Re^2: Are Perl programmers just easier to deal with?

[CloneArmyCommander]

Bingo :)! This is mostly what I was looking for when I wrote my Medition. I guess it was just a bad choice in getting started, it was just that was what was in my mind at the time, and just wanted to get everyone in the moment that I was in. Apologies for the bad lead in.

Mostly, what I was noticing is trying to get help from someone who works with a compiled language, while possible, just seems much more difficult. You have to either wait until the next version comes out instead of asking for a quick fix patch that can just be kluged in and have a nice working product.

Trying to get help from users of interpreted langues, seems to be quite the opposite. I can ask a question here, and I have to finally decide to say "that's enough" becuase I can get so many responses that I nearly drown in them. I figure that what it all boils down to is, seems like a better tech support option, we seem to have enough time to support our projects without having people wait for the next version (unless we choose to do business that way :), and in addition help others get their projects done.

Now, please do not misunderstand me, I am not going down the road of "which languag is better?" because I see it as a dead end. Although, it is nice to stir things up around the monastary, and that seems the best way to do it, hahahaha ;).

[samizdat]

Well, to my mind, there's another distinction. It's long been a joke that (to use Sym@ntec & M$ as an example I remember well) "It's a Microsoft bug" and "It's an integration problem" degenerate into "We'll fix it in 5.0" and {ignore him}.

Contrast that with the open source world, where you get five good user-generated suggestions in five minutes, developer queries overnght, and patches the next morning. I'm not saying it's always this way -- see node embperl and Apache::Session::Memcached which has also been posted on the memcached and Embperl lists without any real suggestions -- but I have almost universally been impressed by the interactive and supportive nature of open source developers.

Part of the problem is that you're never dealing directly with coders or even architects in a commercial company. At worst, you're dealing with Indian "English", at best, you get a script kiddie in a phone farm. I'm all for business and capitalism -- I run a profitable biz myself as well as working -- but corporate bean-smashers ruin the game when a biz "grows up". Sym@ntec is a perfect example of how good products have been placed in a rotten corporate context. Peter Norton created a very useful tool suite, and Act! really is (was!) a well-done contact manager, but Symn@ntec's corporate choices would make a dog howl.

[Anonymous Monk]

Contrast that with the open source world, where you get five good user-generated suggestions in five minutes, developer queries overnght, and patches the next morning.

<rant>
That's so untrue. Yes, there are a few projects in the open source world where that happens, but that's true for closed source projects as well. But people generalise the positive exceptions from the open source world, and the negative exceptions from the closed source world.

Look at Perl. Look at the list of open bugs. Some bugs are many years old. That's a far cry from "patches the next morning". Perl currently is on a 3 maintainance releases/year scedule, but that has been different as well. There were 14 months between 5.8.0 and 5.8.1, with no maint releases in between. Now I understand why this happened, I'm just pointing out that the open source world isn't all that shiny and fast as you suggest. And then, Perl is one of the better projects. How many open source projects are abandoned? How many authors unreachable? From how many is the status totally unknown? Care to do some statistics on sourceforge? Or CPAN for that matter?

Let's face it. No one has an interest in solving bugs that don't affect them. New features give a much bigger rewards, either because it sells, or because it gives more glory. You remember the person creating a popular CPAN module, like DBI. But do you know who solved the most bugs? You might put your list of CPAN modules on your resume. But do you list the patches you submitted to various open source projects? In fact, commercial vendors may have more of an initiative to fix bugs - not fixing them could mean losing customers (of course, if you have 10 million customers each paying a small amount losing 10 of them is less of an impact than if you have 300 customers each paying a million).

In my experience, all software sucks, and most vendors/authors are hard to move to fix their bugs. And it doesn't matter whether they are closed source or open source developers/vendors.
</rant>

[samizdat]

Perhaps I am, as you say, putting "our side" in the best light. However, it is true that the OS world has more doorways, and, even if they aren't always open with the light on, at least one can find other ways to get the code and other places to ask. *THAT'S* what's different in the commercial software world.

[samizdat]

For you downvoting Anonymous Cowards who choose to say that 'all software sucks', I just saw another example of the OS process working. Over the weekend, somebody wrote to the Embperl list that Embperl wouldn't compile on RedHat Enterprise Linux. Gerald Richter, the developer of Embperl, sent the guy a couple iterations of Makefile.pl, first to diagnose the problem and then to fix it. DONE!

[CloneArmyCommander]

It is interesting to get the perspective of someone who is a business owner. I am still in college, so I am working from observation from the outside, which is what compelled me to bring my thoughts to Perl Monks.

It gets difficult to find a middle ground. My personal computer, is big bundle of contradiction. I have Linux dual booted to Windows, so I have software on both ends of the spectrum. I do not want to dive into the quality issue, I honestly see both sides as having good quality software. I have had better luck working with open sourcers, which seems to account for most of my support and loyalty to them. I do not mind working with a corporation, and they have their moments that they listen to their customers, after all they are in the business in suply and demand, but sometimes I wonder if it is because the open source world is starting to give them a run for their money.

Oh, by the way, thanks, this is the first intelligent conversation I have had in a long time, hahaha :).

[samizdat]

I'm sort of in the same boat. The profitable parts of my business are all Apache on FreeBSD, and most of the work I do at Sandia is FreeBSD (well, they eant it to run on Solaris, but there are enough BSD boxen around that I can NFS and drive from FreeBSD. Sandia, however, runs its corp nets on Doze XP and some of the users expect data to show up in Doze formats like Excel. (Thanks, Perl+CPAN!) Many of my corporate users for my websites also run Doze -- one site is 96.5% MSIE clients -- and they constantly send me doc files and powerpoints.

I have had great experiences with Doze-based software, too. Corel Draw has no replacement, AFAIK, and I use a lot of PC board and chip software that is only available on PCs unless you want to drop $50K for entry-level stuff. However, I cannot think of *any* doze software company that I have had satisfying contact with since Peter Norton sold out. Part of my business used to be electronics design consulting, and I'd get raped regularly by OrCAD and Keil and other embedded tool vendors.

That said, however, I believe that there is a really strong groundswell, especially in EDA, for open source development. People are tired of being raped for "maintenance" charges that don't buy anything, and OS tools like Eclipse (EDA framework), OpenOffice (Office Productivity) and Blender (3D visualization) plus Perl, Apache, MySQL, and all the other old favorites like FreeBSD and Linux are providing real competition.

Yes, software companies are being forced to listen. They will die if they don't, bluntly. In EDA, a number of people are trying to release proprietary extensions to Eclipse in order to lock in customers like they used to, but the customers are getting wise to it. EDA is such a bleeding-edge business that many people pay the tab because you just have to get the job done, but almost everyone I know is spending their extra brain cycles boning up on the extensibility features of Eclipse, gEDA, and SystemC.