Gambas futuro o che tipo di Gambas noi vogliamo.

[Ceskho]

Per me questo è un limite. Se voglio fare manutenzione ad un software via ssh da terminale non posso.

non accedere da ssh e' limite, ma piccolino davvero

Spero tu stia scherzando. In ambienti professionali una cosa del genere è impensabile. Buona parte del mio tempo da programmatore lo passo in ssh per capire che cappero succede sui server dei clienti. anche solo infilare un debbugger nel codice di gambas è pressochè impossibile da terminale.

[milio]

Io uso ssh per aggiornare i miei software... certo con un piccolo 'trucco' se cosi' si puo' dire:

Faccio le modifiche in locale...
Compilo e invio via ssh al server...
I client, quando avviano l'applicazione, controllano se nel server risiede una versione piu' aggiornata del software, lo copiano in locale e si riavviano...

Per quanto riguarda i debug dei client, gli stessi prima di morire con il fatidico messaggio 'This application has raised una vaccata' manda un messaggio di debug all'applicazione server...

 
Questa discussione per me e' comunque... fuori discussione!
Non credo che in questo forum, a parte un ristretto gruppo di nicchia (e qui io mi escludo categoricamente) sappia usare gambas in ogni sua parte o comunque abbia la conoscenza per realizzare una vera e propria 'killer application' con qualsiasi altro linguaggio...

Innanzi tutto penso che ogni linguaggio di programmazione abbia i suoi pro e contro e comunque un suo, diciamo, campo di software realizzabili... certo con il C ci faccio di tutto (non io ovviamente ), ma per fare un gestionale o un programma per la manutenzione non ce lo vedo proprio comodo...

Comunque gambas e' cresciuto molto a mio dire e non credo che si riesca a creare un software utilizzabile per tutte le distro di linux senza dover fare modifiche del caso per farlo funzionare, ma lo vedo potente in applicazioni costruite ad hoc.

[pastrank]

Spero tu stia scherzando.

No, dico solo che Gambas e' per fare certe cose, altri linguaggi sono per farne altre. Tutto e bene non credo lo faccia nessun ambiente.

[Ceskho]

Spero tu stia scherzando.

No, dico solo che Gambas e' per fare certe cose, altri linguaggi sono per farne altre. Tutto e bene non credo lo faccia nessun ambiente.

Ci sono altri linguaggi che le cose le fanno meglio. A me dispiace dirlo ma mi sembra una cosa molto oggettiva. Gambas, ad oggi, è l'unico che permette di scrivere software con interfaccia in 12 secondi! Dovrebbero premere su questo.

[pastrank]

Ci sono altri linguaggi che le cose le fanno meglio.

Tra quelli piu' diffusi, in questo periodo sto usando abbastanza Python, alla 3.3. L'impressione, dopo qualche mese, non e' quella di pulizia: ci trovo un ondeggiamento pauroso tra la voglia di standardizzare la scrittura del codice con quella di creare "scorciatoie" nella scrittura stessa, e un po' di contraddizione tra l'estensione della libreria standard e il fatto che alcune funzioni, es quelle per le immagini (che potrei definire basilari!), proprio non funzionano!

Gambas, ad oggi, è l'unico che permette di scrivere software con interfaccia in 12 secondi! Dovrebbero premere su questo.

Qui sono d'accordo. Capisco le voglie di fare cose "strane" per i programmatori, ma a volte la strategia mi sembra perdente.

[Ceskho]

Ci sono altri linguaggi che le cose le fanno meglio.

Tra quelli piu' diffusi, in questo periodo sto usando abbastanza Python, alla 3.3. L'impressione, dopo qualche mese, non e' quella di pulizia: ci trovo un ondeggiamento pauroso tra la voglia di standardizzare la scrittura del codice con quella di creare "scorciatoie" nella scrittura stessa, e un po' di contraddizione tra l'estensione della libreria standard e il fatto che alcune funzioni, es quelle per le immagini (che potrei definire basilari!), proprio non funzionano!

Uso 10 ore al giorno python (più alcuni suoi framework) per lavoro. Posso assicurarti che, confrontato con altri linguaggi, fa quello che deve fare un linguaggio di alto livello: agevolarti la vita! Le immagini per me non sono una funzione basilare come non lo è il suono o il video. Tuttavia ho utilizzato le sue librerie immagini anche per creare barcode particolari e funzionavano molto bene.

Gambas, ad oggi, è l'unico che permette di scrivere software con interfaccia in 12 secondi! Dovrebbero premere su questo.

Qui sono d'accordo. Capisco le voglie di fare cose "strane" per i programmatori, ma a volte la strategia mi sembra perdente.

Le strategie dello strafare non sono mai perdenti se il programmatore che vuole farle non le ritiene tali. Sarebbe come dire a Bolt di correre piano perchè tanto a nessuno serve che lui vada così veloce. Basta muoversi.

[vuott]

....per me non sono una funzione basilare come non lo è il suono...

   

[pastrank]

Tuttavia ho utilizzato le sue librerie immagini anche per creare barcode particolari e funzionavano molto bene.

Beh, intanto PIL non funziona con Python 3, non mi sembra poco, e provando i bindings piu' diffusi le prestazioni sono piuttosto scadenti (personalmente mi sono trovato costretto a usare richiami diretti a ImageMagick): database e suoni ne frequento pochi, ma processando testi (analisi di percorsi utensili per macchine cnc) Gambas (dietro ovviamente a C++ e Lua) da' distacchi clamorosi a Python. Mi lascia parecchi dubbi anche l'uso ricorrente di montagne di moduli esterni per realizzare funzioni accessorie, non si puo' dire che e' politicamente scorretto, ma una volta veniva deriso chi usava troppi ocx nel Visual Basic. A volte mi pare pure che i dir sui moduli importati siano un po' troppo lunghetti
Poi ovviamente ha anche i suoi bei pregi, tipo la diffusione, la quantita' di moduli gia' fatti, tipo quelli su PyPi, la documentazione, la portabilita' tra so, insomma, non ne parlo certo male (del resto lo uso anche io), dico che non mi pare perfetto!

[Ceskho]

Tuttavia ho utilizzato le sue librerie immagini anche per creare barcode particolari e funzionavano molto bene.

Beh, intanto PIL non funziona con Python 3, non mi sembra poco, e provando i bindings piu' diffusi le prestazioni sono piuttosto scadenti (personalmente mi sono trovato costretto a usare richiami diretti a ImageMagick): database e suoni ne frequento pochi, ma processando testi (analisi di percorsi utensili per macchine cnc) Gambas (dietro ovviamente a C++ e Lua) da' distacchi clamorosi a Python. Mi lascia parecchi dubbi anche l'uso ricorrente di montagne di moduli esterni per realizzare funzioni accessorie, non si puo' dire che e' politicamente scorretto, ma una volta veniva deriso chi usava troppi ocx nel Visual Basic. A volte mi pare pure che i dir sui moduli importati siano un po' troppo lunghetti
Poi ovviamente ha anche i suoi bei pregi, tipo la diffusione, la quantita' di moduli gia' fatti, tipo quelli su PyPi, la documentazione, la portabilita' tra so, insomma, non ne parlo certo male (del resto lo uso anche io), dico che non mi pare perfetto!

Non possiamo basare la stabilità di Python parlando del ramo 3. È il ramo ancora in sviluppo perpetuo e non ne, ovviamente, esente da problemi; primo tra tutti incompatibilità con moduli di terzi. Quello che però voglio evidenziare è che se dai alla gente un linguaggio modulare, nel giro di poco tempo ti troverai a disposizione migliaia di moduli esterni e tu, sviluppatore principale dei miei stivali, potrai concentrarti solo ed esclusivamente sul core del linguaggio. Non devi fregartene più nulla di moduli per immagini, suoni, video, parser, report, connessioni e cazzate varie. L'open source ti da queste opportunità? E sfruttale cazzarola.

Io stesso se devo crearmi piccole utility per me stesso le scrivo in gambas perchè così in 10 minuti ho finito ed amen. Ma se devo scrivere del codice da distribuire entro in crisi.

[vuott]

...continua...


" Stepping in this old conversation to point out my take:

-computers nowadays are fast enough to run a VM that's fast enough to
run a normal aplication

-If Benoit says it's complicated, what's left for us

- so instead of trying to port gambas, why not making a installer that
sets everything up for the inexperinced user??? Something like the
SmallTalk approach.

Making a small footprint VM+Linux+Gambas...they're all opensource, we
can get rid of everyhing that Gambas dont need...even more, making a
ligth weight Gambas or tagging components as "portable" and others "not"
Say it's a 500MB installer? C'mon my android phone has "Hello world"
like apps that weight 30MB, half giga is 15m download...even here in the
ass of the world (small town Argentina)

As for porting Gambas, it will be really good.  I have the time (but not
the brain) to help anyone that joins the party and has the know-how for
a start.

martin p cristia "


" The Future and what kind of Gambas do we
want?  Well, this is a question general enough
that I can understand it and put in my modest contribution:
I have just started using Gambas.  A year ago I
started asking a question or two on this forum
about using it to write for an embedded system on
small aircraft.  It is the application I wish to
concentrate my limited time on, not Gambas and
certainly not Linux, both of which I wanted to
use. My questions were pretty naive and simple,
like how to set up and use multiple serial ports
for a processing system.  I was previously and
still am using a Windows 2000 lite sytem with
software I have written in VB6, about which I
know enough to get by.  I also know enough about
Windows to know it was not ideal.  My system is
not commercial (though I thought it might become
one 10 years ago).  But I was not going to pay for Windows Embedded.

In the meantime along come boards like the
Raspberry Pi and the Beaglebone Black.  Raspberry
Pi has about 2 million boards sold.  That's where
the future lies for Gambas - not with computer
experts on desktops, but with people who want to
use cheap systems largely for educational and
hobby projects.  People like myself who were
prepared to dig into VB6, but not go with
successive versions of VB.net and Embedded Windows.

Yesterday, working on Gambas 3.5.1 running on
OpenSUSE 13.1 running on VM under Windows 7, I
finally got as far as writing a small program
that actually received data over a serial
port.  I was almost there several days ago -
thanks to the Gambas examples and helpful input
from this forum.  However, I got held up , as an
ordinary (non-root) user, by not being able to
open the serial port!  A trivial matter, you
might think, but I had to dig into Linux to join
the network group that had permission to access
the serial port!   For me, Linux is simply there
as the carrier of Gambas and I do not have the
time in which to become a happy chappie forever playing with Linux.

I note there is GambOS.  I have not yet looked at
it, but I get the intent.  For home automation,
it seems the right idea.  I have little doubt
that boards like Raspberry Pi and Beaglebone, and
tablets and smart phones  will dominate the
market in future.  Right now, Gambas and Linux go
together.  But if you want a much greater takeup
of Gambas, you must make the installation and use
of an O/S with Gambas easier and more
transparent.  Look at the hardware and users it
is likely needed for.   Here's an image, burn it
on an SD card and away you go.  Right now I would
love to be able to download a current
installation image for the Raspberry Pi that had
Gambas and Wiring Pi integrated.  Similarly for
the Beaglebone Black.  So that the user can get
right down to programming his device using a
custom version of Gambas that suits the
hardware.  I should not have to purchase a small
library of books on Linux.  I would advocate more
effort towards the non-expert user who wants to
use popular and inexpensive hardware.

That's my take.
Carl "


" Sounds like a good idea to me. I'm planning on creating my own LiveCD to
allow users to run my game on Windows and OSX. Not a great solution, but
there aren't many options that the average user will stomach.

For your idea I think we'd need to set up a script or small executable
that would:

1. Install VirtualBox.
2. Extract a premade vdi (Virtual Disk Image) containing only what
    Gambas needs to run its components (probably a lot).
3. Extract a premade vbox file (VirtualBox Machine Definition) so
    VirtualBox knows about the VM and all its settings.
4. Create desktop and menu shortcuts to launch the VM.

I remember Windows had self-extracting executables, like a zip file with
an .exe extension. Perhaps there's an open source program that can
create those on Windows and then run a script afterward? If so that
would take care of everything except creating the Gambas VM. From my
experience with doing the latter, the image always ends up around 7-8
GB. I think getting the file size down to 1-2 GB is going to be tough.

Kevin Fishburne"

[vuott]

...continua...


" That seems like what we try to do with GambOS. It is primarily meant as
a learning environment, a Lubuntu 12.04 dressed down of all the
excessive applications and with Gambas pre-installed and some guides to
start learning.

Exporting it to a pre-made Virtual Disk Image gives a .ova file of about
2.6 GB, but I believe that this can be brought down a bit more.
Making an installation .iso of GambOS (for those who want to install it
themselves) is about 1.2 GB.

I have tested importing the .ova file into a Virtual Box running on
Widows XP and that worked just fine. So for those running Widows and
with little to no know how on Linux, this is an easy approach for
running Gambas.

If there would be a decent Gambas repository for Debain, I would love to
base GambOS on Debian (instead of Lubuntu), so that a Raspberry Pi
version of GambOS could be made. This would be a pre installed
Debian/Gambas on a SD card. All the ideas are there, but the lack of
some needed experts is missing in the project.

I'm still hoping for a Debian packager who can make a recent Gambas
repository (the Debain ones suck), as all it would take is a ftp
location at gambos.org, to put all the stuff and from there all can be
installed and updated. But I'm no Debian packager.

I still need to buy a Raspberry Pi, to see what is possible there and
figure out how to make the OS on SD card. But again, GambOS needs to
have a repository with Gambas for that as compiling Gambas is no option.
People need to be able to easily update Gambas, not compile every new
version, as that seems to take ages on Raspberry.

At the moment I'm about the only one working on the project. The
initiator of the project somehow got to busy, and the person making the
iso has received the latest GambOS.ova to make the iso over a month ago,
but I haven't heard of him yet.

So the project is kind of stuck and meanwhile I'm attempting to make the
iso myself (first run in making an iso ended in a no longer working
GambOS VM).

If anyone is willing to work with me, that would be great.


--
Kind regards,

Willy "


" I do.
I'm a experienced VB6 user, have a very little C knowledge and 60%
Gambas, but learning fast.

martin p cristia "


" Wally:
To return to the RPi and the like:
As I posted on the Raspberrypi.org/phpBB3/ site on 22 Nov 2013 , I
gratefully used an image file loaded into dropbox by "fuller" (see
his post on 11 Aug 2012) for my wheezy-gambas3.img file - i.e. Debian
wheezy with pre-installed Gambas3.2.1.  Any attempt to update the
Gambas to 3.5 failed.  I did update my Linux, but any attempt to
upgrade also failed. Using the .img file to load the CD card was
easy.  Then there was some messing around getting Gordon's Wiring Pi
installed and finally, using a script from "Brian's Life" and info
from Brian Hensley's site, I got the SPI working.  I spent a great
deal of time hunting around on the 'net.  As you note, I'm stuck with
those versions, but I do have a working image at that level which
should do me for a few years!.  You bet I've backed it up and I'm
grateful for the help and assist i received from those sources mentioned.
Burning a new image is so simple even I can do it!  The lack is a way
of keeping current with the developments of both Gambas3 and Linux,
not to mention necessary add-ons like Wiring Pi (with SPI) that are
necessary to properly utilise the attributes of the RPi.  The same
thing would/will apply if and when I get a BeagleBone Black, which
considerably out-performs the RPi.  I know it takes away some of the
important free aspects of Linux and gambas and the other components,
but I do pay for the hardware and I would not hesitate to pay for a
supported OS image which combined at least the Linux, Gambas and
"Wiring Pi" for these boards, so long as the price was commensurate
with the philosophy of the boards themselves i.e. low.  You only have
to see how the few dollars cost of the apps for smartphones and
tablets gets multiplied by the take up to see how just a few dollars
for software can even make a lot of money.   When you can multiply a
very small cost by a million you get quite a large sum!.  That seems
reasonable to me.  Otherwise I'm more of a parasite user than
anything else.  I'll never be able to usefully contribute to the sort
of software development which occupies most of this mailing
group.  I'm not sure if that contradicts the basic licences for
Gambas and  Linux, and all, but maybe shareware "donations" would
help the take up of all this marvellous work by dedicated enthusiasts.
Just thoughts, Carl "

[vuott]

...continua...


" Willy:
My apologies - also to Wally.  My previous on this subject should
have been addressed to you.  It was in the early hours!
I suppose the more basic answer as to getting a greater take up on
this lovely language is that you have to identify and satisfy the
target market.  I would guess by far the largest number of potential
- not current - users are those who have come from Windows and
VB6.  If they are long-standing users of Linux, then they are
probably conversant with C, etc. and don't particularly need/want a
Basic derivative.  There has to be some imperative for them to switch
to Linux. but once that decision is made, their troubles begin!  It's
all in the installation, maintenance and use of Linux and added
components that a novice gets bogged down, not in Gambas
itself.  Well, all that can be done there is to offer a packaged
installation  that's as simple as possible to support and
maintain.  After all, the user had some reason to switch to Linux, so
has to expect some learning process, just as with Windows.
Most smartphone users do not know much about Android - he/she just
immediately starts to use it and learns that way.   I can't see Linux
ever becoming that simple - it's a different ball game - but there's
no doubt that the present process of getting Gambas up and running on
a popular board like RPi and using all the latter's attributes is
just too cumbersome and time consuming.  If I hadn't been retired, I
would not have stuck to it so far.  Gambas is not the end in itself -
you have to look at the end product and make Gambas a part of the
package the user is wanting to use.  I'll get off the air now!
Carl "


" I strongly disagree here. C is not for same purposes than basic languages.
C is good for system programming and things that require fast execution
speed. Gambas is better for just about everything else.

Jussi "


" Well said!

I'm personally mostly using C for every day programming (because I like to
exploit my machine's characteristics and mix in assembly here and there).
My "every day programming" may not be everyone's "every day programming",
anyway.

But there is a time and place for every language and I tend to believe that
an experienced programmer knows this and isn't biased towards a language
just because it doesn't look as powerful in some areas as the language he is
accustomed to.

For me GUI applications are definitely Gambas' domain. This includes all the
front-end applications for command-line programs and all programs that must
control other programs (even without GUI). Also I wouldn't use anything but
Gambas for painting or database-related stuff (of the languages, I know of).
This is just what I observed, scrolling through my recently opened projects
in the IDE.

E.g., for the paper I'm currently writing (about connecting random points in
a circle and an upper bound for the length of the resulting path when using
a special algorithm), the graphics are from a Gambas program, less than 100
LOC - and they're good! I can't think of any other language where this task
is as easy as in Gambas (at least with the visual effects I wanted to have
in the graphics).

The same program gave me counterexamples for weaker bounds and tested my
hypothesis 150 million times within two hours. Of course, my yet-to-write
assembly test program will be much faster (especially if I get access to
some hardware RNG) but no chance to give me those fancy graphics!

But if it's about some "target market" (jeez!), we shouldn't forget people
like me (how self-absorbed am I today again?): Gambas was my very first
programming language. I got to know it from school, taught it myself,
though, as it wasn't on the curriculum and my brain was fresh enough back
then. What I want to say is: there are lots of schools and I know of some
teaching Gambas - and there is no reason why their number should decrease.

It's not that we should concentrate on VB6 veterans (and maybe make Gambas
more VB6). There are also seasoned Linux users who will appreciate Gambas
and there are youngsters whose first language Gambas may be. All the old VB6
users _I_ have met (shortly after their switch to Linux and Gambas,
admittedly!) were really confused about the distributions' way of package
management, Gambas components, etc. and more complained about the whole
situation. I can't remember having heard from them again...

No doubt, migrating windows users (VB6 or not) are potential Gambas users
and their number may be greater than the ones I was talking about but let's
don't concentrate on any particular group and let Gambas evolve naturally. I
don't like the notion of a "target market" anyway.

My two cents... spread over still too many paragraphs, sorry...

Regards,
Tobi "


" Tobias:
Thank you for reply.  I think maybe you misunderstood a couple of
points I was trying to make.
I was not criticizing any language, particularly not Gambas.  Nor was
inferring that Linux is not a great set of op. systems, one or many
of which would be of long-term advantage to a user.  Switching to
Linux does however pose a early barrier that must be surmounted
before many people become comfortable with it.  I have been
programming off and on for nearly 55 years, but only as a scientific
user and not a professional.  I met Unix in the 1970's, but did not
continue to use it.  To put VB6 onto Windows XP, I insert the CD,
click a few times, put in my keycode and it is done.  Heaps and heaps
of online help, both from MS and the net.  The installation process
for Gambas into Linux, plus other components gets better with time,
but is still challenging for many, including those with significant
past programming experience.

You said it yourself: "There are also seasoned Linux users who will
appreciate Gambas
and there are youngsters whose first language Gambas may be. All the old VB6
users _I_ have met (shortly after their switch to Linux and Gambas,
admittedly!) were really confused about the distributions' way of package
management, Gambas components, etc. and more complained about the whole
situation. I can't remember having heard from them again.".

That's my main point.   The original question was about how to ensure
the future of gambas.  Maybe that's already been done by it's
intrinsic quality.  However, to increase the takeup and use - and
hence continued development long into the future - you have to pay
attention to potential users and meet their requirements for
installation and maintenance.  A bit more attention to nearly two
million users of a board like the RPi would not go amiss.   The
BeagleBone Black creates another opportunity.  Why Python and not Gambas?
Carl "


" I cannot thank Benoit(et al) enough for Gambas.
I am very, very, happy with it.
I started with 1.0.4. Did a huge major corporate/mobile thing with 1.0.17/18
and working with 2 and 3 now on other items.
As an old-timer, I can write an old-style program or do super OO with it.
I cannot say enough.
Thank you
-Nando "


" I have to agree 100% with Carl!

I too have experience programming GUI's for about 15 years with VB6. I have a
very small amount of Linux experience.
The reason I am reading this list is that I did not like .Net and wanted to find
an alternative to VB6.  I found Gambas and it looked good.  I managed to get
(after a week of trying) a version of Linux running and got Gambas running on
it. All I wanted to do was create some sample user screens to show someone my
concepts. That was at least 6 months ago.  The whole Linux experience was more
than I had time for or wanted to suffer through.  The first graphical drag/drop
thing I tried to do wouldn't work with Gambas, however was not too difficult
with VB6.

I don't know how to compile a new version/release of Gambas when bug fixes come
out.  Its far to complicated for me.  If I had a few months of spare time,
maybe.  So I gave up, found my old copy of VB6 and was running in about an hour.
I don't feel I can use VB6 for anything for a customer these days, so I am still
looking for something like VB6 that I can use without needing to learn a new
operating system to use it.

From what I have seen, and some of the playing around with Gambas, it looks
very nice and the closest I have seen to VB6.
I just get very frustrated when it takes months of work to do an eight hour project.

I know a lot of people will say "its not that hard" or "I did it in only four
hours" but you are already knowledgeable in Linux.
Tom "

[vuott]

...continua...


" No one has born with skills to use Windows or Linux. But installing Windows
or Ubuntu is almost exactly the same. I didn't notice any significant
difference. Then you go to Ubuntu Software Center and write "gambas" and
click on "install". With modern system this should be done in *much* less
than two hours (depending on Internet connection and amount of updates).
That was my first experience with Linux and Gambas.

I understand if you expect everything work like in Windows, you really have
troubles. You cannot find InstallGambas.exe anywhere and Gambas is not VB6
clone even though it's similar enough to easily convert your projects. But
instead of guessing what were the troubles, please tell us, it's only way
to improve things.

Jussi "


" Hi, Tom.  I have to take a small issue with your last few sentences.  I an
   definitely NOT knowledgeable in LINUX (Ubuntu) and definitely cannot compile
   all the new changes in Gambas as they come hot off the press, but I did
   manage to install Ubuntu in under an hour and create my first 'hello world'
   program.  I did this barely two months after installing Ubuntu (x64) on my
   second development computer.
   I, too, come from an extensive background in hardware & programming having
   started in 1962 for 20 years in the US Navy (that's over 50 years now) and
   state that creating programs in any language is simply a matter of becoming
   familiar with it.  Gambas has the advantage (for me, anyway) of being very
   VB6-like.  I have now (after six months or so) taken several of my in-house
   utilities and converted them into Gambas utilities.  One of the things I
   would very much like to see is the direct importing of a MS .MDB file int
   something Gambas can use.  If that already exists, then I would welcome
   someone pointing me in that direction.  I have an extensive (some 3800
   records) database containing my DVD/Vinyl/VHS/Cassette music & movie library
   and really don't want to recreate it in Gambas.
   I don't profess to know what happens (much) under the hood of LINUX, but as
   I go along, I am learning.  I keep a pen/pencil notebook near my development
   computer that I use to keep track of things I learn about the OS and how to
   manipulate it.  At the end of the week, I transcribe them into a searchable
   PDF file I created to hold all these goodies.  You might try something like
   that.
   In time, I suspect that you will get the hang of Gambas and move more and
   more of your work over onto that platform.  Yes, it does take time.  My
   'hello world' program took three days to complete.  As I gain confidence
   (and skill) sections of code go much faster.  As for migrating VB6 into
   Gambas I simply take a subroutine/function and drop it into Gambas and work
   on the errors it produces.  Most of them are simple to fix.  One big help is
   opening the .FRM file in a text editor and using the values to re-create the
   form in Gambas is relatively easy.  Sometimes, you can do it in a matter of
   an hour or so depending on how complicated the form actually is.
   As I gain experience in Gambas, I keep finding new ways to do things that
   are actually shorter and easier to do in Gambas than VB6.  Who knew?
   Bill "

[vuott]

...continua...


" Jussi said:  "But instead of guessing what were the troubles, please
tell us, it's only way
to improve things."
That's a fair question.
Some time ago I had actually bought a set of DVD's with openSUSE12.2,
after unsuccessfully trying to intall Lubuntu and Ubuntu on an older
computer.  (I'll try again some time).  However, openSUSE 12.2
installed no problems.  Then I looked for a repository for gambas3
and once again, munix9:gambas/openSUSE_12.2 came to the
fore.  However, gambas3 would not install its examples because of
missing dependencies like imlib2, libogg and v4l-tools: I seached the
net, obtaining many pieces of advice on installing these. (Yes, the
process is now consuming heaps of time and energy). I started to
become familiar with "zypper". I tried
http://sourceforge.net/projects/enlightenment/.  I'll say I neded
enlightenment!  By the time I got to "sed -i
"DGifOpen/s:fd:&,NULL:"src/modules/loaders/loader_gif.c  &.configure
--prefix=/usr --disable-static && make" and similar instructions the
cat (the one with four legs, not the Linux command) had the good
sense to leave the house!  This went on for some days - days, not
hours!  Looking now at my notes (and ignoring the expletives) I found
that imlib2 needed libmlib2.so.1 etc etc.  Finally, I hit the
"search" facility at
software.opensuse.org/search?.g=lib&baseproject=openSUSE% and
installed the missing components.  Yes, it's a learning experience in
Linux as it actually is, but no, it's not what I need when I really
want to get on with using Gambas.
If you try to install Gambas 3.5 on the Raspberry Pi running it's
version of Debian Wheezy, the CPU goes into overdrive and stays
there.  That's what happens in the real world of newbies to Linux and
Gambas.  No wonder Tobias "never hears from them again"!
Carl "


" > However, gambas3 would not install its examples because of
> missing dependencies like imlib2, libogg and v4l-tools: I seached the
> net, obtaining many pieces of advice on installing these.
>
All dependencies should be installed automatically... there were something
wrong in how Gambas was packaged.
I guess this is under work, if not fixed already.


> I tried
> http://sourceforge.net/projects/enlightenment/.  I'll say I neded enlightenment!
>
By the time I got to "sed -i


> "DGifOpen/s:fd:&,NULL:"src/modules/loaders/loader_gif.c  &.configure
> --prefix=/usr --disable-static && make" and similar instructions the
> cat (the one with four legs, not the Linux command) had the good
> sense to leave the house!
>
Nothing to do with Gambas and so completely unnecessary. Those commands
seems to be for compiling "enlightenment",
not just for installing it. Not newbie stuff at all.


> If you try to install Gambas 3.5 on the Raspberry Pi running it's
> version of Debian Wheezy, the CPU goes into overdrive and stays
> there.
>
Without details hard to say what happened. Compiling any C program is quite
CPU intensive and RPi isn't very fast.
So maybe it only took very *very* long.


> That's what happens in the real world of newbies to Linux and Gambas.
>
Sounds like I can't recommend openSUSE to newbies.

Jussi "

[vuott]

...continua...


" Jussi:
Re Ubuntu:  I don't know what went wrong - maybe it was my hardware -
I'll try again sometime.
Re the munix9 repository: Yes, I think the packaging was at fault.  I
saw a reference to that somewhere.
There was a useful guide to installation on
http://gambasdoc.org/help/install?en&view&v3, but it's a bit out of
date now.  I did use it to install an earlier Gambas (3.2.1, I think)
on SUSE10 - on the above hardware.
Re the RPi and Wheezy: Not quite as you think: Gambas 3.5 appears to
install OK and finishes.  It is when you try to use Gambas 3.5 that
the CPU usage goes to 100% and stays there.  Other potential users
found the same thing.  The only thing that got me a nearly up-to-date
Gambas on the RPi was the image with Gambas 3.2.1 installed that I
got from a private download.  See previous post.
Re openSUSE and Gambas.  To dissuade potential users from using
openSUSE would be wrong, IMO.  Gambas 3.5.1 installs without any
problems on openSUSE 13.1 and Gambas works very well - once I had the
permissions under control! I find the latter OS with KDE a very nice distro.

To be constructive, here is my take:
1.      There needs to be an up to date list of distros and
repositories, together with some installation instructions, that work
with recent versions of Gambas - what distros and repositories that
gambas will both install and work under.  Just like the list referred
to above at gambasdoc.org, but kept up to date.  (Maybe there is one,
but I have missed it?)  The usual process is to first choose a Linux,
then try to install Gambas.   One needs to be able to choose a
combination that installs sweetly.  Like OpenSUSE 13.1 and gambas 3.5.1.

2.      When you have boards that come out with potentially a huge
take-up, like the RPi, someone or some group (like Willy Raets has
proposed) needs to be able to get out an img file (or something
similar) that has a current Linux distro (or several alternatives)
with Gambas and other components (drivers) pre-installed.   There's
an opening right now with the BeagleBone Black board.  By all means
charge a modest price for the file, to cover the cost of producing
it.  These boards attract newbies, because they open up fresh
computing opportunities.

I have used up more than enough space.
Carl "


" Yes Tobias, you put it in the last paragraph:

> No doubt, migrating windows users (VB6 or not) are potential Gambas users
> and their number may be greater than the ones I was talking about but let's
> don't concentrate on any particular group and let Gambas evolve naturally. I
> don't like the notion of a "target market" anyway.
>

Target markets are things we were fed up with when we came to Linux. Let
us avoid making the same mistakes as the big players on consumer markets.


If that's your only point, Carl

> continue to use it.  To put VB6 onto Windows XP, I insert the CD,
> click a few times, put in my keycode and it is done.  Heaps and heaps
> of online help, both from MS and the net.  The installation process
> for Gambas into Linux, plus other components gets better with time,
> but is still challenging for many, including those with significant
> past programming experience.
>

I 100 % agree.

There are some more really good projects around in the Linux world, and
usually they take care to have their main releases put into the major
distros regularly - and in a way that lets them integrate flawlessly.

We cannot compare Windows with Linux here, however, as Windows is a
(relatively) homogenous environment whilst Linux PCs can differ in many
more ways. Package management is way more complex here.

Rolf "


" I'm one of those experienced VB6 users that used to swim in the Caribe,
and Microsoft sent to the Marianas abiss with .NET ; so I think there is
a huge amount of potential ex VB6 programmers to join the Gambas force.

So Gambas is not THAT easy to install or compile? Well, we're
programmers, lets show some effort...

Linux hard to install? Ubuntu sets a computer up in 20m, and if its a
decent one, the only thing you need to know is your time zone. Otherwise
it is all answered in Askubuntu.

Our goal has to be making Gambas compiled program END users, use
whatever O.S. they want, using a one-file-downloadable-installer.

martin p cristia "