As Oracle axes support for GlassFish, MySQL users might want to pay attention | Hacker News Reader

As Oracle axes support for GlassFish, MySQL users might want to pay attention

parelastic.com

33 pointstanglesome41 comments

The concerns over MySQL might be justified, but have little to do with GlassFish. GlassFish has never been a popular open source project like Tomcat, Jetty or JBoss, and probably not as commercially successful as Oracle's own WebLogic, or IBM's WebSphere, or JBoss. There was really no good reason to keep it alive, other than as the JEE reference implementation, a role which, as I understand, it will continue to serve.

Also, the compatibility between different Java EE application servers is far greater than between different DBs, and migrating from one to the other is usually very simple.

I do hope (and would like to assume) that only the work on the GlassFish server will be reduced, while popular subprojects like Jersey[1] will continue as before.

[1]: https://jersey.java.net/

Jersey is an integral part of WebLogic as well, so development will assuredly continue.

From the jersey mailing list:

> Jersey remains the JAX-RS implementation in WebLogic as well as JAX-RS RI in GlassFish. The decision to not provide a commercial GF 4.x offering is orthogonal to Jersey and JAX-RS development.

> ...and now, just to be safe, I should probably add the standard Oracle disclamer :) Here it goes:

> The preceding is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle’s products remains at the sole discretion of Oracle.

> Marek

I think the difference is that Glassfish is a piece of garbage. Among the few people who have heard of it it has a reputation of being the worst application server you can possibly use.

There are a couple applications running on top of glassfish at $DAYJOB but I can't say I've ever been involved with them. Care to elaborate a little on why it's so bad? I saw the admin console once and it seemed to run like a slug but it did seem to include a lot of features.

The admin console is easily the slowest part of Glassfish, but it lets you configure everything that, if you were using Tomcat, would need you to play around with XML.

I haven't used it but there is NO WAY that it is worse than WebSphere.

In any case, my firm belief is that the whole concept of app servers is flawed. Libs are so much better than frameworks, and app servers are just horribly clunky frameworks, which take away as much control as possible from the dev.

Taking control away from devs is a really backwards way of thinking of development as grunt work, which has been proven time and time again that it isn't.

Death to all app servers, I say!

I actually passed on some repeat client work when they insisted on using WebSphere.

My initial project for them was an eternal struggle to get basic things (XML parsing, Hibernate, CDI, etc) working in their environment because specific versions of these things were already included in the app server, and always ended up conflicting with libs we needed to include for our own app. I remember it distinctly as one of those times where I just felt helpless to make anything work.

I was in a similar boat. I abandoned a potentially very lucrative startup because the owner (my friend's brother) insisted on using Websphere and all the associated apps (he was a former IBMer). There came a point where some misconfiguration occurred when attempting to restore a previous version of the application after that point nothing worked. I tried for days on end to fix the problems but nothing would compile without massive errors. The piles and piles of cruft that Websphere added did nothing but obscure the real source of the issue. Eventually I gave up and said screw it and left.

The startup was already on its last legs at that point, as I was the only active developer since the other developer I worked with had already left after getting fed up for other reasons. After I left the startup limped along for a while but as far as I can tell was abandoned. If they had just used a more sane environment I could have easily developed the entire thing myself as it wasn't anything more complicated than a straightforward CMS + payments. I still kick myself for not making a better case for a more sane environment but I was wet behind the ears at the time and I didn't trust my instincts in the face of guys with masters degrees and X years at IBM and other top companies. I know better now.

App servers solve so many problems so you don't have to. Taking control away by solving problems to let you focus on the bigger ones is great.

Can you be more specific on your issues with WebSphere? I remember the buzz when it first came out, but haven't bumped into it in quite a while.

In all honesty, I've never heard that about glassfish before. The people I know that use it, are very happy with it, and praise it over JBoss.

Would you care to explain with an example.

Definitely not the worst. I work with it every day and most of the time it behaves quite well. Glassfish 4 (Java EE 7 support) behaves relatively well, too. I do wish they made it more light weight and faster, but clearly it is good enough to be bundled with NetBeans 7.4.

I've been a fan of Glassfish and Netbeans for 5 years. It started when I was testing a few "stacks" for web services, and that combination worked amazingly well out of the box; or rather, it was the only one that did.

As opposed to MySQL, which is respected and certainly not a target widespread of hatred and mockery?

Have you ever worked with WebSphere or WebLogic? When I was, I wished I was using Glassfish.

Why, specifically, is it a piece of garbage?

Pure FUD. Oracle has owned Innodb since 2005 and now that they own the whole of MySQL impressive releases are coming at a steady clip. To date, Oracle has been a better steward of MySQL than Sun was.

A lot of the big distros have transitioned to MariaDB already (https://mariadb.com/kb/en/distributions-which-include-mariad...). Debian probably will soon as well, which will trickle down into Ubuntu probably.

I think there's plenty of room outside Oracle if they were to ever shut down MySQL. I moved to MariaDB, and it really was a drop-in replacement.

Compatibility document for the curious:

https://mariadb.com/kb/en/mariadb-versus-mysql-compatibility...

TL;DR: It's 100% binary compatible, but PHP manages to have problems anyway

/facepalm

I think it's only a problem if you're using mysql_ instead of mysqli_.

https://mariadb.com/kb/en/installation-issues-with-php5/

Since mysql_ is being deprecated soon (so they say, maybe PHP6 haha), it'd be good to migrate to mysqli_ anyway.

Shouldn't everyone be on PDO by now? If you're still on mysql[i]_ you must be used to having problems.

Probably. I'm a pretty poor developer. I'm a sysadmin who sometimes writes PHP apps for work. So, I'll look into PDO. :D

Definitely recommended! It will also help you when you swap out MySQL for something else later on... little to no rework!

If you've got the time, try writing one in python and see what you think.

lmm: (For some reason your post doesn't have a reply link) These are usually webapps for department-specific needs, and I generally find PHP to be easier. I have a few python things that I use on the backend, as well as a plethora of bash scripts. But, I admit to knowing PHP more than Python. (I'm in a smallish company, so I end up wearing a lot of hats)

generalpf: I will do so! Do you have any go-to references that you prefer? Or shall I just dig through the PHP docs?

> For some reason your post doesn't have a reply link

I think that for a comment at some level of nesting N, the ability to replay is delayed by Y seconds. IIRC this is to prevent stupid back-and-forth arguments.

It's really quite simple, read this and get to work! It also explains the 'why'.

http://wiki.hashphp.org/PDO_Tutorial_for_MySQL_Developers

I maintain an early 2000's php website that is pretty functional and has a lot of features, but uses the classic spaghetti code. I can't exactly just switch to PDO without considerable time investment, and otherwise things are working just fine. I don't think this is an uncommon scenario.

I converted an OSS application that I wrote back in '03 to PDO in just a few hours. It actually helped me find security holes because of strong typing in bound parameters. Have a go!

Should be, absolutely.

Are, nope.

PHP's long-lived popularity means there's a huge amount of ancient PHP articles and tutorials still out there on the web that use the old library. Because these articles are both old and widely linked-to, they tend to come up when newbies search for PHP help. So they learn how to do things the wrong/deprecated way.

Resources like PHP the Right Way (http://www.phptherightway.com/) help counter this, but it'll continue to be a problem as long as using the old library doesn't cause any obvious problems like immediate crashes, screaming in-your-face error messages, etc.

More details: https://mariadb.com/kb/en/installation-issues-with-php5/ It's apparently not an issue with 5.4+ and has several fairly simple workarounds.

Why? MySQL has a well-maintained drop-in replacement outside of Oracle control (MariaDB). It's even mentioned in the article.

(And no one's ever heard of Glassfish, probably the reason why Oracle are dropping it)

Agreed, every MySQL project I'm on is actually a MariaDB project. While I prefer Postgres and have arguments to present to clients who rely on MySQL, a fear of Oracle is not one of those arguments because I just don't think it matters what happens to MySQL proper.

The way they closed down OpenSolaris told me all I needed to know about Oracle's plans for every other OSS project.

We did use GlassFish with JRuby on a project early on in my career, that's about it.

I wouldn't worry too much, since MariaDB is out there.

I'd love to make the switch from MySQL, except Amazon RDS doesn't support MariaDB yet. But since they just added PostgreSQL, maybe there's hope? Presumably MariaDB support would be vastly easier, as easy as a new version of MySQL?

I think that would be sufficient motivation to get me to use PostgreSQL. I doubt I'll go running out to buy an Oracle database when there's sufficient alternatives (including MariaDB).

I think what's more likely is that MySQL goes the way of Hudson/Jenkins. Oracle strangles the open source community and pushes people to use a fork of the original software. Oracle's in the database business, and I just don't see any reason for them to kill off one of the most popular databases in the world. More likely is that they make it serve their purposes to the exclusion of the open source community's.

I switched over to MariaDB on my most recent Linux install, and honestly have no idea that the database is different. Moved my Windows install from MySQL to MariaDB as well. Was a bit more complicated, but not by much.

This kind of makes sense; they have WebLogic to sell to enterprise customers.

Reply on news.ycombinator.com