I have reviewed Blumentals Rapid PHP Editor twice before, first a casual review of the 2016 edition and then a slightly more detailed examination of the 2018 edition. This discussion will focus on the areas that are most important to me.

In that context, I recently published Expression Web: The Final Denouement. In that article I explain events that have convinced me that I cannot continue to use Expression Web (EW). This brought into sharp focus my need to decide upon a new set of tools. Having recently purchased Rapid 2020 and it being my primary "other" tool already, it was only natural that I would examine it very closely to see if it could take on the role of my primary Web development tool.

It can't.

Notable New Features

Rapid 2020 is mostly an update release, with minor improvements in most of its major sections. As usual, Blumentals' descriptions of new and updated features is vague and fluffy, making it difficult to understand exactly what has changed.

The most notable upgrade is to search, something I have complained about in previous reviews. Search has undergone an extensive rework.

The other important improvement is the ability to publish folders via FTP.

The IDE

Rapid PHP ("Rapid") is still deficient.

File Explorer

In the review of Rapid 2018 I listed things that could not be published from the File Explorer. All of these items can be published from Expression Web and most can be published from other IDEs and editors (e.g., PhpStorm, Visual Studio Code (VSC)).

• A folder.
• More than one file.
• A collection of files and folders.
• A collection of files/folders found as the result of a search.

For Rapid 2020, the following things still cannot be published from the File Explorer:

• More than one file.
• A collection of files and folders.
• A collection of files/folders found as the result of a search.

In other words, 2020 has added the ability to publish a folder, an important improvement and certainly a step in the right direction. Unfortunately, folder publishing is unreliable. I regularly get errors that stop the publish operation. The error messages are not descriptive and do not indicate why the failure occurred or how far along the publish was. The behavior leads me to believe that the FTP client built in to Rapid is not robust. This kind of unreliability and uncertainty is not acceptable.

In short, the ability to publish a folder is not an improvement at all.

When publishing a folder, Rapid launches a dialog listing all files in the folder and allowing the user to make a selection of contained files. I found this inconvenient because my expectation is that publishing a folder means just that. To publish a selection of files, I want to be able to make multiple selections of files/folders directly within the File Explorer. The Rapid solution is cumbersome.

I'm getting the impression that the file explorer needs an overhaul.

Another operation that is common to my style of work is drag and drop copying of a file or files from the Windows Explorer directly in to the IDE or editor I'm using. It works perfectly in EW, PhpStorm, and VSC. It crashes Rapid. By "crash" I mean that the program exits with no warning, no error message and, of course, without completing the copy. Failure to complete means the state of the publishing target is unstable.

After mentioning this in the Blumentals' forums, I was told by a WeBuilder user that this type of drag and drop operation was not supported. Drag and drop is a basic "explorer" operation and its absence, which I admittedly did not notice before, is a problem. But if it's not supported, the IDE should tell me that when I attempt it, either by prohibiting it (displaying the not sign when hovering over the destination folder) or by emitting a message to that effect.

Searching

Rapid's Search & Replace functionality improved greatly, mostly by integrating all forms of search into a single dialog. It's not quite as smooth as what I'm used to but I'll excuse that because Blumentals does have a customer base used to its style of search. The company thus needed to find an integration strategy that delivered improvements without breaking what people had come to expect. This improvement is long overdue but very good. I expect refinements in the future.

Along with improved search comes the dockable search window. This is a compact way to provide search with a dialog that docks to the edit window. I think it's a very good improvement and one that would be very helpful in a lot of situations. The organization of the compact window is very good.

Search results still lose their currency the first time an edit is made. I guess that's an item on a future list.

Project Management

There is no change here. Peter Klein's Project Workspace extension works in Rapid 2020 and remains an essential add-in.

Performance

My interest in performance is focused on two things.

The first is search. Rapid is faster than EW but not faster than VSC or PhpStorm. It's been delightful having that performance at hand.

The other is startup. I want an IDE to launch as fast as possible and, hopefully, faster than my twelve-year-old Expression Web. Modern IDEs examine the entire project being loaded and make all the connections so that elements like Rapid's Code Explorer can be populated for easy code navigation. EW doesn't do that.

Even doing all this work, Rapid gets running very quickly. Believe it or not, it's faster on startup than PhpStorm or NetBeans. 

Startup is a big deal. It directly affects productivity and is very important because the IDEs do not allow multiple instances. One project at a time, please. 

Conclusions

What has changed since my first review of Rapid PHP 2016 is that I can no longer depend upon Expression Web for its capabilities. I need to find a solution that provides the same level of productivity and capability as I had using EW alone and then the combination of EW with Rapid PHP in recent years.

The flaws in Rapid's publishing model make it unacceptable for my level of professional work. This is not to say Rapid has no place in the PHP world, especially given its price. But without improvements at a core level, Rapid is not good enough.

What's Next?

That's a good question.

Recently, in honor the tenth anniversary of PhpStorm, JetBrains offered aggressive, $45 pricing on a new license that I could not resist. (Sorry, the deal's gone.) Although PhpStorm is now a subscription product, the license includes a permanent "fall-back" to the version obtained at the time of purchase; no matter what, I have the right to use this version forever. I made the purchase so I could examine PhpStorm at my leisure, to understand it better, but I did not buy it with the intention of using it full time. I am now examining it in more detail with the thought that it might be the next step. It's a great product overall but has a publishing model that makes me uncomfortable.

I rejected Visual Studio Code out of hand five years ago. In this business, that's a long time and a lot has happened. The PHP support is better and there are some publishing extensions that look promising. I'm digging in to learn more and a friend is doing the same thing; we plan to share what we find. The big plus here is that it's all free. I remain a fan of Microsoft, believing it trustworthy, so sticking with a Microsoft product is appealing.

I can't ignore NetBeans, mostly because it's free. Studying it will be more complicated than other tools. It's too soon to say anything about it.

I look forward to making a decision.

Will I Continue to Use Expression Web?

Yes, but only for WYSIWYG editing.