When I am telling people how much time I invested in the development of Padre in the past year they usually think I am nuts. Why instead of working at clients making money I am spending my time writing Padre?
On the other hand if I told people in the Perl community it will immediately eliminate the little karma I have as they would find out I am not doing it just an hour in the evening but sometimes for whole days. So they would find out - if they did not know earlier - I am not such a good programmer.
So in order to convince these people I am not totally nuts I have to give explanations that go beyond the I am having fun as that alone does not bring food to the table.
Earlier I already wrote Why am I writing Padre? but actually with that post I mainly wanted to draw attention to Why others write Padre and only as an extension I wrote about my reasons. Those reasons could have been mistaken - to the untrained eyes - to be altruistic. They are not.
This time let me try to write down some of the business reasons I am writing Padre starting by the most obvious ones.
Consulting and training
If Padre can help make Perl 5 more accepted in the enterprise and if more people use Perl it will mean more companies need training in Perl 5 and more companies need consulting services. Some of this hopefully will reach me as well.
At least here in Israel there are 3 or 4 training companies that provide among all the other technology stuff also Perl training. They usually employ people who either don't know Perl, or don't care about the community aspect or tell they don't give a shit about this Open Source thing or simply teach PERL. Even though I am relatively well known in the local Perl community and people do recommend me, when it comes for companies to pick a trainer for a Perl course many times the training managers will prefer the well known names in the training industry than someone like me who does not have a marketing budget or a sales agent.
I am told writing a book about Perl could help in this situation but I think it is beyond my capabilities and my patience. Hopefully they will be more ready to take their risk and have the man who wrote the IDE as their Perl trainer, even if my role is slowly diminishing in the pool of Padre developers. (I just checked today. Only about 25% of the total commits on Padre are mine and to my joy this is dropping rapidly.)
I can imagine that once, when Padre becomes used by many people in many organizations they will want to make local customizations. I think that we should connect Padre to both various open source version control tools, bug trackers and other systems related to development. That will allow many people to use these connections out of the box but there are lots of home made customizations in the world of configuration management. So I hope Padre will reach a point when companies want to integrate it with their home made solutions. Who better could they hire for such task than one of the Padre developers?
I am still studying what applications are developed on the Eclipse platform and if we can provide an alternative platform but with Perl and Padre as the framework? If we can do that, it will open the doors for a lot more development projects. Both for everyone in the Perl community and for me specifically.
After all if someone has Padre installed she already has wxWidgets installed so building and distributing a desktop application is a lot closer than before. With all the emerging plug-ins of Padre a lot of the specialized modules on CPAN can be exposed to simple use via a GUI.
Test automation is a field that I think Perl can have a lot more use than it has now. As test code is not distributed outside of the company this field - similar to web applications - does not require compilation-for-code-hiding. Something Perl 5 lacks. With that disadvantage out of the way and with the enormous set of tools on CPAN we could provide a great platform for software testing.
I think one of the biggest selling points of tools such as the HP Mercury WinRunner, the IBM Rational Robot or the TestShell of Qualisystems is the record/replay part or the Code-free test creation part even if in the end the majority of work will be done after the initial code-free session using Visual Basic or similar language.
We already have Perl as the backend language. We only need the GUI part to make it more user friendly and more manager friendly.
After the German Perl Workshop I started to build a Padre plugin that would become a GUI tool to record a replay web access using Sniffer::HTTP. It could already do the recording but then I ran out of time and it is still only half written in our SVN repository.
So I think if we can enhance Padre with a test generation and running system and package it nicely (see the first steps in Padre stand-alone installer for Windows - first beta version) we - as in the Perl community - will see lots of renewed interest in using Perl.
Applications for Windows system administrators
Look at this code:
use Wx::Perl::Dialog::SingleChoice; my $choice = Wx::Perl::Dialog::SingleChoice::dialog( title => 'Select one', values => ['this', 'that', 'else'] ); message(text => "The item you selected is $choice\n");
This will show a window with 3 options. The user can select one option and the selection will be displayed in a new window.
This is a simple straight forward script - no callback or any other GUI magic - that shows how easy it is to write small utilities for Windows that hide the ugly Windows command window. This is now part of the Padre distribution but we can move it out to be a stand alone package (at one point it was) and let anyone use it. If we can wrap the various Windows system utilities in a similar GUI we can start making further inroads in the Windows small utility programming world which can also lead to larger applications.
Who needs GUI anyway? - Irrelevant to you?
I am sure many people will feel these are totally irrelevant to him. Sure most of the core Perl developers are using Linux or Mac and are comfortable with the command line but it is not true for the majority of the programmers and the majority of the people using computers. Especially in the Windows world.
Part of the above solutions are aimed at those but I believe the command line people will also see the fruits of such efforts in increased number of projects and probably higher status as they are the ones who know the guts of the system.Published on 2009-07-16 by Gabor Szabo blog comments powered by Disqus