Web aps with Perl (6) vs. Longhorn and the Windows API?

[spq]

Recent mental meanderings:

Lately I have been despairing that Perl 6 will ever happen. Or at least not happen in time to be useful. I don't really follow perl6-language very closely anymore. I was very excited for the first couple of years, but how long can a person be expected to remain enthusiastic? I still think Perl has huge potential. But will Perl 6 take so long that other languages will have eclipsed the value of the changes and new features? I'm just a code grunt without enough time to keep up with the latest in language theory or industry trend.

A lot of my worry is fueled by the dearth of programming jobs I've found with Perl as a significant component. Particularly for web systems where PHP seems to be rapidly outgrowing Perl. I love working in Perl, and am reluctant to look for real work in another language. How long will be too long though? Will we need a new name by the time Perl 6 is ready (and I am still assuming it will be completed some day) so that it isn't getting written off as some last gasp of a language that has already become passe?

I found a link to an article today that both gave me hope and increased my worry. http://www.joelonsoftware.com/articles/APIWar.html I've observed a bit of this myself at work. If a stable Perl 6 gets out the door in time and is as good as it's sounds like it will be, how to we we increase the ... buzz? :) How do we, as a community, take advantage of that (and more) to drive Perl as a premier choice for developing web/distributed applications? How do we make it so that when HR/managers think LAMP, they assume Perl instead of PHP?

A stable version of mod_perl for Apache 2 seems incredibly important. Efforts like Class::DBI also seem to provide incredible potential; yet so few people out there even know it exists, including developers I know who work with Perl daily in their jobs! I think Perl is already superior to PHP, at least in most ways; where is PHP getting more traction than Perl from?

Sorry this is such a ramble. It's just my thoughts and concerns, for whatever they may be worth. I love Perl and am lazy enough not to want to switch to something new ... at least until I find something even better. So I find myself wondering what can we do to get other people to understand what most of use believe? Where can we make Perl even better for web apps, a space we still have a fair foothold in?

What are other open source (and not?) projects doing to get 'media' time that we aren't? ... Can anyone suggest a good primer on writing tech articles? Many, if not most of us have worked on cool projects, or used Perl in key systems at work. Maybe a good step is to start talking to people outside the choir? Can anyone suggest magazines/websites to submit publications to where Perl articles might broaden our exposure?

[Ovid]

FYI: Patrick Michaud is actually doing work with the grammar engine for Perl 6, tests are being written and work on the actual language is finally happening. Many share your frustration, but I honestly expect to see significant progress on an alpha Perl 6 in the coming year. It's finally becoming more than vaporware.

[spq]

Thanks! It's been a long rough season and I'm a bit out of touch. I'm really glad to hear real work is moving forward (not to say all the planning and brainstorming weren't work)!

[Juerd]

where is PHP getting more traction than Perl from?

mod_php is more useful for bulk hosting than mod_perl because mod_php is less powerful. As a system admin, you don't want a user you don't personally know to write a handler, or to be able to do internal Apache things (like play with subrequests and internal redirects). And even though having an eternal interpreter and an API to communicate directly with Apache make mod_perl so fast, not having an eternal interpreter is what makes PHP much easier to work with and much more reliable in case of sloppy programming, and not having that API would comfort sysadmins more. Perl interpreters are very heavy, and that would need to change before a mod_php-ish thing can be made with Perl. The closest thing you can get now is letting the process that handles the request die. You then skip startup overhead by CoW forking, but don't have the risks involved with having your interpreter live on.

Summarized: mod_perl is too powerful to compete with mod_php on the bulk hosting / idiot programmer market, and unfortunately, that's where popularity lies.

[theorbtwo]

I should think a good middle-point is having each individual script having it's own interpreter -- what's wrong with that? (You could have them last forever, and rely on the kernel to swap them out as it wishes, or you could tear them down if they aren't used for a fixed period.)

[Juerd]

I should think a good middle-point is having each individual script having it's own interpreter -- what's wrong with that?

It's a good solution, and several programs use this technique. However, for things to work correctly, still you need to pay attention and not program sloppily. Another problem is that this puts very heavy load on a machine if you use one script per page (as is very common in PHP), regardless of what the kernel can do.

[brian_d_foy]

The Perl jobs are out there: you just have to be willing to take them. Like any other job, you have to go where the job is. I know one Perl shop in Chicago is constantly looking for new hires, and Google, Amazon, and Yahoo are looking for Perl programmers. The Perl Jobs list has a steady stream of posts.

Perl gets a lot of media time. It's not as much as Java, but it's a lot more than most other languages. Most people in the industry, including HR types, know that Perl (or PERL or Pearl or something that rhymes with hurl) exists. Ask them about Ruby or Python and see if you get a reaction. It's not that Perl is losing stature, but it's virtually matured into the venerable position of C, and no one seems to be worried that C is going to disappear. I see Perl in a lot of places, but because it was useful for a particular part of a problem, not necessarily because the people wanted to give Perl center stage, and I think that's good.

If you want some resources, check out the Perl advocacy mailing list, O'Reilly's Perl Success Stories, The Perl Journal, The Perl Review, The Perl Foundation, Yet Another Perl Conferences, Perl.com, CPAN, and it's activity level, Perl Mongers, and The Perl Home Page. No other open source group has it pulled together like the Perl community. Despite recurring myths of its demise, Perl is rather healthy and relevant.

On the other side, you also need to see Dave Cross's "Why Perl Advocacy is a Bad Idea", Nat Torkington's "Be an Advocate not an Asshole", and Mark Jason Dominus's "Why I Hate Advocacy".

[spq]

Hmmmm... I was subscribed to The Perl Journal for years (and have all the paper copies AFAIK), donated to the Perl Foundation, been to a couple (and presented at one) YAPC, have been hosting the Boston.pm tech meetings for a while now, etc. etc.

I'm not so much worried about Perl's demise, at least not in the immediate future, as I am interested in seeing Perl be the first thing businesses consider when deciding which technology to use in developing their web application. This is as much selfish silliness (I don't want to move my family to Chicago) or ego as it is basic advocacy (I do like Perl best).

I agree that we (I) have to be careful about not being over-zealous in advocating Perl. I try to reserve that for discussions on places like Perl Monks. =P I'm asking, is Perl 6 going to happen sometime soon, and if so can we plan on generating media outside our usual haunts when it does? If web applications are a 'next thing' and Perl already has a large role in that space, is this a good time to look ahead at increasing Perl's share of that space, hopefully to everyone's benefit? What might be good ways of doing that we can try or suggest to the wider community? I'd like to publish some articles (which in my case will require a really good editor! ;), but should I aim for the Perl Review or someplace more widely read?

Thanks very much for feedback and the links, a few were new to me.

[chromatic]

... what can we do to get other people to understand what most of use believe?

Donate code, documentation, tests, smoketest reports, or money to the Perl 6, Parrot, and Ponie projects and convince other people to do the same.

[Juerd]

If you read chromatic's node and want to donate money right now, here's the link: http://perlfoundation.org/. Even donations of $ 1 are possible. If you don't want to spend much, remember that $ 1 is still Inf times $ 0. If your employer uses Perl, ask them to transfer a small portion of the profit Perl has made them. It can be seen as charity or an investment. Note that on the Fund Drive Details page, sponsors are listed for 12 months. There were some logos of big sponsors until a few days ago; they expired. This is your chance, if you (as a company) have the money.

[Jenda]

I don't think that's the answer. The question (as I understand it) is more how to increase the "brand awareness" and improve the image of Perl, not how to help developing Perl 6. Though of course it's also important. I don't have a good answer though.

All I can and will do is force an update to the "marketing web" for the project I work on to give Perl the credit it deserves. (In case the recruiters wondered what is behind the system that delivers their ads to job boards.) And do a presentation for fellow developers in other teams on the Perl based tools we have and what could the Monster build process have been if only someone paid attention. A bit too late since they go the ASP.Net direction even regarding the messagizing/i18n, but a little envy can never hurt.

[spq]

You understood my question quite well Jenda, thank you. I think the main thrust of my thought (however poorly presented), was that if the idea put forth in the linked article is correct; that the Windows API is diminishing in relevance and web applications are on the rise. And if Perl is one of the best choices for developing fast, extensible web applications, if not the best, then how can we increase Perl's percentage in that growing market?

This is largely selfish. I like programming in Perl, it's my language of choice. I'd like to have my skills and experience become more and more in demand. I want to pick where I will live, not drag my family across the country looking for a job in one a large Perl shop.

I have, and will continue to donate code, money, etc. to the best my resources allow. But that chromatic, in and of itself, not going to generate more Perl jobs in general. I think Jenda has the right idea. Getting your company to donate money is great. But get them to let you publicize that your work was done in Perl, wherever reasonable. Whenever I do work in mod_perl, I put one of those little 'powered by' icons & links. I don't think it has to be huge things. I don't think we need to go around trumpeting the glories of Perl and annoying everyone. Just keep our eyes open for ways to let people know what Perl not only can do, but has. Someone mentioned O'Reilly's Perl Success Stories. I'm very grateful for all O'Reilly has done, but if your not going to perl.com, where do you see those stories? Maybe we should have a 'powered by Perl' icon that links to that page?

Thanks! These are just some thoughts (meditations :) I felt like sharing, and I really appreciate all the feedback!

[dimar]

May you be at ease and worries dissipate ...

I'm just a code grunt without enough time to keep up with the latest in language theory or industry trend.

No need to be so self-dismissive. You know what "Longhorn" is, you've heard of joelonsoftware, Class::DBI, and you're on perlmonks. This proves you at least know how to look stuff up. Unless you are in academia, most 'industry trends' consist of simply interpreting this season's new fashionable 'buzzwords' and translating that into what you already know.

(I am) lazy enough not to want to switch to something new ... at least until I find something even better.

Unless you are willing to at least compromise on this, this single attitude will be the biggest source of innumerable woes, worries and frustrations for you. It will give you grey hair. It will rob you of sleep. Many have been there before. Why? Because being the "best" technology is neither necessary nor sufficient to being a "widely adopted" technology. In fact it frequently happens that popularity is inversely correlated with "elegance". How much you wanna bet (for example) more people know what is XSLT than know what is Haskell?

There's no need to succumb to the popular fall-back of assuming "those other people" ignorant, uninformed, or idiots. Lots of intelligent people recognize the grass-roots greatness of perl, but who also recognize the circumstances of mass-marketroid top-down corporate-driven "innovation" and adapt to it as necessary without compromising their ideals.

Some people really dig the metric system, base 12 arithmetic and Dvorak keyboards. It would be a shame, however, for them to lose sleep and worry just because not everyone else 'gets it' ... you might as well worry about bio-engineered quantuum nano-bots that render *all* programming languages obsolete. Unless and until that happens, you are definitely safe to know and contribute to perl, which is *not* going away any time soon, and don't be unwilling to play the 'buzzwords game' if that is what it takes to put food on the table.

[Solo]

where is PHP getting more traction than Perl from?

phpBB is what I observe as the PHP killer-app. And many introductions to PHP come from trying to add new pages to or install plugins for that board system.

[Juerd]

phpBB is what I observe as the PHP killer-app.

But many didn't expect it'd really kill ;)

[Aristotle]

That's much like the position Ultimate Bulletin Board used to occupy, then.

[zentara]

Want a reason to advocate Perl over PhP? Do a google search for "Santy virus". I fear this is only the beginning of the problems for PhP.

[b10m]

Ah yes, and we, the Perl People, don't suffer from Matt's spammable formmail.pl (and close to every other insecure Matt script). Please grep your web daemon's log for a POST on formmail.pl|cgi|...etc

Yep, I know of the NMS Project, but still hordes of insane people easilly install CGI scripts they find somewhere and don't understand. Also, no one is forcing you to use a horrible thing like phpBB. There are PHP alternatives.

Like always: the more something gets used (phpBB, Microsoft Windows), the more malicious attacks will be written for(/against) it...

[zentara]

Like always: the more something gets used (phpBB, Microsoft Windows), the more malicious attacks will be written for(/against) it...

Well that is sort of the point I was making, as to why Perl is better than PhP ( or linux is better than MSWindows). It allows you to have more variation in your scripts, making it much less susceptible to attack because everyone is using the same "canned code". Nature dosn't make everything "all according to the same recipe"...it introduces tremendous variations so that one attack vector dosn't wipe out everything. Perl's strength is it's "variations".

[Aristotle]

Here's another point that I'd've made here even without the recent appearance of a conveniently related article on Perl.com: mod_parrot will probably be a huge boon in the web arena.

PHP/Parrot, Python/Parrot, Ruby/Parrot and others more are in various stages of development and will probably see the light of day much sooner than Perl6. Because these all target Parrot, mod_php can be written in PHP/Parrot, mod_python can be written in Python/Parrot, and mod_perl will be written in Perl6. Since the other languages can target mod_parrot now (as in much sooner than Perl6), mod_parrot will mature long before Perl6 hits the stage. These other languages will also provide experience with the implementation of mod_foolanguage on top of mod_parrot.

I forsee that as a result, once Perl6 is implementable, an initial implementation of mod_perl will happen in next to no time flat.