Slashdot Mirror
Web 2.0 Mashups Almost Ready For Enterprise
Dion Hinchcliffe, in a blog post over at ZDNet, talks about the increasing business value of 'Mashup' projects. Some of these, he believes, may soon or already be ready for use in an enterprise environment. He demonstrates one of these upcoming projects, showing off IBM's QEDWiki in a Flash demonstration. The software allows users to create their own mashups from canned widgets, turning data into simple applications with fairly straightforward functionality. From the article: "The motivations for mashups are quite different inside of organizations, where application backlogs and demand for more software that will improve collaboration and productivity are often rampant. If this state of affairs is true, far from having too much software, most enterprises don't have enough to satisfy demand, despite the prevalence of mountains of existing enterprise systems, many of which are underutilized. The arguments for letting users self-service themselves with end-user application tools and getting IT out of the critical path for the backlog of simpler applications are extensive." How important do you think 'self-made' software will be in the future?
I'm not going to ax for extra credit or anything but, I wrote mashups in the 80's. FTA:
I would claim this specific notion (mashups) not only originated from the enterprise and trickled into internet consciousness, enterprise "mashups" existed many years ago. I know, I wrote them. It was (or at least we called it) surround technology.
We took vital pieces of different applications and wrote wrappers which allowed users with very simple interfaces to access more data more accurately more quickly. One example was a service order writing routine for small business that routinely took over 30 minutes... using our "mashup", we accessed the necessary enterprise applications and melded into a single app presentation and shortened the 30 minute process to less than 5.
I could go on, there were at least three other major applications we wrote (small team of 2, sometimes 3), that were "mashups". The advent of browser technology simply gave us another presentation tool, the notion and mechanics of mashing was still there.
I've played with Google "mashups", and Amazon "mashups", they're really nothing new.
There was a (don't know if they're still there) a Strategic Computing Consortium based in Boston, Ma, and they were huge advocates of surround technology and not only taught techniques and reasons for approaching solutions this way (I won't go into it -- it was a six-week class). And they provided and sold tools and consulting for putting these new applications together... the CEO (I believe) was John Donovan, author of a few college texts on OSes, and another major contributor was Stewart Madnick, one of the original authors of CMS (IBM's Conversational Monitoring System).
I'm won't claim they were the "founders" of mashups, but what they espoused and taught was mashup technology, and they were teaching it in 1986 (that's when I attended the consortium). The more things change, the more they stay the same.
(Also, as an aside, the article implies this new magic allows for "easy" creation of new applications. This is hardly so. All the care and due diligence of putting an application are still required. The effort can still be significant... There is certainly time saved if a team leverages existing critical applications but to toss this out as magical and easy for any end user community to leverage is probably glib and misleading.)
Deleted
Several attempts have been made at this in the past. In many companies, there is one or two Excel uber-users (or even, gasp, actual developers), who are able to understand parts of the enterprise's accounting, ERP, CRM, etc. databases and make tools...er, workbooks, that facilitate some of the necessary analysis or other operational needs of the department they work in, even if it means "enter data from here in the app form X to this cell here", i.e., manual screen scraping.
The company I am contracting at is trying to do something like this with an enterprise rules engine by TIBCO. Others provide various kinds of APIs that hide the gory details of the database or application interface, whether it is SAS, SAP ABAPs, etc.
It might work in a general sense, but it will still involve developers at some point to bridge the gap between functional experts (i.e., accountants) and the application, in order to fit the application to the business, and not the other way around.
Is a mashed-up trailer for Star Trek Enterprise, a la "Ten Things I Hate About Commandments" or "Must Love Jaws".. That might make that crappy show entertaining.
http://www.youtube.com/watch?v=t4UIJTt-vdU
A witty saying proves you are wittier than the next guy.
Yep, we see that every few years. Strangely enough, it coincides with the latest new "paradigm".
I blame Star Trek. People want technology to be magically easy to configure and re-purpose. But it isn't. Computers don't "think" like people do and it takes a lot of work for a person to think the way a computer does.
Being pretty much accurate for most of the data most of the time is what you get when the untrained person attempts it.
Please, let this horrid buzzword die. Right now. All we have to do is convince the Slashdot editors to stop injecting it into articles. Last time I turned around, "mashup" meant some kind of Frankenstein DJ set. Now it means the same thing as connecting software packages together by end users? WTF.
*** DRINK MORE COFFEE ***
I like mash-up potatoes ;-) They're good with butter and salt.
But where in the world is this Web 2.0 taking place in the rest of the world? Perhaps in small shops, but certainly not throughout any enterprise or even close to being ingrained as a solution in any major IT firm that I know of. In fact, the IT industry has gone the opposite road from "intuitive and creative" and has wrapped itself around the "software axle".... making policy based on software instead of choosing software that is intuitive to policy. Irregardless of why this has happened, it has happened, and I dont see any corporate CIO or mid level manager coming to accept the wonder widget that bob from accounting replacing the application they foolishly purchased from slick willy the software sales guy.
20th century Marxism is not progress...
we're playing buzzword bingo here, right?
"I blame Star Trek."
I don't.
"People want technology to be magically easy to configure and re-purpose. But it isn't."
Let's ignore the faction that benefits from the status-quo.
"Computers don't "think" like people do and it takes a lot of work for a person to think the way a computer does."
It's easier to change computers than it is to change people.
"Being pretty much accurate for most of the data most of the time is what you get when the untrained person attempts it."
They're usually better domain experts than the turf-protecting programmers.
nt at all
My turnips listen for the soft cry of your love
pretty imporant, and i cant help but think that it is being vastly overlooked in spite of the promising work ive seen so far at multiple different conferences. there was one really promising demonstration given by cyberdyne systems last month, although i cannot quite recall the name at the moment. AirWeb or something of the sort. anyway they seemed to think that it was going to have broad implications in the years to come.
I remember the whole justification for the expense of web services and the xml-ification of everything was the promise of doing just this. No more Com, OLE, COM+ etc where everyone must have MS Office installed to get things done. Software is available in any browser, any device any where at any time.
You make a public stateless web service. RSS feeds of content. Internet enabled APIs. Mashups are the logical result of being able to pull in data from anywhere, control it and use XSLT etc to change the layout.
I don't see this as any novel or amazing concept. It's the final result of a decade of hard work finally getting noticed now that the web standards have stabalized and the internet bubble popped and removed most of the random money chasers from the pool of talent.
"How important do you think 'self-made' software will be in the future?"
About as important as it's been up until now. The vast majority of people who have to use computers are totally incapable of using them beyond launching applications with them. In an era where people have to be trained on specific keypresses and mouse clicks for specific applications, there is exactly zero chance of these people developing any kind of software, using any kind of environment, to solve any kind of business problem.
There is no way this will even remotely impact the role of professional software developers. None. Zero. Zilch.
TAGS: washups
A mashup is a music term, meaning a song made up from the parts of other songs.
... and then they built the supercollider.
The next great thing from the The Department of Redundancy Department!
Now enterprise software developers too can take a hodgepodge of technologies and combine them into a poorly designed application of Frankensteinian proportions.
Wait, how is this different than before?
its 'mashups' have come up.
Read radical news here
The idea is pretty cool, but my problem with all the IBM tools it that it looks very bad :)
You can instantly see that the icons (cool green colors) are made a by a graphics designer, but the rest of the website looks like it could be made by any of the millions myspace users. Horrific!
It uses 6 or 7 shades of blue that don't match...
Err.. yes, I'll stop nagging like a woman now.
My blog: http://www.redcode.nl
"Mashup" is most possibly the worst word that has ever come out of the technology sector as a buzzword.
First it sounds like "an amalgamation of multiple different components into one" but when I look at all of the sites/services that are referred to as "mashups" none of them fit this description. QEDWiki is a wiki, it doesn't appear to be "a wiki with a calendar attached" and it certainly doesn't appear to be built from 10 different components or easily integrated.
the article mentions zillow, which is an online real estate directory.... It has no "mashy-ness" about it at all.
Anyway, its a stupid word that doesn't mean anything
"Web 2.0 Mashups Almost Ready For Enterprise"
:-(
What a disgusting, vapid headline
That is all.
No, let's look at cars. The heavy equipment that usually takes a new driver a few months to "master".
And yet tens of thousands of people are KILLED while operating these every year. And I'm not even talking about crippling injuries, non-crippling injuries or property damage.
The fact is that even when their LIFE IS AT RISK people fail to handle the technology they have correctly. Even after being trained on it.
So why would they spend more time and effort learning how to program effectively?
"The fact is that even when their LIFE IS AT RISK people fail to handle the technology they have correctly. Even after being trained on it."
Let's pretend that you didn't make the assumption that these failures were caused by carelessness. As opposed to other factors.
"So why would they spend more time and effort learning how to program effectively?"
Similiar comments were made in the early days of the personal computer by the previous "priesthood". Looks like those who benefitted from change, themselves resist change.
Has anyone actually watched the flash demo? Sadly, I have wasted a good ten minutes of my life that I will now never get back watching it. In doing so - I took notes on two terms that I found interesting:
Situational Application: Come on people, WHAT fucking application on the planet is NOT situational? I've NEVER used an application that was NOT situational - be it a game (entertainment), word processor (solving a business need), or anything else for that matter.
My other favorite:
Data driven application: As opposed to what?!? A bullshit driven application? Ah yes, that is officially MY new buzzword: Bullshit driven application. You heard it here first folks....
Most programs are driven by user input. For instance, an accounting program does not create an invoice because of the contents of a data package; it creates an invoice because a human being carries out a sequence of input events called "raising an invoice." (For web based businesses, the process is initiated by a buyer clicking on an event button.)
Not that I am saying this distinction is being made here. But "Data driven application" can be a useful term - especially if (like me) you develop data driven applications and have to explain to people that the competition's application is purely event driven.
Pining for the fjords
Is it not some kind of create your own application?
No sig for now.
>> "People want technology to be magically easy to configure and re-purpose. But it isn't."
:P
> Let's ignore the faction that benefits from the status-quo.
While you can make good technology that works well, ultimately it does rely on a user who knows what they're doing. There are plenty of untrained users who can't figure out anything beyond the wall plug. You can make up new meanings for words like "faction" all you want, but it won't change the fact that I know these people and I answer their illogical questions constantly. Alas, it's what I do all day, for the most part.
>> "Computers don't "think" like people do and it takes a lot of work for a person to think the way a computer does."
> It's easier to change computers than it is to change people.
And who's going to change the computers? Oh, right, the same people who don't want to or can't adapt. Actually, it's easier to train people than to try to code up a DWIM instruction. Been there, done that, plenty of times. You bend over backwards trying to make the computer explain everything to them, they get some message because they tried to do something that makes NO sense whatsoever, and then they ask you what the message means. The root cause here is muddled thinking at least as often as program error.
>> "Being pretty much accurate for most of the data most of the time is what you get when the untrained person attempts it."
> They're usually better domain experts than the turf-protecting programmers.
Doesn't help much. I have a huge mass of legacy code written by "domain experts" who were not programmers. There's no error checking, they apparently don't know how to allocate memory (malloc() is apparently unknown to them) and do nice things line making an array of 1,000,000 of a certain struct. They have huge masses of effectively dead code, require people to enumerate all possible options in a configuration file (then proceed to silently overwrite all of the data in the file with hard-coded, recalculated values). But maybe that's not so bad, because they slurp in the file with scanf() and don't bother to check ANY part of anything for stupid things like error codes. And no, it's sure as hell not some speed-critical loop, nor is the application carefully isolated by other things which DO check for errors.
Fact of the matter is that you have to be both. If you're not a programmer, you'll create a brittle, WTF of a program. If you're not an expert on the problem you're trying to solve, well, you probably won't finish your application and no one will use it because they prefer the old piece of crap they've been using the whole time.
But anyhow, the proper role is to have the domain experts write the specs and do the testing, while the programmers write the actual code. Otherwise, I'll probably end up submitting your code to The Daily WTF if I'm unfortunate enough to come across it
In a side note, given the nature of his site, I'm honestly not surprised the Daily WTF's maintainer is the fanboy of Windows that he is...
"The arguments for letting users self-service themselves with end-user application tools and getting IT out of the critical path for the backlog of simpler applications are extensive.""
And the arguments AGAINST it are very serious and extensive as well.
Look at all the crap Excel spreadsheet "systems" and badly-designed Access database "applications" that exist in every company.
This stuff is under no one's control except one or two employees. It is sometimes used for mission-critical decisions. And the reliability and accuracy of the application is not controlled by anybody, let alone the issue of whether proper backups, data vetting and security are being done in such "end user developed" applications.
This has proven to be bad news in the past for many companies, and will be proven so again, I suspect.
Applications that aren't that important for a business, such as applications that merely improve the productivity of an employee's personal use of their computer, aren't that bad. But applications that are important for the CORRECT performance of the employee's JOB should be developed by people that have some clue about the issues that surround application development (assuming such people exist in your IT department - which isn't always the case, unfortunately.)
Richard Steven Hack - This sig is TOO GODDAMN SHORT TO DO ANYTHING USEFUL WITH! MORONS!
Deprived of the command line, lacking the ability to simply pipe information between processes, the web generation has rediscovered the simple fact that many tasks are best solved by letting the end user combine small, best-of-breed tools.
"Mind you, this is not a bad thing, just another variation of an age-old trend ... ask any carpenter or plumber out there and you will get the point."
Mind you I should point out that the "Do-it-Yourself" business is a several billion dollar business, which wouldn't exist if those "superusers" had to get the permission of every carpenter or plumber every time they wanted to fix a toilet, wire a lamp, or hang a shelf.* It's about time for the same changes to come to the programming field. Otherwise you all are no better than the AMA and their guild.
*A similiar argument applies to any endevour that bypasses the "establishment". e.g. music, publishing, printing 3D.
"Look at all the crap Excel spreadsheet "systems" and badly-designed Access database "applications" that exist in every company."
Some people shouldn't throw stones.
There is a reason that companies have an IT department. There is a reason that they hire computer experts. The simple fact of the matter is that every Tom Dick and Harry doesn't have the necessary skill set to develop and MAINTAIN their own applications. Companies need to ensure that they have data integrity and ensure that everyone is working with the same dataset. When you start giving users control over something as mission critical as data applications you are looking for a headache. At the end of the day, you are going to have a bunch of pissed off users and a bunch of pissed off IT guys. The users are going to be pissed because their applications break. The IT guys are going to be pissed because they are expected to support applications that they didn't even develop in the first place.
If you need to give users access to data, give them a copy of Crystal Reports and send them off to class to learn how to use it. I haven't come across a single situation where a non-technical person needed data out of any system that couldn't be presented to them with Crystal Reports.
How is this any different than OpenDoc, OLE or some of the NextStep OO "kits"?
This sig intentionally left justified.
SIGSEGV caught, terminating
wait... not that kind of sig.
I'm about to be bored watching this all happen again. Sorry if this sounds like flamebait, but it really isn't intended to be. Here's how I read this -- we've got very few new ideas. We've got RSS, aggregators, social networks and content and we've got endless permuations as always. In an effort to generate excitement (and VC) we're inventing new buzzwords and "paradigm shifts" as fast as we can say "Flooz". There's nothing "new" about aggregating various technologies or sources into a single offering. (Hello? This is Slashdot people!) If it makes anyone feel "2.0" to call combined technologies and different content-sources wrapped up in userfriendly packages with community features "mashups", then make yourselves delusionally happy. I'll wait for the "mashup" explanation right after you finish defining Web 2.0.
I think I'll recombine the same technologies next year and call it a "smoothie".
------ The best brain training is now totally free : )
I don't see many commenting on whether or they think their users will actually create anything with anything at hand. Where I work, we have management analysts that need to creatively use spreadsheets, web apps, or anything, in order to do their job well. Well, they don't. If it's not served on a silver platter, they are not going to even try to use it. To be fair, the amount of work that they are expected to deliver is now much, much higher than, say, only 3 years ago. I generally think that it's only the perception of a bad economy that keeps them from fleeing the premises. I don't see anything that requires non-developers to take action ever taking off when people have to take time off to learn. The fear factor is big. Cragen
How important do you think 'self-made' software will be in the future?
Not very.
I find laziness to be an excellent motivator.
And the argument against it can be summed up in one (admittedly hyphenated) word:
Sarbanes-Oxley
"And what it means for regular users when there's no budget or appetite for cavalry-calling is that they suffer at the hands of recklessly clueless Excel fiends who think they're God's gift to project management."
As opposed to IT that suffers from delusions of BOFH.
"Those same manager folks would, even if there was plenty of budget for software development, staunchly insist that their million-column spreadsheets are working fine, while forcing underlings to spend half (literally, half, I counted) their time maintaining and updating disparate copies of the same data in different sheets used by different managers."
Maybe you all should create better tools, instead of complaining that the present tools are being used in ways you guys don't approve of.
"Make it easy enough for an idiot to do, and every idiot will."
Make it hard enough that an expert is required, and you'll have a job for life.
Funny how that angle is never protested as much as your angle. Proving once again this isn't about making computing better, but protecting one's turf.
> How important do you think 'self-made' software will be in the future?
...
... then ask why you have IT "churn" because your staff do nothing but "firefight")
Only as important as the poor IT d00ds who have to support the whole mess once the fly-by-night "finance manager" has left the company
(for "finance manager" substitute any role in the company