Would like feedback from paying customers

[Olivercomputing]

Frequently (as I and many others have observed), a question about a problem about Isomorphic's software goes unanswered on the board - even if it is clearly a real problem and a real bug. It's a bit of a mystery to me why that should be the case, since software quality is of importance to any publisher, and even if an open source user hasn't paid for it, they've expended time in development and their discovery of problems is a benefit to Isomorphic. I deal with one open source project that is not even making much from their software (yet) but whose developers are extremely diligent and timely about addressing problems.

I would like to ask those who have paid for the software and support to comment on their satisfaction with Isomorphic's response time and general level of support, for their questions. I am pretty happy with SmartGWT but there are lingering bugs in too many cases for me to really satisified, and I would like to have an idea about whether they will be taken more seriously if I buy the software and support.

[Isomorphic]

Where are the bug reports you're referring to?

When bugs are filed that are real and well demonstrated and apply to the most recent code, very very often we fix them regardless of whether the user has support. In some cases this is not immediate (as it is for supported customers) or we will triage the bug because it's an edge case that is less important than better samples, docs, or new features that affect many more users.

If you think there are unaddressed bug reports, look them over carefully to see if there is a minimal, runnable, standalone test case that has been tried against the latest version, and all other required information has been provided (browser, GWT, SmartGWT version, logs and diagnostics as applicable).

Otherwise, they are likely to be ignored. There are many false bug reports, and it's better for the community and product that we improve the product rather than chasing dubious reports.

[Olivercomputing]

Very often a bug can be identified which manifests itself within a large, working system, and it can be difficult to try to isolate it to a small "test case". That does *not* mean the bug doesn't exist. The notion that it should be ignored unless you have a neatly prepackaged instance of a problem is very questionable.

To take one recent instance that I mentioned here (without response): registering a right-click handler to replace (and suppress) right mouse click in a window was not working with all browsers. Serious problem for me, and still unresolved.

I have other issues I've encountered but the routine lack of response does not particularly encourage me to spend the time posting them here.

[Isomorphic]

Likewise, just because it's happening in your application does not mean that the bug *does* exist. It's just as likely to be a usage error or other non-framework problem.

We have had too many experiences chasing and chasing a report only to discover that a user has some third-party JavaScript library involved, posted false or stale information, was actually not on the version claimed, etc.

These wild goose chases do not help the product. If you had the data we have, you would not consider this questionable at all.

Regarding this particular report, are you referring to this?

If so, the most fundamental problem is you haven't mentioned anything about the content in the HTMLPane, which could be an IFrame, or some random HTML with it's own event handlers for all we know. And you also omitted versions for everything, despite the forums prompting you for it every time you post (in multiple ways).

It's pretty stunning to see a critique of support based on this.

[Olivercomputing]

The latest GPL version of SmartGWT, 2.4.

Assuming the problem is in the end users' application is convenient but not indicative of a serious interest in finding real problems. Frankly, even though I am impressed with your software, it is manifestly not bug free (I have a number of problems that I haven't listed which I have tested in a number of ways.) I mean, I pointed out to you guys recently that your own search settings in the forums excluded search strings that were "too long" to find various method names, which you had not previously identified. Pretty simple stuff - did you want me to debug something there too? Or was it sufficient to just point out the problem?

Software is pretty competitive and I have observed - here and in others' observations in outside venues - the dismissive attitude of Isomorphic about problems. I can say that Instantiations (recently acquired by Google) identifies SmartGWT as being substantially more problematic than Ext GWT (Sencha) in integration with the WindowBuilder product.

So a suggestion: less being "stunned" and more being interested in debugging your products. Lots of software companies, and indeed open source projects, have a much better attitude. I wouldn't care so much if there wasn't such a significant value with your software overall.

[Olivercomputing]

And incidentally, the loaded HTML does not contain IFrames, or any Javascript. Which I would readily have stated if the question were asked in that thread.

I would note that my original question was addressed to Isomorphic's customers in order to get *their* feedback of *their* paid support experiences, and I hope to see some.

[Isomorphic]

We love to find bugs and fix them. We are very aggressive about this. You can found countless instances of overnight fixes to bugs that were reported by people without support. You will find lots of mentions that SmartGWT is surprisingly low on bugs.

Your problem report was ignored, not because we like to ignore bugs, but because we have a voracious appetite for bugs and your post really doesn't look like it's going to yield one.

When we see the combination of a user who couldn't be bothered to post basic version information, didn't correct other fundamental issues with the report even when prompted, seems to believe he should get special attention (implying it's going to be pulling teeth to get all the required information),
and there are another ~100 possibles waiting to be looked at, not to mention confirmed bugs to fix, then what we're going to do is move onto those other reports and ignore your post.

Because we want to actually fix bugs.

All you need to do to get attention is follow the guidelines linked from right next to where you type in each post.

Finally, note that Instantiations is a very, very special case related to our product being a GWT wrapper - it's easier to integrate with ExtGWT not because we have more bugs (we have far less) but because it's a more typical GWT-based framework (and of course, has far fewer features). Google has indicated plans to let Isomorphic take over the code related to SmartGWT. Then, we'll be able to wipe out it's bugs, too.

[Isomorphic]

Crossed posts. If you think you've got a bug with the HTMLPane, just go post in the original thread and show a test case that includes your settings and some minimal HTML content that reproduces a problem. This will take far less time than arguing about whether the original, severely incomplete post should have been answered.

[Isomorphic]

Oh, also, try enabling the EventHandler log category and watching what you get when you click. The likely explanation here is that something is swallowing the event without it ever reaching SmartGWT. If so, we'd need to understand what HTML content can end up doing this, or whether it's a browser plugin you have installed, etc.

[Olivercomputing]

I'll be sure to include the basic information in any future posts.

My point with this thread was to ask for user experiences with *paid* support. In e.g.: http://forums.smartclient.com/showthread.php?t=11713, Isomorphic replied:

----
Either purchase commercial support or understand that you are relying on fellow community members for individual support, not Isomorphic.

It's fairly obvious [?] which posters have support and they receive quick and complete responses, which they often compliment. Paid support receives very good reviews.

If you are hoping Isomorphic is going to respond to questions when you do not have support, are not using functionality in the commercial editions, and are not contributing in any way, that is very unlikely. We can better serve the community as a whole by improving the product and documentation than by answering your individual questions.
----

[?] mine. Similar sentiments are expressed multiple times across the forum. Note that the above does *not* state that bug reports *will* be taken seriously even if tons of information is given. It gets across one essential idea: if you haven't paid for support we aren't going to take you nearly as seriously, period.

I'm not saying that's wrong. Just don't imply otherwise. Given the product value I am potentially willing to pay for a Pro package with support but I want to see *customers* willing to give feedback about their experiences. I hope you don't have a problem with that. And I hope to see some responses along those lines.

[Isomorphic]

Of course we don't have a problem with you requesting that customers provide their experiences. Don't worry, we haven't done anything to *prevent* customers posting in this thread :) You just haven't got any interested yet.

But no, your summary of Support's charter (about being taken seriously) is definitely not correct.

What's stated in the FAQ and was summarized in this thread:

When bugs are filed that are real and well demonstrated and apply to the most recent code, very very often we fix them regardless of whether the user has support. In some cases this is not immediate (as it is for supported customers) or we will triage the bug because it's an edge case that is less important than better samples, docs, or new features that affect many more users.

.. is more accurate. To sum up the difference, the priority is:

1) commitments to customers
2) our assessment of what will best improve the product for everyone

If you're not a customer you can still be taken very, very seriously under point #2. Hence all the immediate, free fixes.

[Olivercomputing]

Well, ok. I consider the direct quote from Isomorphic's reply on the given thread (and some others) to be somewhat contradictory to your summary here, but I will take it to be the definitive statement of your position.

And I do hope that somebody is sufficiently enthused about your support to want to acknowledge it here.

[Isomorphic]

The two statements are in harmony. The difference is bug reports (what's been discussed here) vs how-to questions (what the other statement is about - if you review SDorogin's posts, they are almost all big, open-ended how-to questions, many covered in docs, samples or prior forums discussions, others with too little context to answer).

Also note the word "relying". Put together solid bug reports and you'll likely pick up a few free fixes. But you don't have a guarantee, so don't tell your boss or customer you have a guarantee unless you've bought support.

As far as enthused customers, there's quotes on the website and a trip through the forums will find a whole lot of thank-yous or wow-that-was-fast remarks. People are busy, and unlikely to trip across this thread and have time to tailor a response to your specific question, so we'd recommend you go hunting instead.