Is it a review? Is it a feature? Is it an editorial? It's all of the above, as AAS's Krisse goes off on a well-timed rant against low quality Java titles that damage the whole smartphone ecosystem. And, if you happen to be the author of a certain lottery number generator, then prepare to be shamed.
Read on in the full article.
It has been said once that if you create a language every idiot can program with, only idiots will. We have seen this happen with Visual BASIC which became a telltale sign of badly written software. We are seeing it with Java now. Not that there are no good J2ME games, but statistics are strongly against them.
So why is this product being sold via this site?
Comment the app in the shop!!!!
Now there is only one comment, very positive, probably made by autor!!! - 5 star (sic!)
Why would anyone purchase blindly such a simple application when there are probably dozens of freeware equivalent applications out there, or at least shareware ones you can try before you buy, is beyond my understanding...
That being said, I do agree with you that the complete lack of quality control in retail sites is worrying. This was only a $10 gadget, ther would be worse situations...
Take, for instance, MobiForm sold on SymbianGear (http://www.symbiangear.com/product.html?pid=19415,19416,99383). That's a 139 Euros (178 US$) Rapid Application Development (RAD) tool, and it is listed as E61 compatible. I was deeply interested by this one as it could have been a good way to develop on-board using the E61 alone, even if there were many limitations. So after digging as many informations about this tool as I could, from sources like the developer's website, I shelled out the money, only to discover that the E61 is, in fact, not supported, and no S60 3rd Edition is currently supported due to a change of Java Virtual Machine for that version of S60. There's no way it's going to support E61 in the future, too, as was confirmed by the author later. Yet it's still listed in the E61-compatible applications.
The author is a helpful and friendly person, and I was refunded very quickly when it became evident that this tool was not going to fit my needs and the the E61 compatibility was misleading, but that was already months ago, and it's still listed as E61 compatible on SymbianGear. Since the author had been nice with me, I didn't write to SymbianGear about this, but maybe I should have?
When I complained about the lack of software, I had this good vs. junk apps issue in mind as well, and years ago I was complaining to see the retail sites flooded with ringtones and themes which artificially increase the number of "available apps" yet provide little to no added value, but it seems that every platform is equally plagued by this problem.
Yet I think that the security model chosen by Symbian/Nokia might be a hindrance to the emergence of low or medium value, but freeware applications on the market, specifically designed for S60 3rd Edition, instead of generic Java applets that work on every device but are largely under-optimised to say the least.
My 2 cents only, though, I know that many here do not share my opinion... 😉
As a post-scriptum, does anyone know about a clock program for S60 3rd Edition that would display the time (with seconds) as big as possible on the screen? Like BigClock on PalmOS devices... A quick search on the net from the E61 itself didn't bring much results appart from Handy Clock which didn't match the requirements. And no, thanks, I don't want a Java application, only a true Symbian one... 😉
The software in the AAS shop comes through from Handango although I *think* Rafe can override items he doesn't want. Rafe?
Steve
Steve's right the software comes (partially via Handango) and a certain amount of stuff gets added automatically. However I can remove stuff as I see fit. I hadn't noticed this particular one, but I'll make a note to zap it. I currently working on a way to show the more popular / more downloaded / better quality applications first. Long term project as its not as simple as pure numbers.
I don't think the problem is the ease of writing software. Making application creation easy is a good thing, it means developers can concentrate on good ideas and good design rather than fixing bugs in code. As you might guess from the article's title, the problem IMHO is all down to the apparently total absence of Quality Control.
In the early 1980s it was incredibly easy to write commercial software, there were hundreds of cases of bedroom software houses going on to become famous developers, and that was great. But it's important to remember that bedroom coders almost always went through commercial publishers. For every bedroom coder who did have a big hit, there were hundreds more rejected by the publisher, so the good stuff was able to make it to the shop shelves and the bad stuff wasn't.
There was also another quality filter on the bedroom coders, the distributors of software would also inspect games before deciding how many copies to order from the publisher.
If commerical Symbian developers went through a commercial publisher, that would immediately weed out a lot of the rubbish, and if the legal download sites also applied quality control then that would weed out even more of the bad software. It seems like neither of those things is happening right now.
Rejection because of quality control can actually be a good thing for developers, because it'll give them an idea of what works and what doesn't, and may spur them into improving whatever areas were lacking in their rejected software.
"When I complained about the lack of software, I had this good vs. junk apps issue in mind as well, and years ago I was complaining to see the retail sites flooded with ringtones and themes which artificially increase the number of "available apps" yet provide little to no added value, but it seems that every platform is equally plagued by this problem."
It is a potential problem for any software download platform, and as you say all smartphones, not just Symbian, suffer from it. That's why I used the phrase "smartphone software" in the article rather than "Symbian software".
This is the heart of what I was trying to say in the article, that releasing junk just makes things worse. It's far better to have a limited range of good, useful applications than thousands of bits of utter rubbish.
It's interesting to note the approach taken by the download platforms present on all of the Next Generation game consoles: they're NOT going to flood it with thousands of files all at once, but release a limited number files at a constant rate so that the quality isn't overwhelmed by pap. I just hope that the console platform holders don't get tempted into opening the flood gates to rubbish, because there's bound to be the temptation to mistakenly think that just doubling the number of titles available will double their profits.
Making application creation easy is a good thing, it means developers can concentrate on good ideas and good design rather than fixing bugs in code. As you might guess from the article's title, the problem IMHO is all down to the apparently total absence of Quality Control.
In a traditional environment, where writing software actually requires expertise, a company is forced to hire experts and have a QA team. Of course, it does not come cheap, thus you have to insure that your are not wasting funds on a badly designed product. Also, the price of a copy of your software is going to be high enough to justify the investment.
In the new, Java-based, mobile environment, you can outsource writing (and designing!) your software to a bunch of really cheap Indians or Chinese. They will, of course, have *some* kind of QA internally, but you won't have any idea about it. Nor will you have to worry about hiring and managing programmers: the outsourcers will do it all for you, although the people they hire will most likely be chosen based on required funds rather than quality. You just need to approve their preliminary designs and the end products. They are real cheap, too, when you compare them with the local workforce just about anywhere.
Because it is so cheap to develop applications now, you can have a lot of them and sell them for peanuts. The obvious problem becomes the quality deterioration. Yes, those nameless guys *can* write Java code and they *can* design some resemblance of a game, but their products are always going to feel half-baked.
If you go back in history, you will see about the same situation with Atari 2600 games, although the reasons were different that time. You may also find that the abundance of crappy A2600 games basically killed the console industry. Consoles did not rise again until the NES, which was of course a totally different system hardware-wise.
Interesting observation about bedroom coders and their resellers being their quality department. One reason the retailers could do this was because they took a large slice of the money, 30% or something like that. The developer was lucky if he got 10 % of the retail price.
Also, the only thing a developer could to to sell their software was by going through the retailer. You made money from your software if the retailr stacked yoour cardbord boxes.
Software was therefore expensive, a couple of hunders of dollars.
(I am skipping the shareware scene, which came later, but before the internet).
With the Internet came electronic dictribution, which made it possible for software developers to sell their software directly to consumers, or through content aggregators. Because it is much cheaper to sell using the internet, prices have come down a lot, and software authors can still make more money compared to selling cardbord boxes through a retailer with shops made of brick and mortar.
As a side effect, the quality control that was there in the cardbox box world, and that was paid for by the 90% the developer did not get, has gone.
Because it is much easier to create distributable software (you don't need to produce the CDROM (or floppies 😉), the quick start leaflet, the manual, the cardbord box, the physical distribution of the cardbord boxes) more people are now able to produce distributable software. Changes are, average quality will go down because of this.
A content aggregator has now two problems:
1) distributable apps are a lot cheaper and there is therefore less money to test. But because apps are now cheaper, the risc of buying a badd app has also gone down. Instead of wasting 100 euro, you waste a tenner.
2) there is much more software to test. There is now more software for sale for Symbian devices than there ever was stacked in cardboard boxes in shops.
Finally, most content aggregators I know have a 30 days money back guarantuee. Users now have to be their own quality testing department, and instead of paying more money, they need to spend more time reaserching and evaluating it.
Or, pay a lot more money and let other people test the software for them.
Sander van der Wal
www.mBrainSoftware.com
Really, people, my poor programmer's heart weeps: We have a quality problem with Smartphone applications, and that is because today's programming languages are too easy to work with?
Do I understand the right way to get more quality Symbian software is make building programs harder again? Make Java illegal, perhaps?
Literally hundreds of millions of completely untalented people could write books as they were given in an earlier post as an example transposed into the realm of literature, with only a few pages of mostly nonsense. So where are the enormous quality problems with bad books?
Do I understand the right way to get more quality Symbian software is make building programs harder again? Make Java illegal, perhaps?
Literally hundreds of millions of completely untalented people could write books as they were given in an earlier post as an example transposed into the realm of literature, with only a few pages of mostly nonsense. So where are the enormous quality problems with bad books?
PS: Please, do not conclude that I propose to outlaw graphomania.
To my mind, the problem with (Symbian) software is not caused by the ease of programming, though this may play a role. The main problem is the fact that much Symbian software which would have been free on other platforms (SIBO, EPOC, Palm) is *not* free.
There is nothing particularly wrong with a poorly written piece of software if a. it does not do any damage to the device on which it is running and b. it is free.
What you have now appears to be an abundance of software which does not meet these criteria. Java just exacerbates the problem by being non-device specific, lowering the quality (if there was any to start with) even further.
Another take on the situation: the main problem with native Symbian code is that it's too difficult (incompatibility between versions, horrible API:s etc.). This means that people won't write software for fun, so there's no freeware. Hence simple programs, that would be produced as freeware for other platforms, have to be developed commercially.
As for the ease of buiding native Symbian apps: Yes, it is often definitely un-funny to program for Symbian, and the plethora of OS versions and UI flavors really does not help.
Yes, "Lucky Lotto" would be much less of a problem as freeware. Everybody should be free to program whatever he or she likes and publish it as freeware, and complaining about crappy freeware should be kept to a minimum - you just get what you pay for, sometimes.
"Lucky Lotto" would also be much less of a problem as time-limited trialware, with a chance to try it before you buy. But here even I become nearly sympathetic with the author of the program: Ever tried to implement a time limitation for a Java or a Symbian program that is not achingly trivial to circumvent? I mean, Symbian, Nokia and SE give us hundreds of classes and methods to use in our programs, but say, where is proper support for try-berfore-you-buy shareware?
I think we're straying off the point here, the issue isn't one of programming or implementation of demos but one of quality control.
As far as utterly useless applications like Lucky Lotto go, they should never even be allowed into the sales channel. Something like LL is completely devoid of usefulness or entertainment and doesn't represent potential value to anyone, so it can do nothing but damage to the sales channel and the platform.
The fact that apps like this are allowed in implies there's absolutely no quality control at all, beyond simply checking that the SIS file installs. This situation has to change.
"In a traditional environment, where writing software actually requires expertise, a company is forced to hire experts and have a QA team."
I just can't see why any company with any level of budget wouldn't be capable of vastly improving upon something like Lucky Lotto. There's just no need to outsource something like this, it's a flipping number generator for goodness sake! 😊
As I pointed out in my article several times, the amount of work required to make Lucky Lotto into a useful application is absolutely tiny. Allowing users to set the range and number of figures generated would have allowed me to actually use Lucky Lotto to generate my lottery numbers (I couldn't because this option wasn't there). This feature could be added by pretty much any programmer with any level of experience, probably in less than an hour. As Steve pointed out in his notes, he wrote a lottery app for Symbian in 10 minutes.
Other extra features I suggested such as an odds calculator would also take very very small amounts of effort, it would just involve inserting the number of lottery balls and the range of the numbers on the balls into a simple mathematical formula. This could be done in ten minutes too.
The fact that they didn't do any of this but still published the app and demanded $10 for it speaks volumes about their intentions for the product: they didn't even try to make anything useful, they just published the first thing that came into their head with no thought about the needs of potential customers.
"I am not sure why you would do something this drastic. But a software distributor like Handango that tests every piece of software submitted and does not accept ones deemed unplayable would be a good point to start. In fact, I would suggest employing a community of people who would test submitted programs in exchange for getting free copies."
Definitely. That's exactly the kind of thing that should be going on, and it's what goes on (or is supposed to go on) when console platform holders decide whether to allow a console game's release. It would cut through all the problems with development and outsourcing mentioned earlier in this comments thread.
The suggestion about allowing people free copies in exchange for testing is a good one too. To avoid personal taste coming into it they could take a vote from several random testers on whether to accept a piece of software. There could even be a feedback system where the testers say why they've rejected something, which would help the developers fix any problems (it'd be a compulsory beta testing stage, effectively).
The fact that apps like this are allowed in implies there's absolutely no quality control at all, beyond simply checking that the SIS file installs. This situation has to change.
That's exactly the kind of thing that should be going on, and it's what goes on (or is supposed to go on) when console platform holders decide whether to allow a console game's release. It would cut through all the problems with development and outsourcing mentioned earlier in this comments thread.
The suggestion about allowing people free copies in exchange for testing is a good one too. To avoid personal taste coming into it they could take a vote from several random testers on whether to accept a piece of software. There could even be a feedback system where the testers say why they've rejected something, which would help the developers fix any problems (it'd be a compulsory beta testing stage, effectively).
On the most fundamental level, for me there is no principal difference between Symbian software and Windows software. Because of this I think that we can gain insight about where the Symbian software scene may head in the next few years by looking at the Windows software scene today.
I see all kinds of phenomena in that realm, just by looking at the big download sites, e.g. the presence of a lot of good freeware, a lot of bad freeware, a lot of good software for money, a lot of bad software for money.
I see also a couple of mechanisms that allow the users to find good software. Free software and shareware can be tried first, of course. Software from a handful of well-known companies can be trusted more or less by default. And then there are user feedbacks, user ratings, and already the download count can be a important indicator of the quality of a program.
What I really do *not* see is quality control.
What do you think - if a big Windows software download site has several tens of thousands of files, how much crap is among these? Maybe 50% or more. That's what I would call a quality control breakdown. Does it matter? Well, obviously not that much.
That's why I personally am convinced that we can dream about all kinds of schemes of imposed quality control in the world of Symbian applications as much as we like, but they just won't happen.
Maybe they could happen if the number of Symbian applications would stay low for the foreseable future, not much higher than today, because then in theory somebody would have time to look at them all in some quality control process. But who wants the number of Symbian applications to stay low just to make quality control feasible?
I would say: Trust the market. Things like "Lucky Lotto" hurt today, and such things will probably stay with us, but they won't matter that much anymore, because the people will know how to evade them and find the good software that they need.