How do you explain the benefits of the configuration depicted in the graphics of the SUM4JTA Technology page?
Think about having many testers automatically and without cognitive mistakes reporting their test results to a support engineer. Think about having no barrier in communication, instantaneous communication, and not a bit of testing experience being lost.
Now think about the ability to accumulate all of these experiences in one automatically correct analysis.
Now imagine that all the users of your application automatically become testers working for your organization without any effort on their part while they run your application in SUM Enabled mode.
What is "SUM Enabled mode", and "Black Box mode"?
By "Black Box mode" we mean the traditional mode of application execution. A SUM Enabled Machine allows you to switch your running applications between the two modes. When being executed in SUM Enabled Mode your application is automatically analyzed for the presence of Race conditions. When Race Conditions are detected they are automatically reported by the SUM User Interface running either on the local machine or on the Enterprise Edition - SUM Knowledge Station "mother ship" machine or on both.
What is the difference between the SUM Enabled Machine and SUM Knowledge Station installations of SUM4JTA?
One SUM Knowledge Station, that is an Enterprise version of SUM4JTA, can support many SUM Enabled machines.
Both versions can run SUM Enabled applications and see the results of SUM4JTA analysis. However the Enterprise Version - SUM Knowledge Station - has extended capabilities of being able to see and further integrate the results of analysis originally performed on different SUM Enabled Machines.
The SUM Enabled Machine version does not need to have GUI capabilities if it reports to SUM Knowledge Station machine (which does need GUI capabilities in order to see the aggregated knowledge).
The license for the SUM Enabled machine is less expensive. You can benefit by having number of SUM Enabled computers passing their knowledge to one "mother ship" computer - SUM Knowledge Station computer, which will process the collective information.
What is the basis for ROI calculations?
The effectiveness of SUM4JTA was studied on popular open source projects. The results show that the most stable versions of these products, which have been supported for 11 to 15 years, still host over 10 Race Conditions per 100,000 LOC.
A study conducted by the NIST reports the average time to fix one bug as 17.4 programming hours. http://www.nist.gov/director/planning/upload/report02-3.pdf
That translates to over two days of software engineering time for an average bug.
However Race Conditions are not your average bugs and are much more difficult to detect. They can live within your system for many years and manifest themselves only occasionally, which can be at worst possible time. Keeping all this in mind, and in order to be conservative, we keep the price of one bug to an ISV as $1,500. Estimating 100,000 LOC being supported by one Software Engineer, we gather that SUM4JTA can save you 20 * $1,500 per year per Software Engineer or $30,000 per year per Software Engineer in savings on improving software reliability.
Can SUM4JTA reproduce Race Conditions in order to analyze them? "No", and "Yes", and "We Don't Need To".
"No", because Race Conditions are not reproduce-able. We cannot force a method or a portion of a method to take the absolute same time to run as it did before. We cannot force all the involved pieces to align in absolutely the same way as they aligned previously.
"Yes", because once we have detected a Race Condition, SUM4JTA can emulate it in order to provide a visual explanation of this condition to the user.
"We Do Not Need to" because the analysis is already performed automatically at the time of Race Condition detection.
Are there any additional benefits from using SUM4JTA?
Yes. A specific piece of code does not have to have Race Conditions within it for SUM4JTA to provide benefits.
Can one locate all race conditions present in a code at once using SUM4JTA? (*)
No. You cannot do it with any tool. SUM4JTA is a dynamic code analyzer. It uses your application experience. Think of it as a net that catches all experienced Race Conditions and does not catch any that are not a Race Condition (it has 0% false positive rate). Think of it as of a tester that is always there and makes no mistakes.
A Static Code Analyzer would attempt to do that, but it would give you in best cases a 20+% rate of False Positive results. Additionally it will miss some of the Race Conditions actually present in your code. False positives unnecessarily rob you of your time while you try to locate things that are not present. Additionally, they undermine your trust in a tool.
To get the value out of SUM4JTA you need to have it running all the time. You will get even more value if it continues running in production. Get your users to work for you, perpetually improving the reliability of your applications!
Do you provide a trial period?
The free trial period is 7 days and it demonstrate SUM User Interface in action on the example of popular open source project by NASA "WorldWin". (see your own house as it is seen by Earth-orbiting satellites)
The trial license provides for 4 "bullets" (4 resolutions of races, combined) in the Java applications you would want to try.
For the purchased licenses we provide Monthly and Yearly subscriptions and and free updates and a Guarantee of not reporting any False Positives results, backed up by a free additional license period .
The Guarantee works by requesting from you the proof of allegedly False Positive result. This can be done by, controlled by you, Email Notification mechanism built in the SUM4JTA.
The same mechanism can be used by you to securely communicate any specific finding of SUM4JTA to your tech support person inside or outside of your LAN.
Additional, built-in automatic communication to your technical support computers (SUM Knowledge Stations - enterprise version of SUM4JTA) see this configuration in the Purchase link) is also included in the trial product.
How would I cancel a subscription if I decide to?
Your Google Checkout receipt has a link "Cancel your order". That will stop your subscription from being automatically charged for the next month or for the next year.
Are floating licenses being provided?
Yes. We will provide you with a floating license server when you opt for multiple licenses and a discount for a corporate license.
What is the overhead?
The overhead is in many cases low enough to allow using SUM4JTA in production.
The exact percentage varies from application to application. The application algorithm is what controls the overhead of any dynamic tool. In many cases the results we see show under 5% overhead. Put a bit more memory in your machine and it will run faster than before your applications became SUM Enabled. But this order of overhead is not guaranteed for every application. You would need to see for yourself.
The option to interactively switch the mode of execution between SUM Enabled and Black Box modes allows you the capabilities of using SUM4JTA in all phases where your application runs: development, test, and production.
How does one get additional licenses?
You can purchase additional licenses at any time via Google Checkout by selecting the "Purchase" button from the main menu. Volume discounts are subject to the specific volume at the time of purchase.
Do you provide Technical Support?
Technical support and free upgrades are included in the price of the license.
How does the referral program work?
You can earn a bonus for referring new clients to TSI after they make their purchase.
The amount of the bonus, which will be in the form of the check that will be mailed to you, will be calculated based on the volume of the purchase made by your referrals.
Your referrals would need to provide your email address on the form automatically generated during the purchase of the license.
