PhpStorm evolution stopped?

Seems like after 2.0 version phpstorm evolution stopped. 2 years past but still no framework support, no project management. And PHP api is still closed so you can't write your own plugin to add your framework support. I see a lot of small improvements and I apprecitate them. But what about new features, about very important features?

6 comments
Comment actions Permalink

We do have big plans and huge features in mind - much more than we ever publically announce. Some of that are quite revolutionary.


Yet throwing the new features in imposes a huge trail of bug fixing. And it requires an infrastructure to base upon.

While some features are "highly wanted/voted" they are in the same time a) really hard to do properly, b) have workarounds, and c) have small impact on *actually* making users choose our IDE instead another. Example - FTP support, project management. Unfortunalely these two are low priority for us in the resource allocation list.

We are focused on better refactoring, code inspection, technology interconnection and code writing/modification experience. Focusing on PHP/JS/HTML/CSS part.

So we do recogenise the impotance of framework support - that's the place where IntellJ IDEA is also shines above all competitors.
But this promises to be most resource-grabbing and we *need* to make sure that we doing it effectively. We don't want to have rewriting it from the scratch for each new framework quirks. Its actually requires architecting proper API facade and extension points to plug stuff in. This work is in progress.

Framework support is also intersects with having 3rd party plugins leveraging our PHP PSI (Program structure interface) - its still evolving much and beeng so heavilly modified will break all 3rd party stuff almost each EAP. This is pretty hard to manage judging our experience. 3rd parties usually a) highly overestimate their urge to cope with such problems and b) utterly underestimate the complexity of work requried. A public API will hinder intertnal evolution and we can't allow that to happend yet.

Still, we have some advancements in these areas in *nearest* plans - i.e. a) pluggable class-by-name selection strategies and b) generic factory support on the top of the list.

With frequent releases and upgrade subscription we bring users a steady stream of improvements in all areas.
0
Comment actions Permalink

Hmm, I've seen lots of progress in PHPStorm...so much I was really hoping 4.0 would be released before April 28th[when my last resub expires].. and I plan on resubscribing in a few months after I've got some more project work.

A lot of things which are "stalled" are, in all honesty, not so important.   For example, the ability to create a set of folders/directories based on a template.  Considering that PHPStorm supports Phing natively, being a PHP coder I certainly COULD do it all with a phing task - so really a lot of "must have" features ARE within my control.

At the end of the day, I've tried over and over again to like Eclipse....nothing works.  It just doesn't meet my mindset in how to work.  PHPStorm and Netbeans work incredibly well for me, and PHPStorm edges netbeans a bit out for PHP coding.

0
Comment actions Permalink

Sounds great, but I hope you do not mind one sceptical remark (wish) form me:
I hope you will do some extensive research into realistic and optimal work-flow(s) for web developers and web designers before commiting to huge changes and feature implementations? Working with web projects requires some (not all) very different tools and structures than for example Java or C++ programmers.

c) have small impact on *actually* making users choose our IDE instead another.


If you pull it off to actually cretae a web-ide that is structually different form other programming IDE's that take the way web developers and designers work (and not force them to work like programmers) I think that would actually add another advantage that php/webStorm has over other IDEs.
Project management is one of those things: if, and only if, implemented correctly with a firm understanding of what project management would have to look like in our line of work.
So again, it sounds great but I honestly hope you will do proper research into workflow and special needs and take seriously into account that our tools are not exactly the same as the tools of other programmers and scripters.

:)

0
Comment actions Permalink

web developers and designers work (and not force them to work like programmers)

I am an experienced developer and I really have no idea what this statement means. I'm curious, because I LOVE PHPStorm and it is far and away the best product on the market for PHP develoment (heck, I even use it for terrible language called 4GL, with no lexer and still makes it very usable.) I also do C# development and I'll be honest, and PHPStorm's work flow management is 100 times better than VS in my opinion, and I hate having to "manage" all of those projects from one IDE.

So, I what exactly is it that you want? If you are "switching" servers, why do you need that in the same view? I do this all the time, and from PHPStorm and it works great.

Seriously, not trying to slam you or all the comments in this issue, but trying to understand what it is that is so lacking. I've used Netbeans, Zend Studio and Eclipse extensively and their "project management" is not any better and is really just more cluttered (and PHP code writing is leagues behind.)

Don't get me wrong, I'm didn't drink any koolaid (see all the tickets I've filed and complaint that I've had over the last couple of years), but this is just one area that appears to be of big concern that I don't get.

0
Comment actions Permalink

Hi Keith,

Let me start by saying that I think everybody has his/her own way of working so my view of specifics is hardly to be considered universal. What I do think however is that the issue you link to (http://youtrack.jetbrains.com/issue/WI-1432) is so popular because people:

1) have been used to working with IDEs that give them ability to have more than 1 project open at a time in the same window, and

2) not everyone thinks of project management to be the same thing.

I personally think of project management as something that transcends the IDE. It is everything related to a project like estimates, invoices, time-tracking, documentation, contact details, etc. etc. However, I do not think that all this should be centralised in one program, let alone an IDE (why would I want to create an invoice from PhpStorm, that would just be silly). What I do think however, is that certain areas of project management do belong inside an IDE, like status, category, version control, synchronisation, passwords, resource libraries, etc. And furthermore it is my philosophy that anything project management related should have an easy, universal API.

For example, I have all my projects listed in an MySQL database. This Database is used by a program I wrote myself that links my projects to my bitbucket account for VCS, to my google account for contact details and another program of mine also connects to this database in order to link administrative tasks like invoices, estimates, project status, etc.

And finally I have a third program running (all those programs in php so they can run locally and remotely) that links all that to a database that stores the location of files (file management) like designs, documents, assets, source files, remote server, test server, etc).

For me, it would be enough if I could you establish a 2-way communication with PhpStorm for project management. So I can store and retrieve stuff like login credentials, status, server- and file-locations,  etc. For others it would be enough to be able to open 2 or more projects at once.

For example, I have currently over 100 projects and therefore more than 150 servers defined in PhpStorm. Without the ability to group them as as active or archived it is horrible to maintain if I need to switch between developemnt and test or production server. While I only have to do this my selecting a setting from a dropdown box in my own program to do so.

Example: When I set a project to 'archive' in my program it automatically creates archives of all document and assets and checks if all invoices are paid. If so, a complete backup is created of everything related to that project. The source files are made available to the client in a shared folder online so if something happens to me, he/she has access to it. Any automated processes like regular database backups of online checks are deactivated.

All this is done simply by making a selction froma drop-down box. If I can make this php-script communicate with PhpStorm I could also change the source location (from active folder to archived folder) change the category of the project (from active to archived, decluttering certain views), as well as the servers, so they no longer show up (until I actually need them). etc.

I am from 1967, and I have been lucky to have had access to computers from age 10 upwards. One thing I grew up with and has always been my philosophy is the promise computers brought us in those days:

store once, retrieve infinitely

Unfortunately, computers never delivered on that aspect. This is mainly a human flaw. take contacts for example. I do not know about you, but I need to keep track of multiple contact lists (google, skype, msn, telephone, etc.) sometimes you can sync (with errors) but most of the time you cannot.

The time and subsequently money wasted by the inability to create an international standard for something as common as contact management is no doubt huge.

Strange enough all these things are always thought of as 'extras' and are mere afterthoughts while I think they are at the base of everything. Tools and techniques are nothing if there is no structure. Structure is the foundation onto which things are built.

It's one thing to have great idea how you can better edit a php file with code completion and what not. But better would have been to understand the position and timing of that php file in a project.

Again, practically my views are my own and I do not suspect anyone to agree with me on those. But the basic theory of structure is -I  maintain- solid. The work-flow is important because it determines when to use what and how.

My argument is that no IDE builder probably ever sat down and made such a list or talked about practical management problems an IDE could solve. It is a step most people seem to think does not belong to an IDE while I think that since an IDE itself is PART of the workflow it should 'fit' into its place within that workflow and therefor know what that workflow looks like.

If you found a way of working that does not involve writing down stuff on massive amounts of paper, switching between programs at least 3 times each minute in order to copy, paste, look up or whatever, and it does not involve redundantly repeating stuff in order to accomplish 1 task: then I would love for you to share your workflow with me please?

0
Comment actions Permalink

Im in the "Works fine as it is" camp.

I used to use eclipse and found it weird not to have everything I ever work on all visible in the editor, then I got used to it and its fine now.  Admittidly I sometimes have 3 instances of it open at one time but putting each one on its own desktop (see KDE) with the associated browsers for that project works great. If i need to copy from one project over to the other a ctrl+c in one instance can be ctrl+p into the other.

One thing i do see appearing on occasion is that when I open the editor for the first time, all of the last sessions instances open.  Im hoping this is not the result of 'Project management'.  because that would suck.

in the file manager any project that is a PhpStorm project has a different file icon than a normal folder.  By 'Project Management'  are people wanting a place to view all of those folders without any of the surrounding operating system folders?

Eclipse suffered from slowness and the need to limit searches to 'workspaces' in order to just search the current project.  Please don't build that into PhpStorm.  ctrl+shift+f is good for searching the entire project now without having to define where the project starts and ends.


-- edit --
sorry Gary, this wasn't exactly in reply to you, more just to the thread.  oops pushed the wrong button.

0

Please sign in to leave a comment.