Future features of ColdFusion

Want big features?
Let start asking big:
1) Bundle CF with A Limited connect server and LCDS Deparmental.
2) Add CFMeeting and generate a connect meeting room on the fly
add eventgateways to support listening for behaviors inside the meeting (like voting, chatting, enterRoom etc)
3) Eventually an interface to communicate with a SIP gateway?
4) Unlock all XDD limitations for CF9
5) Add policy server capabilities into CF
overall I think CF should be the aggregator of Adobe Enterprise Technologies

Hang on, so what you asking isn't necessarily "What would make CF better in the next version?", but "How can we get you to buy CF9?". For me, these would invoke two different responses.

For me, CF9 should contain what it does now, but improved. Almost a CF8++. For sales though, you'll need that "killer" feature. For me, I don't know what that is, as I don't come across a huge amount of problems day to day, other than doing heavy lifting on the data tier.

1. The ability to create and send mail items with <cfexchange>
2. A Flex based interactive log reader, kind of like http://flogr.riaforge.org ;)

That's all I have right now.

and upgrade flashpaper to AS3 format please!

Just thought of something.

Being able to code 100% in a script based format. Either Actionscript 3/4 or CFScript. I also think that would go a long way to improving CF's reputation...

My comment got swallowed so I'm going to give a rehash. Sorry if I was just impatient

Video: I still think that a suite of video tags and functions will be a killer feature. I know it might cut into Premiere but I don't think they really are in the same market. Allowing companies to create their own YouTube has some value. We've been doing video for awhile, and CF would have been a huge help to us: http://spike.wharton.upenn.edu/media/

Standards and Excessibility: Recylced from before but certifying that ColdFusion HTML output will be compliant and accessible would be a great feature.

Search: Alternatives to Verity would be great. I recently kludged Verity to work,because I didn't want to bother with Lucene, but that would be a cool additional tool to have in my toolkit.

Enterprise Integration: I frequently have to drop down to Perl to deal with an legacy application with a bizare interface. In most cases I could have coded an alternative if I had a cftelnet, cfsocket or cfsnmp tag to use. I would think being an easy bridge from old technology to all of the new options you have included would sell some units.

Identity and presence Management: I spend a fair amount of time coding the connections to tie this IM account with that domain account with this application user. I don't know if there is anything ColdFusion can do to make it easier, but figured I would add it.

Was that more of what you had in mind, Jason?

+1 on ditching YUI and adding EXT 2.0
I love the SIP integration idea.
Ability to consume gzip'd content with cfhttp, cffeed, etc.
+1 on multiple search engines:: lucene
Leverage Java Frameworks such as ehCache and Acegi

I guess my first set of suggestions weren't juicy enough (http://www.cfinsider.com/index.cfm/2007/9/14/What-...)

Round 2:
11) cfDrupal

12) I'm not an OO zealot, but for those who are... give them as much OO capabilities (keep them *optional*) as possible. We don't want to continue the trend where OO developers from Java or .NET backgrounds completely ignore/ridicule CF because it isn't fully OO.

13) CF .NET - I know you can integrate somewhat with .NET in CF8, but Adobe should put together a BD .NET version that blows the BD version away because it has all the CF8/9 capabilities, including Event Gateways, Flash Forms, Flex and AIR integration, etc.

14) hooks into PHP and Ruby (and possibly other major languages)

15) to add to other comments about a "real IDE"... I'm guessing they're talking something as robust (though better than) Visual Studio .NET.

16) cfCaptcha

17) thinking of ways to bring folks from other languages to CF.... maybe like .NET let's you write in C#, J#, VB etc. and it compiles to .NET, CF could do something similar with cfscript syntax?

18) cfGoogleAPI and cfYahooAPI

19) cfAIRWidget - wrap it around stand-alone CF/HTML/AJAX/Flex code. Save it, and it compiles into an AIR widget that you can distribute to customers and the like... Then, us CF developers can start distributing our own ideas for FTP clients, mini browsers, IMs, IDEs, media players, DB clients, etc... and we wouldn't have to learn .NET, C, or Java to make it happen.

20) cfVOIP

21) cfMakeMeDinner

22) cfGetMeAGirlFriend ;D

+1 to KTML if only it were possible.

Hi Jason,
Some of this is a little high-level/abstract - I'm not sure what level of detail you are looking for.

1) Better integration with Adobe Chats. What I want to be able to do is offer webcasts from inside my branded site.

2) Workflow extensions. I know that Adobe has workflow tools - can we integrate them into the CF framework? Many CF installations support intranets. Adobe has great document support. CF now has document/PDF integration. The next logical step, for me, would be introducing workflow.

3) Extend CFPresentation - allow better user interaction, for example form submission. I want to show a user some slides, and get their feedback via form submission. I was told I could do some Flash remoting for this, but the CF integration with a flash form would be better. Also, maybe branching based on user feedback (to different slides)

4) A built in Survey tag (CFSurvey), that supports branching.

5) Built in SAML support (CFSAML)

6) Reporting enhancements. I have to be honest, I have to look further into CF 8, but this is what I'm thinking. Flex has great flash based graphs, but they are quite a bit of work, when it comes down to it - with labels, tips, etc, etc. CF reporting is easy, but a little too static.

Here's a quick example of how to enhance charts, is to enable "binding" on charts. If I click one area of a CF chart (or CF report) another chart will change / get populated. Kind of like with Flex, but it doesn't need to be super complex.

Also, introduce AJAX features into CF Report. (yeah, sorry, I know I'm vague)

I'm in the healthcare industry, and I can't tell you how important reporting is.

7) AIR Support
I know the last one is a hot button category, but consider this: The end result of a CF app is HTML/AJAX. I know that its server based. However, if there were guidelines for creating AIR apps (example, not in-line SQL, need to use back end CFC's). The CF/AIR "wrapper" will need to know the base URL for the CFC's and will access them as web services. I'm not expecting it to be a "one-step" process.

Well, you asked for "big picture" ideas, and there you have them! I'm thinking off the top of my head here - so some ideas are a little incomplete (or just plain crazy). I'll post more thoughts if I get them, or maybe we could catch up for a beer at MAX (I'm on intronetworks).

Cheers,

David

So in addition to the developer-specific suggestions I left <a href="http://www.cfinsider.com/index.cfm/2007/9/14/What-... the other entry</a>

The video suggestions are really something... Yahoo Video, GOogle Video, YouTube, etc... are all based on Flash Video, and who knows more about manipulating Flash Video than Adobe?

Seriously, ability to input, merge, reencode, and clip video together, then output a video player swf would be pretty cool. Of no use to me in my daily work, but then most of the "Killer Features" really aren't.

While I think the <cfgoogleAPI>, <cfdrupal>, KTML, EXT2 are fun ideas, It seems odd to me to integrate support for specific software packages that only a select group of folks are using into the core language. One of the big advantages of ColdFusion used to be that it worked with any database server, and you didn't have to change your code (much) if you were moving from Oracle to MSAccess to SQL. Likewise, it would be nice if <cfsearch> worked with Inktomi, Google Applicances, K2, or Lucene, and <cfexchange> was deprecated in favor of <cfpim> which would talk to Exchange, Lotus Notes, iCalendar, or any of a variety of PIM softwares.

Sorry for chiming in again Jason - but in general, I'd like to see CF9 be more integrated with other Adobe technologies. Obviously, I mean integrated further. Flex and PDF features are great, of course.

We're researching the design/develop workflow here, and on the face of it, the CS3 tools look great (specifically, we're looking at starting with Fireworks). The end result seems to be HTML, which is great, but I'm not too sure it takes advantage of the great new CF features - cfimage, just as an example, or some of the AJAX controls included (there is spry integration, though, I guess).

Again, I'm being vague, I know, but I wanted to spark some thought (as you requested). Let me put this into perspective - we're trying to be more of an "agile" development department, and the particular methodology we have selected hinges on creating the UI templates first - storyboarding, if you like. Eventually, we'll have to get that over to CFML, and the details in between are a little murky right now. So, it may not be a CF9 feature, per say, but could be an integration effort.

Again, hit me up if I can be more specific, or just hit me, if I'm talking jibberish.

Cheers,

Davo

I was excited when Adobe choose FCKEditor because I thought they did it for it's available Coldfusion Image Library and File Browser plugin components. I was really surprised to find out that the built in variety of FCKEditor didn't include either of those. So for CF9, I hope that the Rich Text Editor makes uploading files and photos really easy as well! It would even be great if the photo uploader was enhanced to make it easy to resize (and reduce file size) the photos using CFIMAGE.

I know that's all something I could do myself today with CF8, but it's needed over and over in so many applications that I'd prefer to just have the tool prebuilt in CF9 so that I just need to configure and run in my own apps.

Preferably the File and Image browsers would support multiple libraries as well. (So that we could have different libraries per user rather than one big library per application.)

I would love to see CF9 be supported on a Debian based Linux distribution.

This is more based on a preference over Redhat based distributions.

I've noticed that CF8 seems to plug along very nicely on my Ubuntu development machine, so I am hoping you guys are already half way there.

A better way of loading Java classes and libraries than the usual system we have of placing them on the cf server class path, restarting, etc. (And not using Mark's JavaLoader). CFJAVALOAD

A way to do java scripting.

A transaction manager.

What about a integrated version of FMS? - similiar to the LiveCycle Data Services ES that was included.

video conversion tags

Server level application firewall security similar to mod_security or even mod_security support

gzip compression - ability to turn on/off gzip compression at the server/application level

What about adding Stored Procedure support inside the report builder? I know you can pass in a query to an already built report, but it would be nice to show the list of stored procs along with the tables and be able to use them in the report building process.

I might be a little late to the party on this one, but how about a free production grade version of ColdFusion Server? This would be a version that would specifically address the "Why would I buy ColdFusion when ASP/PHP is free?" arguments. I have run into these arguments all to often because executive management generally does not consider the total cost of ownership when it comes to these platforms. This is especially true for companies that are not IT or software companies.

I have found that the cost of ColdFusion at times falls into an unfortunate middle ground of expenses. It's not the typical item that has to be carefully considered in a multimillion dollar annual budget (like a datacenter expansion). At the same time when the question is asked "What can we cut from the budget?" the expense of ColdFusion does stick out enough in an itemized task group budget that it often gets singled out (at least in my experience). And these discussions often do not include advocates of the platform. It is the Directors and VP's that have a budget to balance that make these decisions.

This proposed free version of ColdFusion would have to have enough functionality that it makes sense to use in a production environment. This version would be targeted at small to mid-sized businesses, particularly companies that are not IT/software companies or companies that do not run mission-critical applications on ColdFusion. The current Standard/Enterprise editions would still be targeted to business that run mission-critical applications.

My second suggestion is a bit more radical. It would be to take a more Microsoft-type approach. ColdFusion Server would be free (all editions) but Adobe would build and sell a world class IDE for ColdFusion. This would be designed to outpace all current free or inexpensive offerings available to ColdFusion developers (specifically CFEclipse, Dreamweaver and HomeSite+). In order for this to work, however, the IDE would have to be nothing short of life changing for ColdFusion developers. This is because the tools we currently have (at the minimum) reach the level of "good enough".

The potential advantage to this approach would be an apples-to-apples cost comparison with the Microsoft tools. No longer would the cost of the server come into consideration. At this point, management would consider the cost of Adobe tools against the cost of Microsoft tools. Management would never ask the question "Why would I buy ColdFusion when ASP/PHP is free?" because management would be purchasing the same type of product whether they go with Microsoft or Adobe.

I admit that these suggestions are radical, and in the end may not be practical, but I think Adobe needs to do something radical in order to take the ColdFusion platform to the next level. I look forward to the days of ColdFusion version 15 or 20 but Adobe may need to consider an unconventional approach to secure the future of ColdFusion.

Jason/Matt, it’s an interesting model, and you can argue that it works with Flex Builder. I was at MAX recently, and I’d say (unscientifically) 95% of flex developers used Flex Builder. So, it CAN work – it would need to be kickass, though, as Matt pointed out.

Many in the community call for a free version of CF, with scaled back functionality – personally, I think this will just dig CF into a hole. The FUDers will simply compare ASP/PHP to the free (scaled down) version of CF, and we’ll be spinning our wheels.

The personal concern I have is regarding the future of the enterprise version of CF, which is something our company uses. Now, as a package (JRUN, CF development platform, CF Monitor), CF Enterprise is one of the cheapest J2EE solutions out there. Adobe needs to get out of the “Challenger” category and into the “Competitor” category though. I’m concerned with the announcement that JRUN will no longer be developed – will I need to buy BEA or Web Sphere to support CF Enterprise in the future? If so, that would be a HUGE burden to shoulder.

I don’t know how MS gets away with its “Free” tag. I mean, Visual Studio Pro costs as much as CF Standard. How is that free? At least CF scales well – add another developer, costs stays the same, but not so with MS solutions.

Still, I’d impress in Adobe to push the pricing model for CF9 – get it to less than $1k, if possible, and show the IT world a move in the right direction. It’s a tough line to walk, though, better features cost money, and I’m glad we have all the features we have.

Cheers,

Davo

- Matt Rileys idea of a low budget CF with only a few features and a simple (as usual) installer especially for web space providers

- IDE again: I'm still using Homesite+ but Adobe should CF stronger bundle with eclipse for an easy quick start

- CFCONFERENCE: A CFML web specification - to have CF in the news (even BlueDragon, Railo and Smith are good for business)

- I personally don't like CFSCRIPT but it seems people are too familiar with messy code looking like php or asp. So give it to them.

- Don't forget: http://www.teratech.com/blog/index.cfm/2007/10/18/...

Jason,

I've said this several times now, but the two features I see making a huge impact for CF9 are thus:

Pluggable search engine... make it easy to remove Verity and use other engines, from Lucene to Bob's Phat Search Server... maybe even integrate the API with Google's search API? I know it can't be 100% open because you'll have to implement a wrapper for every engine added, but, perhaps in the spirit of event gateways you can expose a Java search API that could be extended by anyone to use a new search tool.

This is my big one... my legacy if you will:

Add Administrator extensions and tags in the vein of CFQUERY/DSNs to allow people to more easily integrate payment gateways into their applications. Provide a list of the most popular ones like you do with databases, then config options for auth keys, etc., and add <cfauthorize />, <cfpayment />, etc., to make using ColdFusion THE PLATFORM for doing ecommerce sites! w00t!

One feature that I would love is ability to set a flag not to automatically
create variables for URL and FORM scopes, so that I have to type URL
or FORM.
Same goes for query when looping.

Also, new framework for building pages.
for example the page would look like:
<CFPAGE EXTENDS="MyBasePage">
<CFFUNCTION Name="OnLoad">
<CFFUNCTION Name="OnInit">
<CFFUNCTION Name="OnAuthenticate">

<CFFUNCTION Name="DoSomething">
</CFFUNCTION>

<CFFUNCTION Name="OnRender">
<CFBUTTON OnServerClick="DoSOmething" Value="TEst">
</CFFUNCTION>
</CFPAGE>

Hi Jason,

Integration. I think further integration with other Adobe and non-Adobe products would make Coldfusion a much stronger product. Before you think 'more of the same', integration can fundamentally shift the role of Coldfusion - most importantly into the consciousness of key decision makers.

Most radically, consider including features from or bundling express versions of FMS and Livecycle. Most people has no idea how easy Flex combined with FMS makes 2-way videoconferencing. Nor do they see much of anything about Livecycle. Allow people to taste these products in Coldfusion. This increases the value of Coldfusion, but it is also an entry point into the dedicated server versions.

Regardless of whether you bundle the functionality or not, other Adobe server-side products should be ridiculously easy to use in CF. Its great that you included LCDS express in CF8, but it still isn't trivial to use. AMF, Flex Messaging, data push, Actionscript proxy etc are all terms that confusion the hell out of me. Simplifying the CFADMIN screens and including testing tools (eg. a testing flex client) would be make the whole technology more accessible.

Imagine writing a Flex 2 or Air application in MXML on a CF page. What does CF add to that equation? Maybe the CFMXML can have remote connections to CF or LCDS server cooked into the SWF. Maybe flex libraries (SWC files?) can be automatically generated by the server based on a data CFC?

Similarly for FMS, I can imagine that having defined a (bundled or remote) FMS server in CF admininistrator the user could immediately start working with high-level CF-style productivity code like <CFVIDEO> (inserts a flex movie player in the current page) or <CFWEBCAM> (inserts a flex application for 2-way only video conferencing).

I don't know a lot about LiveCycle, but since it is an Adobe technology, CF should be able to talk to it natively using powerful tags to recieve and update information. I think Adobe should consider including PDF form rendering in CF, since investing heavily in PDF forms probably increases the value of LiveCycle, rather than the opposite.

For AJAX integration, although I appreciate how far we are down the road of using YUI, I think CF should seriously consider forming a strategic partnership with Spry. What stands out about Spry as a library is that it ridiculously simple to do amazing things (like Coldfusion). From your perspective developing CF tags for spry is a simple activity too. Being an Adobe product you can influence its development, share some budget and maybe rest easy knowing that your entire javascript codebase is consistent, well managed, compatible with other libraries and has a stylistic appeal that sometimes gets lost in other brute-force CF solutions.

(And have you noticed the number of designers working in Spry who really need an easy back-end solution? )

I think deprecating CFSCRIPT and moving on to the latest flavour of AS is another radical idea that has merit. How many flex developers find themselves writing some form of backend code? I would guess a lot. Increase the incentives to choose CF as a backend server by using AS on the server side...

... Which implies the availability of a complete API exposing all the functionality exposed by tags. Recognise that CFScript (and AS if it comes) is generally used by power users. There is a tendency for CF to lose valuable developer mindshare when they "graduate" from Coldfusion. Enhancing their language tools and access to high-productivity libraries means they won't wake up one day and discover they work mostly with Java objects and may as well be working in a pure java shop.

A high level API would also allow CF developers to retain the CF productivity boost into their 'power-user' days, and I imagine it would help with the CF Team's development and debugging efforts as well.

Anyway, I have other comments RE integration and what it might do for Coldfusion, but I need to go to trot on. This is turning into an opus anyway :)

I've been out of the CF world for a couple years but two features I always wanted were:

- Url Rewriting built into the server. Rules could be set in the CF admin.

- Scripting syntax for writing CFCs ... actually, making a lot of the non-UI tags available to cfscript. I always found the tag-based syntax clunky for non-html files.

The killer server: Bring together Flash Media Interactive Server 3 and Coldfusion Server Enterprise!!

I seem to be late to the party but I support much of what has been said.

I've been doing CF development since version 2 on floppies but after using AS3 for the past few months with Flex the suggestions about replacing CFSCRIPT with AS3 sound great. Otherwise I'd vote for making CFSCRIPT a first class citizen with non-udf access to everything available in CF tags.

Either way I'd like to see more effort put into making the Flex/AIR - CF RemoteObject setup simpler and easier for beginners to access (the XML and cross domain hoops could be better documented and integrated into the CF administrator / Flex/Eclipse interface).

Obviously full featured LCDS functionality would be great to have included in CF. It would be a gift to developers (one install process gets you the whole local dev environment without Tomcat etc.) and would give CF server a built in advantage over PHP / Rails / .NET which would still need WebORB, etc. Including LCDS in CF server could make those services available from existing shared CF hosting providers at a much lower cost of entry, greatly increasing their exposure and use by the developer community. CF server should be leading the way in this area for Adobe.

If Adobe really wanted to boost Flex / AIR development they could merge CF and LCDS into one server product, make AS3 (or 4?) an available standard language across everything, and even reduce the server price tiers so that the PHP / Rails / Java / .NET alternatives would look much less attractive in comparison, especially to small shops / independent consultants.

I'm sure there are many financial and technical hurdles to this kind of thing happening but this is a wish list right ;).

First up I'd like to say that it's great that the CF team are so open to the cf community to have their say.

I agree that cffeed has to support gzip, and that a free 'Express' type IDE is a great idea to encourage new users to go for CF over other languages. CFEclipse is great, but for someone just starting out with writing code are they really going to download Eclipse, then get the CFEclipse extension + Aptana etc when they can have a simple download and install msi/exe?

Video support would be a killer feature, but why not also include audio as well?

One thing that I would love to have that hasn't been mentioned is a group attribute for the cfloop tag that works like cfoutput query="x" group="y". It would make my code so much neater!

How about multi language support a bit like the SetLocale. I not exactly sure what I mean but being able to build multi language apps would be a real selling point for CF.

CFC paths - can we do it from the web app root or be able to navigate up?

AMF support for CFC webservices.

Sorry, one more - you've got me thinking now! How about a good way to serve large files as cfcontent doesn't quite cut it.

Make all the cf tags that create output generate valid markup, perhaps in the format specified by some server setting (most don't generate valid xhtml currently)

Some more *big* ideas:

<CFAJAXPUSH /> - Push data to an Ajax client for use in bind expressions using an dimensionless flash object

<CFAJAXTIMER /> - Sets simple JS timeout

<CFAJAXPROXY cfc="mymanager" jsclassname="MyPrivateNamespace.mymanager" /> - Allow private object namespacing

[All Ajax Tags] - Automatic JS dependency inclusion with JS require() function (no cfajaximport required). Graceful degradation.

<CFAJAXIMPORT scriptsrc="/my/implementation/of/cf /js/interface" /> - All JS libraries loaded dynamically through CF bootstrap. Open CF JS API for customisation

<CFCACHE action="put" scope="application" value="#mycfcinstance#" key="myCfc" /> - General extension CFCACHE to handle everyday application, session, etc caching in a safe way without CFLOCK

<CFCOOKIE> - Extend to use flash shared objects
<CFIF hasJavascript()> - Self explanatory

<CFMESSAGING action="send" topic="" > - Include a reliable messaging agent with CFADMIN support for use with Flex-messaging, inter-CF, JMS, SMS, Ajax etcetera.

<CFSET application.taglibs = {} />

<CFFILE action="copy" retries="nn" compare="true" />
<CFFTP action="put" retries="nn" compare="true" />

- Ant-style file-comparison for CFFILE AND CFFTP. Option retry count for robustness

<CFANT /><CFANTTARGET><CFANTTASK name="" type="cf" />- Integrated ANT or other deployment/automation tool. Can call CF code

I'm all for Aaron's comments about OO. We do need more OO capabilities (yes, keep them optional.) I'm not one who ridicules ColdFusion, but some more OO capabilities (or options) would do me just fine.
I do think that the video and stuff can be added maybe in a future release. To be honest, you should consider doing an 8.5 where everything is just enhanced and wait a while before you do a 9 release – making the 9 release a real big deal. That's what I would do anyway.

Some more:
1. A general update and review of all tags leveraging third-party code to include
a) latest versions
b) more functionality from the third party

Eg. <CFCHART>

2. On the subject of CFCHART, the use of the charts to setup drilldown capability is hampered by the limitations to the URL parameters. If you could sneak an ID into a CFCHARTDATA tag for use for this purpose it would be great.

3. <CFMAP long="45.4" lat="54.6" zoom="8" googleid="XXX" /> (this would probably be cooler than it would be useful)

4. The ability to create PDF forms.. PLEASE! If the Acrobat team give you any grief, well, ignore them :)

5. I said it before but it still makes sense: Consolidate your server range (Livecycle,FMS, LCDS). Think of all the millions of dollars that could be in your budget!

6. NTLM support for CFHTTP, scheduled tasks, etc. Some of us live in M$ networks, afterall.

7. Support for some additional formats in CFPRINT: RTF, TXT, HTML, etc. (you could transparently PDF them first).

PS > Missed you at WebDU!
PPS > Where do we go to find out more information about the open process initiative?

I've got a great feature request... how about fix all of the little bugs and idiosyncrasies in ColdFusion?

For example, it drives me crazy that you can't conditionally disable a <cfform> control:

<cfset disabled = "false">
<cfinput type="text" value="can you edit me?" name="test" disabled="#false#" />

Since an HTML input control with a disabled tag with ANY value (even "false") is interpreted as disabled="true", cfform should not pass the disabled attribute through to the <input> tag when the value is false.

I'm sure there are lot of other little bugs and gotchas like this... but none come to mind at the moment. Going through the livedocs comments would probably reveal most of them.

i know i'm a bit late on this one, but how about the sql to show up when in debug mode and there is an error with the execution of the sql. i'd like to be able to turn debug on for the sql and know what it looked like in order to determine why it had the error.