Enterprise PHP

[Dr Livingston]

1. MS is the best way for everything
2. If you don't use .NET you're an idiot
3. .NET is better than everything
4. Anything Open Source is useless and crap
5. Proprietary's is by definition better than anything open source
6. PHP has nothing to offer
7. Enterprise is me. I define it all.
8. PHP's devleopment model is useless, and has nothing positive at all, in fact you're better off coding using Assembly than using PHP

Well dip me in ****e and roll me in chocolate ? The above I quote is the exact vibe I get from your posts.

Okay ?

[Taoism]

I am using build 131 of PHPEdit here.... it is the best IDE IMHO (the support list is shaking down 132 right now, and 133 is coming along)

Cheers,
Keith.

[Jeremy W.]

Well then I must have said it a hundred times then, right? I mean, because it's obviously me who's closed minded, ignorant, biased, etc, because since I believe MS will eventually save us all from every evil in the world (in fact, Bush should send MS after the terrorists, they'd fix it in minutes... well, hours as it'd take a few hours to install win2k3), it's impossible for me to have an open mind.

Perhaps, once again, people should read my posts instead of inferring what they think I mean.

[Dr Livingston]

PHPEdit 1.3.1 ? Umm... But are you using Win2000Pro ? If so then I could easilly assume it's my box that's acting up no ? I had no problem using an earlier version of PHPEdit on Win98 a while back...

Though personnally I didn't take to it too much - only downloaded the latest version to see if there were any improvements really

[Phil.Roberts]

Off Topic:

PHPEdit ? Umm... I would need to disagree with that. I installed the latest version on Win2000Pro the other day there and bang; Windows locked up...

Using PHPEdit v1.7.132 on Win2K Pro hasn't given me a single problem so far..... Don't use 1.6 even though it IS marked as "stable". It's buggier than hell and generally not very good.

[Taoism]

PHPEdit 1.3.1 ? Umm... But are you using Win2000Pro ? If so then I could easilly assume it's my box that's acting up no ? I had no problem using an earlier version of PHPEdit on Win98 a while back...

Though personnally I didn't take to it too much - only downloaded the latest version to see if there were any improvements really [img]images/smilies/wink.gif[/img]

I meant 1.7.131

I use Win2k at work (with some Unix on our AIX servers thrown in for good measure - but only using telnet), WinXP Pro at home....

Cheers,
Keith.

[Dr Livingston]

Must be my box then folks yes ? I built the box myself about 18 months ago I noticed though on Win98 I installed Empire Earth first and it worked great - what a smart game btw - and then I installed it again later after an uninstallation and the damn thing doesn't work.

My box is haunted folks... Either that or I need to strip the box and begin a new build again ?

[M. Johansson]

PHPEdit ? Umm... I would need to disagree with that. I installed the latest version on Win2000Pro the other day there and bang; Windows locked up...

Now I'm using DW MX; Still some way to go IMO ?

just to set things straight for ya: DW MX is not an IDE. It's not comparable.

Jeremy W - I am wondering if your on the payroll of Microsoft by any chance ? You seam to intent on promoting .NET and Microsoft to the point of being shear ignorant of the fact that there is other alternative technology available.

You know this is the lamest argument that can ever be produced in a OSS debate, right?

[Dr Livingston]

DW MX an IDE ? Umm... Don't recall actually that I stated that DW MX was an IDE ? I just mentioned the fact that was using DW MX correct ?

I know DW MX isn't an IDE. You know it's not an IDE; hell the whole world knows it's not an IDE...

[M. Johansson]

Oh, okay then. I misunderstood - Phil said PHPEdit is turning into a fine IDE, and you said it was crashing on your system, and that you now use DW MX, which suggests that DW MX is an IDE. I just wanted to set things straight, so nobody would be confused.

[brainpipe]

What IDE(s) are you using?

None. We are discussing Zend Studio, but most of us use EditPlus and/or vim. (Also, I was wrong in my estimation of the records generated per day: it's more like 40-50 million.)

My two cents on the rest of this discussion: I think the answer to the entire argument, as some people here have already said or at least hinted at, is the definition of enterprise ready. The company I work for is a multi-million dollar company serving clients much, much bigger and clients much smaller. The system we've built is a replacement for and upgrade of several different applications that I think most people would call "enterprise-ready;" products from companies like Microsoft, HP, Sun, etc. Can we do it using PHP, Perl and MySQL? Yes, we can (to my surprise). Should we? I really don't think so, and most of the developers I work with don't think so either. We constantly have to make adjustments or rewrite code to deal with MySQL crashing or disappearing; to handle low-level application functionality -- essentially creating things like java beans in PHP.

So if enterprise-ready means you can make it work for enterprise systems, then yes, PHP is there. If it means ready to work for enterprise systems out-of-the-box, then no, PHP has a very long way to go. But I really don't see the point in arguing about it. Why not throw all that mental muscle around making PHP more powerful?

Once you know a language inside-out, and can call yourself (honestly) an expert, it starts to lose its luster for you. It's easy to get caught up in a language that's relatively new to you and think it's the ultimate solution. But there is no ultimate solution. PHP is a wonderful language imho, and I absolutely love it. Even so, if I were designing the system I've helped build at work, I would have chosen Java and either Postgres or Oracle for the lower-level work, and PHP for the front-end. To me, that doesn't mean PHP is somehow lesser. It's different, and I'm glad it is. Same goes for MySQL -- I'm personally kind of disappointed that they're adding all these new features in v4+, because I'm worried they'll lose what I like so much about it: pure, unadulterated speed. If I want a web site that uses a database for persistent storage, and I just want to store some preferences for my users and some content (or even a bigger CMS), I'll generally go with PHP/MySQL. Doesn't mean Java is lesser, it's just not the best tool for the job.

[M. Johansson]

Excellent post, brainpipe! I see much good coming out of your presence on these forums.

[Dr Livingston]

Good points - I would have to agree with you on the point about PHP being enterprise ready ''out of the box'';

A lot of folks shy away from having to install and configure Apache (for example) with PHP and a database simply because they have little or no experience or lack of time or in most cases just don't see the point that they should have to do this task.

ie This should automatically be done for them on a click of a button; there are some solutions around for installing the 3 (Apache, mySQL and PHP) though I've not looked at those myself so cannot comment on their effectiveness ?

[Karl]

I've looked at a few of the sollutions for when I can't be bothered to setup a Windows test box manually, but I'm not overly impressed, so I usually end up doing things manually anyway.

Re mySQL, at 40-50m new records per day, I'd say you're nuts using mySQL But you obviously think the same too.

[Dr Livingston]

Would like to know the number of hits on Yahoo's Databases...

That'd be interesting huh ?

[Jeremy W.]

It's all selects, you don't want to be creating that many records with mySQL.

[Dr Livingston]

SELECTS ? I know but just how many ?

Apart from maybe Google, there is no other Site that could take this kind of pounding right ? Not even IBM could come near the number of hits Yahoo has to take every day.

[Selkirk]

Well, I used to sit on a five person committee that evaluated and recommended programming languages, databases, libraries, and other development tools for a manufacturing company with $2 billion annual sales, so let me add my experience in enterprise software arena:

First, The Yahoo decision was probably as much a result of them being an Open Source shop as anything else. If they were a Microsoft shop it would have been ASP.NET. If they were a sun shop it would have been Java. If they were an anti-microsoft shop (they exist) it would have also been Java. It takes a tremendous advantage to overcome familiarity.

Second, enterprise class products come with enterprise class sales forces. For an enterprise class purchase, they will come to your office and a dog and pony show. They come in pairs: A tech guy and his pimp. They will load your data into their product and answer ignorant questions from your boss. They make everyone feel good about a purchase they were going to make anyway. When you call them and ask them a question, you get an answer back the same day. When you are unsure about their future, they hand out NDAs and setup conference call with internal developers and product managers. I doubt that Zend has this kind of sales force.

The sales force in a large purchasing decision acts as an imperfect proxy for what kind of support we think we will get after the purchase. That and free lunches!

Third, large companies can be very concerned about stability. One mission critical application we had was completely dependant on a library from a small company that went out of business around 1994. In 2000, I had to hire a contract programmer to write an assembly language patch to the binary to get it to work under MetaFrame. Yuk. I wont even talk about y2k.

Backing the wrong solution in an enterprise setting can damage your career. Experienced corporate decision makers are conservative and have a herd mentality. They like to go where everyone has already been. They have enough risk in their lives without inviting more. This is summed up in the old phrase: "nobody ever got fired for going with IBM." Its true. PHP is risky.

One of the biggest factors for evaluating a development tool is the pool of available programmers. This isn't just limited to certification. Frankly, PHP just doesn't have a good pool of enterprise class programmers. (present advanced PHP company excluded of course. ) My interpretation of the Zend survey is that most PHP programmers are more familiar with HTML than with transactions. I think this has to be what Sun sees in the recent PHP agreement: a source of cheaper labor.

Connectivity is a big enterprise issue. Look at all the different things that java can connect to.

The biggest things that PHP has going for it are a shallow learning curve and cheap abundant hosting (via apache). Shallow learning curve means nothing in an enterprise environment where you might have 12 people working on something for 7 years. Cheap hosting doesn't mean much when hosting is dwarfed by development costs.

Enterprise PHP is a catch 22

Look at the confusion that occurs in these forums about how to do MVC. The PHP community has not yet worked out a good set of enterprise patterns for using PHP. The catch is that until PHP gets used in more enterprises, these patterns are going to remain obscure and elusive and PHP will remain unready for the enterprise.

The best thing for enterprise PHP would be to get involved in The Middleware Company Petstore cases study.

When Microsoft first built the .NET petstore to compare .net to Java, the Java crowd laughed at them because their software wasn't constructed with enteprise patterns. Microsoft got serious and started putting work into Enterprise patterns and architecture and revamped their pet store for the enterprise.

Similarly, PHP participation in this will probably draw the same type of initial scorn from the enterprise java folks, but the process and criticism would help to focus PHP as an enterprise solution.

[M. Johansson]

Good points - I would have to agree with you on the point about PHP being enterprise ready ''out of the box'';

A lot of folks shy away from having to install and configure Apache (for example) with PHP and a database simply because they have little or no experience or lack of time or in most cases just don't see the point that they should have to do this task.

ie This should automatically be done for them on a click of a button; there are some solutions around for installing the 3 (Apache, mySQL and PHP) though I've not looked at those myself so cannot comment on their effectiveness ?

I might be misuderstanding you again again here, but we ar not talking installers. By saying that PHP is "not enterprise ready out of the box", he's referring to the fact that your developers have to spend a great deal of time (possibly hundreds of hours) on building a framework, so it can be used for development on a larger scale. .NET and Java already has those frameworks written.

[auricle]

I'm late to this discussion, but I couldn't let this one pass (from page 4)...

I see that the same thing seems to have happened to CD for example, at the time it was in direct 'competition' with DAT, DAT being on the market earlier, offering digital recordings at 48Khz up to 4hours on one tape, along came CD, offering only 44.1Khz at 25% of the recording time and you couldn't record it yourself!

The original DAT spec never allowed 4 hours of recording time. It was up to 2 hours at 44.1k, 48k and 32k. One manufacturer (Sony or Tascam, but I can't remember which) introduced a double-length 'version' which use non-linear 12-bit recording at 32kHz instead of linear 16-bit.

Still, due to keen campaigns we all believed the CD was it, we would 'never have a scratched LP ever again', we would 'enjoy the crisp sound of digital recordings in the comfort of our own living room', etc, etc...
I do have a couple of hundreds CD's, but if I can get my hands on it, I'll go for the vinyl, every day of the week and twice on sundays .

Vinyl, like any medium, has its own distortions -- at least it was full-bandwidth. What amuses (and disturbs) me is that there's a whole generation of people (perhaps some of them are present here) who will never know high-fidelity sound because of MP3 and other data-compressed formats.

What's the relevance to the current discussion? I like it when I can't hear the medium working -- that's what makes linear formats (CD, DVD-A, SACD) so appealing. Still, I have a lot of vinyl and as long as I perform the cleaning ritual every time I play one, each of my records should still perform without my awareness of their performing (surface noise, mastering compression, tape saturation from the original recording) being too obtrusive.

Similarly, I like it when I'm not aware of the mechanics of a web site -- the site just works. Whether it's CF, ASP, JSP, PHP, whatever, it doesn't matter. When the site crashes (which, in my experience seems to happen a lot with ASP apps -- YMMV), I'm out of there!

I suspect that a lot of people would gauge so-called "enterprise" applications by their reliability.

[Dr Livingston]

M Johansson - Misunderstood me ? Seams going by what you've posted now I've mis-understood yes ?

It happens and I can easilly admit that folks On this point, PHP is seriously hindered by the lack of a common and standard framework.

This IMO is not going to change for the forseeable future either; What I ask is what makes Java's Struts so great that everyone who develop's Java wants to use it ?

Then based on these findings cannot a PHP working group then design and develop a framework based on those findings ?

[Chris82]

selkirk: excellent, excellent post. (can't give you rep - must spread around some more) It is great to hear your insights in this matter.

I was about to comment on the many available libraries and the missing "best practices" development approach. But you already said what I wanted to say basically.

[Dr Livingston]

Yer, there is a LOT of sense in that there post; For PHP it's an uphill struggle but as always, isn't it so ?



I'll give you some respect instead ? ...

[Phil.Roberts]

It seems that the big stumbling block in any kind of enterprise development of PHP is always going to be lack of leadership from the top down. Ultimatly the PHP group itself needs to recognise its shortcomings in this area and take action on its own. It seems like a lot of time has been wasted in the wrong directions (i.e, trying to emulate Perls CPAN with PEAR). Outside groups will always suffer from the problem of not being "officially recognised".

[Dr Livingston]

Yer, but also we need a kind of standard that all other PHP development is for the most part anyway modelled on yes ?

M Johannson was talking about frameworks... Where on this Earth is PHPs framework ?

... We don't have one is the answer of course.