I read the very interesting article about the Systems fair:

http://www.freepascal.org/wiki/index.php/Systems_2005

One conclusion was that the web appearance needs to be improved: "..
The websites are a serious problem. They are quite ugly and not very intuitive, and this is the case for years. The members present had quite some difficulties to explain how to get the latest full installer for Lazarus, for example. Not everybody knows Freshmeat, and the main address www.freepascal.org won't help those people really much to get to the right download location. sg will handle this issue as soon as possible. ..."

Are there any plans so far?

I've been extremely busy since Systems, but things begin to get better now. As far as I am concerned, I will continue this topic this or next week.

- Sebastian

I think there should be more independent websites promoting freepascal in the mean time. When I volunteered and asked why there wasn't a "contributed programs" section in addition to the "contributed units" section, and why there wasn't more contributed units categories, they simply stated the answer to my problems was creating more pages on the FPC wiki.

I've given up suggesting ideas for the FPC website since then, and now work on my own FPC based websites and put effort into them, instead of suggesting new ideas for FPC website. I think other people are encouraged to build FPC related websites if they can't help the freepascal site directly.

I think like how torry.net, Delphi 3000, etc. work (but you obviously can create smaller websites than that) people should be building independent websites to promote freepascal.

I would diagree. Up to now there a tons of sites concering FPC/Lazarus: freepascal.org, sourceforge, stack.nl, wiki, ftp-sites, mirrors, cvs, etc. That produces much confusion. There is a good rule by marketing people: "One face to the customer".
Moreover people are used to see Free Pascal and Lazarus as *one* project. But it isn't.

some discussion points:

- one canonical site
- consistent style/color on all sites
- strip off old bones (TP tutorials, MOS)
- merge FPC and Lazarus sites

this is about the same we've discussed during Systems... :)

Giving ideas doesn't help. We need people to do the work. We can't extended/debug fpc itself and rewrite a website at the same time.

>I think like how torry.net, Delphi 3000, etc. work (but you obviously can create >smaller websites than that) people should be building independent websites to

promote

Mirrors and cvs are not at all like what I was talking about.

That produces much confusion.

Of course, mirrors and cvs do cause confusion. And all the clone documentation sites out there do too. That's not what I was talking about. In order to see PHP in use, people build sites with PHP and put a PHP logo on it. PHP isn't successful because of the PHP.net website!

"One face to the customer".

I'm talking about genuine, unique, freepascal user websites. Look how many PHP or Visual Basic sites there are, with "I'm a happy PHP user" written all over them. The extension myfile.php is marketing itself. If your websites had a myfile.pp or mysite.fp extension, that would be marketing via brand naming. The PHP extension on files itself is branding. Every time you see a PHP extension on a file, you are getting brainwashed with the word "php". That's marketing.

Okay, I come from an internet marketing background.

If you have one single domain name, your search engine rankings and traffic are pretty poor. If you have 5000 websites promoting freepascal, your search engine rankings and traffic improve greatly. It's similar to having 5000 ads in the paper versus one nice ad. And mirrors bring your search engine ranking down. If you have 100 mirrors, your search engine rankings and traffic do not improve greatly. Mirrors can cause google to see your website as clones, and this sometimes brings the ranking down. That's not really a big issue right now with the FPC website, since the mirrors don't seem to be affecting it's ranking.

I don't think the GNU C compiler website looks all that good? In fact I don't even know if there is a GNU C compiler main website or homepage!? I probably wouldn't ever visit it, since there are so many other ways to download the GNU C compiler. I'm more interested in the people who USE and have had real world experience with the compiler, right?

I don't think the GNU C compiler is popular because of one nice website. I think the reason GNU C compiler is successful is because of all the fan boys, their applications (gzip, midnight commander, linux, and 50,000 other applications made in C), and their websites, and of course directories like debian.org which link to several C applications.

So in summary: one nice website is nice, but from a realistic internet marketing perspective, more importantly are huge databases of content, examples, and program/unit directories from the users. FPC documentation already exists well on the search engines, but examples of the applications or websites that are run off FPC engines or FPC code, do not exist.

Don't get me wrong: one nice website, is not harmful in any way. It's not going to harm anyone. I just seek a realistic marketing perspective, where thousands of people are downloading software from word of mouth and real world examples, rather than a nice looking website.

L505 wrote

I don't think the GNU C compiler is popular because of one nice website. I think the reason GNU C compiler is successful is because of all the fan boys

I'm quite sure it's mainly successful because it's the default (and often only) C compiler on Linux systems, as well as ported to pretty much every other system out there.

Jonas

Op Thu, 24 Nov 2005, schreef L505:

So Pascal Server Pages should have a .pascal extension or something to the executables... Maybe :)

There is not much wrong with our search engine rankings. We have a lot of external and internal links to our site, it does its job.

http://gcc.gnu.org

I think our website is better by leaps and bounds than GCC's website. GCC's website is definately not the reason for its success.

Yes. Every .c extension is marketing for the C language and the compiler to compile it is preinstalled on most systems :)

We need to advertise better. The 2.0 marketing was done quite well IMHO, OSNews was a good choice, it gets 450000 hits a day, and with a story on-line it was easy to get sites like Slashdot interrested.

However, we cannot live from one big campaign once in five years. The problem is we are not as famous in the generic open source world as we are in the Delphi world for example. We need to get more known in the open source world and the majority of work that needs to be done here is outside freepascal.org.

Daniël

Ok, even if outside websites about free pascal help more then working on free pascal website directly, why don't you make it possible for people to see the website code and submit patches for it?

This way the developers can focus on coding the compiler and others that wish to improve the website can do so.

I'd specially like this on Lazarus website.

Felipe Monteiro de Carvalho wrote

Ok, even if outside websites about free pascal help more then working on free pascal website directly, why don't you make it possible for people to see the website code and submit patches for it?

The code is in svn, in svn.freepascal.org/FPC/svn/html

Jonas

Op Sat, 3 Dec 2005, schreef Felipe Monteiro de Carvalho:

We did add the Wiki for this purpose, and are looking into ways to increase user contributions to it, but it should be a controlled system, i.e. we don't want people modifying the home page etc.

Submitting patches is and has always been possible, the whole website is in svn, http://svn.freepascal.org/svn/html .

If you need more permissions on the community site, i.e. change stuff, write content, translation, moderator please say so.

This way the developers can focus on coding the compiler and others that wish to improve the website can do so.

Realising this means we means to concentrate more on the web site, and a lot more, only after the functionality is there you get the benefits.

I'd specially like this on Lazarus website.

That is another issue that Lazarus people are also looking at.

Greetings,

Daniël

I agree..

Actually, then, I should more so compare FPC to something like python or perl rather than GCC. Because GCC was available years ago when FPC was not. Whereas perl an python are newer. I think *part* of the reason perl is popular is because of all the websites out there saying "perl is so great because I did this and that with it". Of course there is a problem comparing a compiler with a scripting language, since scripting languages are used more for sysadmin.

But what I'm getting at with my "fanboy" point, is that a lot of the FPC users are very quiet people and do not have as big of a mouth as they should. A lot of perl and python programmers do have big mouths and speak up about their language. But still, Pascal has it good - there are way more Delphi/freepascal projects than there are Smalltalk ones, for example.

Regards.

There are way more Pascal projects then Python or Perl projects. The only thing being that the majority of this projects are created with Delphi and not Free Pascal. Things like Skype and Age of Wonders computer game were created with Delphi.

The comparison to Perl and Python is valid only in the Linux community. On Windows I don´t see these two languages, ever. The competition is with .NET and Windows is still a much greater market.

What I think is happening now is Borland has being doing wrong steps for quite some time (like abandoning their own project of a framework in favor of .NET). Of course there are many factors on that, like Microsoft constantly buying their best programmers and buying pieces of Borland Capital. Delphi is slowly losing space, and this could be occupied very well by Lazarus, and unless they hear about Lazarus first, I bet those developers are going to Java if they wish to build multi-platform apps.

On the Open Source world things are different. It is more of a meritocracy. On the technical part there is also a large room for FPC and Lazarus, specially for building GUI applications, but I think people won´t pay much attention to it unless there is a very large open source project using FPC. Something as big as Firefox, Open Office or KDE, because this projects draw lot´s of attention to them. KDE for example is a major propelent of Qt. It even includes a Qt IDE in it´s packages!

Felipe Monteiro de Carvalho wrote

There are way more Pascal projects then Python or Perl projects. The only thing being that the majority of this projects are created with Delphi and not Free Pascal. Things like Skype and Age of Wonders computer game were created with Delphi.

At least the Mac OS X version of Skype seems to be a combination of C+
+ and Objective-C, with no FPC-generated code in sight.

Jonas

Jonas Maebe wrote

At least the Mac OS X version of Skype seems to be a combination of C+ + and Objective-C, with no FPC-generated code in sight.

It probably is. But the initial version, the Windows one is a Delphi app. The Linux and Mac OS versions are not ports of the Windows version. They rewrote everything on each platform. This information is easy to confirm by googling.

http://www.codecomments.com/Kylix/message684809.html

Of course it would be much easier for skype developers if they used FPC for the Linux and Mac OS version, .... but this did not happen.