next stage will be to explore SQlite, then MYSQL, while
having LibreOffice Base as a source of reference?

In terms of distribution, giving it away free will not be taken
seriously by current educational administrators and policy makers. It
suits me to market it and if successful, support this community plus
other social causes of choice.

So based on all that have been said, where, specifically (if possible)
go from here in creating such a product.
 
S

Sam Duncan

The first thing I would do for something like this (taking the pragmatic
approach), is design your information schema. You need to design (not
code) your schema such that it provides the kind of analysis you are
after with the least amount of work. It also needs to be flexible in the
sense that you will inevitably want to start storing new kinds of
information, and analysing the data you have in new ways. The data is,
after all, the basis and purpose of your application.

Whatever you decide to do it in, and I am guessing you will end up with
a conventional relational and hopefully free database, you can assume
that it will be comprised of tables. Tables will contain fields. The
tables and their fields will represent some encapsulation of an entity.
The tables, and the fields therein, will have relationships with other
tables and their fields which will provide you with the basis for your
analysis and reporting.

If you are looking to roll your sleeves up and actually get started, do
so at the data end. In that scope the collection, presentation, and
reporting layers are largely irrelevant, that is to say; your data store
shouldn't really care what is accessing it. Avoid the IDE and framework
religious wars, don't bother with logos and About Us pages. Dive into
your data store schema with a pencil and a notepad today. Anybody here
who has used a database of some kind (and there appear to be plenty),
will be able to help you turn your sketch into a real implementable
schema. Of course, you will inevitably have to wade through ever more
esoteric hooha about the specifics of that implementation ;]

Pretty much I'm saying that I'm not a rocket scientist, and if I wanted
to build a rocket, I would intuitively start with determining how I
would power it.

Sam
 
P

Phillip Gawlowski

Ok you are clearly channeling Steve Balmer here. Stop with the bullshit.

The only requirement to redistribute your source is when you modify
the source and then distribute the resulting product.

No, the requirement is to publish *derivative* code. And since the GPL
doesn't outline under which circumstances a specific work becomes
derivative, with no court cases to clarify the issue, the question is
out in the open:

http://www.gnu.org/licenses/gpl-faq.html#NFUseGPLPlugins

Hence the LGPL: It makes clear that using a library doesn't taint *your* code.
If your product is an online service then you are not distributing the
product and there is no requirement to release your source code.

The Affero GPL *does* establish such a requirement.
See a lawyer about that.

Absolutely. Or use libraries licensed with a non-copyleft license
(BSD, MIT, and similar).

--
Phillip Gawlowski

Though the folk I have met,
(Ah, how soon!) they forget
When I've moved on to some other place,
There may be one or two,
When I've played and passed through,
Who'll remember my song or my face.
 
E

Everett L Williams II

[Note: parts of this message were removed to make it a legal post.]

Mike said:
Everett L Williams II wrote in post #980275:

We're trying to help a person who doesn't really want to grapple with
all sorts of technolgies. Access is much more... accessible.

I think there's a bit of snobbishness here. SQlite is trendy whereas
Access is a rank loser. Well Jet has sophisticated view mechanisms,
supports complex sub-queries, does transactions, referential integrity,
security, replication etc etc. I've worked in some of the largest
financial companies in the World and you'd be surprised how many biga
multi-site many-user Access systems are in use everydayAnd, .

The attraction of tools like Access (and Excel) for non-professionals is
you can experiment by seeing all the time what you are programming and
what effects it is having.

It's not Ruby but don't dismiss this style of IT
I was dismissing it because there are better, faster, simpler tools out
there that are equivalent in cost and cheaper when distributed. And,
referentiaql integrity is only maintained if you check all the right
boxes and know about the holes that exist and correct for them. And
record and page level locking is such a mess that I will never trust it,
having been lied to by MS on so many occasions on the subject. They say
it works, and then you find out the list of exceptions and so forth and
so on, and so often, it is not in your control. The Jet engine was
borrowed from Fox technology when they bought that company, and has been
sadly neglected ever since. I really hate to trust MS on anything about
Access.

Everett L(Rett) Williams II
 
J

Justine B.

Phillip Gawlowski wrote in post #980190:
Hi Phillip,
What I want to create is a database that can measure the performance of
all entities in a school district. The closest software that exhibits
some semblance is that of Microsoft Access. Where, as I understand it,
the input entry of a single data can be housed and then derived, through
a set of queries, then further analyzed through/by Microsoft Solver
software.

That's a feature all relational databases share: You store data, query
the data, and do something with the results.

However, you don't *need* Access (which is an expensive toy to deploy
software on for a school) for that, but a plain' ol' database.

Investigate SQLite (excellent for "embedded" databases, since it's
light-weight and can be used from pretty much every programming
language, and is Public Domain) if you want to write software that
ends up on a PC, or whatever is popular on the web: MySQL/MariaDB or
PostgreSQL.
The difference with my proposal would be that based on selected
indicators [which will be dynamically influenced by changed event(s) and
policy(ies], which would be able to measure success. I have been exposed
to a statistical software named SPSS and having worked as an economist,
has influenced my outlook on creating an approach/database/software
which would indicate in real time, measured results.

Sounds like a standard use-case for OLAP:
http://en.wikipedia.org/wiki/Online_analytical_processing
Mike Stephens recommended Mendix as a possible solution to my woes. Do
you know of such arena?

I don't trust tools that claim to write software for you. ;)

Mind, a visual tool can be very helpful (I like to easily visualize
SQL databases and their key relations, for example), but I'm quite
sure that Mendix itself won't help you in your case: There's no
business process per se in what you want to roll out, and all those
"UML to Software" tools have fallen flat, requiring manual
"optimization".
Therefore, I figured that, doing it all by myself may be the best
solution. However, some of my concerns are: "Why reinvent the wheel?',
How can I create a sustainable system that does not compromise quality?,
What curriculum structure should I follow that will meet my needs
without, straying from my goals?

Therefore, this is my dilemma, which seems to be going in circles. Any
suggestions.

Yes. Incorporate a business before too long, and start looking for a
technical co-founder now. What you want to do isn't impossible for a
single person. Said person needs a bit of experience in a lot of
technologies. While you can grab database engines, libraries, etc. off
the shelf, you have to check licensing (Anything GPL licensed will
*force* you to publish your own source code, for example, while the
LGPL doesn't have this problem), learn and deal with a lot of
technologies (databases, client / server computing or web programming,
OLAP, reporting, data entry), which is... well, just a tad much to get
started with a business of any sort.

And you have a good litmus test for a technical person: Can they and
do they want to teach you to code. :)

Mind, you should find a technical person you can, push comes to shove,
bind with an NDA either way, simply to buy expertise in rather tough
areas that you shouldn't deal with as a beginner, like application
security (doubly so if you want to launch a website!).

--
Phillip Gawlowski

Though the folk I have met,
(Ah, how soon!) they forget
When I've moved on to some other place,
There may be one or two,
When I've played and passed through,
Who'll remember my song or my face.


Mendix doesn't claim to write software for you. In fact, no code is
generated; the visual model is the application! You can extend the
application with Java if you so choose, but 99% of the time you don't
need to. You can build a complete working application (including the
data, presentation, and logic layers and security) in just days with
one-click deployment.

In this case you can use Mendix to setup a view of your data within 5
minutes ... just download a free trial at http://www.mendix.com (Try it
Now button). If you wish to view external data, the free database
replication widget is available in the app store.

Or attend Mendix training (there is one next week in Boston) and they
will show you how! http://www.mendix.com/company/events/
 
H

Hilary Bailey

Sam Duncan wrote in post #980397:
Sam, I think you are correct in your assessment. In terms of the
database, should I indulge in LibreOffice Base or use Sqlite/SQL? After
such decision,is it possible to add components from PSPP and LibreOffice
spreadsheet plus others, to create a software? In other words, just like
making a car, where we add different parts to create a product, can the
same logic be used to create a software.

The first thing I would do for something like this (taking the pragmatic
approach), is design your information schema. You need to design (not
code) your schema such that it provides the kind of analysis you are
after with the least amount of work. It also needs to be flexible in the
sense that you will inevitably want to start storing new kinds of
information, and analysing the data you have in new ways. The data is,
after all, the basis and purpose of your application.

Whatever you decide to do it in, and I am guessing you will end up with
a conventional relational and hopefully free database, you can assume
that it will be comprised of tables. Tables will contain fields. The
tables and their fields will represent some encapsulation of an entity.
The tables, and the fields therein, will have relationships with other
tables and their fields which will provide you with the basis for your
analysis and reporting.

If you are looking to roll your sleeves up and actually get started, do
so at the data end. In that scope the collection, presentation, and
reporting layers are largely irrelevant, that is to say; your data store
shouldn't really care what is accessing it. Avoid the IDE and framework
religious wars, don't bother with logos and About Us pages. Dive into
your data store schema with a pencil and a notepad today. Anybody here
who has used a database of some kind (and there appear to be plenty),
will be able to help you turn your sketch into a real implementable
schema. Of course, you will inevitably have to wade through ever more
esoteric hooha about the specifics of that implementation ;]

Pretty much I'm saying that I'm not a rocket scientist, and if I wanted
to build a rocket, I would intuitively start with determining how I
would power it.

Sam
 
E

Everett L Williams II

[Note: parts of this message were removed to make it a legal post.]

*Hiloary,

Sam's advice is sound as far as it goes, but if you are as you represent
yourself, your odds of producing a salable project any time in the next
couple of years seems remote, regardless of the tool you choose. That is
not meant as an insult, but an evaluation of your questions. Find that
tech pro that has been spoken of here, and let that person select the
tools that they know well enough to accomplish the task or at least take
a long look at their advice before selecting tools. Rather than try to
build any technical data descriptions, I would say that you should write
up as detailed a description as you can manage of the data that you
believe must be collected and from where it is to be collected, and also
write up as detailed a description as you can of the evaluations that
you want to be built into the product and those that you want to be
available dynamically. Most of the targets that you will have in this
situation will probably want to export selected data to a spreadsheet
and than manipulate that data with said spreadsheet. If you want to
select LibreOffice for that or MS's product, that will be something that
you can probably leave to the users. Standard analysis can be built into
the product. I really see this as a DB with data entry forms and a few
tools for data selection. That can be done directly in several of the
free SQL products, or if you want to build a web based product, the
something like Alpha 5 from Alphasoftware.com could manage it in a very
short window of time. Doing large amounts of custom programming on this
would seem to be a self defeating effort.

Everett L(Rett) Williams II
*
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Members online

Forum statistics

Threads
473,731
Messages
2,569,432
Members
44,835
Latest member
KetoRushACVBuy

Latest Threads

Top