Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
CppCat, an Ambitious C++ Code Analyzer from Tula (viva64.com)
82 points by AndreyKarpov on Feb 14, 2014 | hide | past | favorite | 25 comments


Developers are a notoriously bad group to sell software to, and this interviews only confirms that once again. The authors feel they have to apologize for the price €5250 for a team of 9 developers. 9 office chairs cost more than that! Then the interviewer follows up with "such a high price, are there really any customers?" Geez.

With specialized products like these you need to find your exact target audience. If your audience is small then you need to deliver tons and tons of value.

Back of the envelope, you have to reach 1000 new customers a year at €5000 each in order to pay for development, an office, and all other overhead. Difficult, but not impossible. There still are tons of C++ shops out there.


Very true. And one of the challenges with static analyzers in particular is that most programmers think "oh this isn't really so hard, I could hack something like this up myself over the weekend". That said, the value certainly is there, and some companies realize it; back in the 1990s, I was architect for C/C++ static analysis tool PREfix, and our minimum sale price was $25K for a team of 10 people.


"oh this isn't really so hard, I could hack something like this up myself over the weekend" - Yeah! Bravo.


The authors feel they have to apologize for the price €5250 for a team of 9 developers. 9 office chairs cost more than that! Then the interviewer follows up with "such a high price, are there really any customers?" Geez.

I think you are largely right, but my knowledge of economics is weak and I'm confused by what happens when I invert your point and extrapolate:

1) Their market is more or less limited to software development companies willing to pay €500 for an office chair.

2) Pictures of their new office space would indicate that they themselves spend far less than this on each chair: http://www.viva64.com/en/n/0095/

Does this indicate that there is something suboptimal about their pricing model. Or perhaps it's just unrealistic to assume that a company should be able to afford its own products? Or perhaps this is a measure of the degree to which the "global market" is only partially present?


"Their market is more or less limited to software development companies willing to pay €500 for an office chair."

This is actually a huge market. Every company that I've ever worked for in the past were in the 10,000 plus developers and spending $500 for office chairs wasn't that big of a deal. Their target audience is certainly enterprise or startups with decent funding.


"Developers are a notoriously bad group to sell software to"

I sort of knew this when I started my startup but I was kind of surprised by how difficult it can be. People will automatically make assumptions about what it can and can't do without even looking at it, because in their mind, what's possible is what they know. My copywriting for my website is certainly to be blamed but I also know this is a tough crowd to please.

It's also why I'm more focused right now on getting the installer and documents ready so it can be integrated into their workflow. Speaking as a developer, we are pretty methodical in our thinking and we like to believe we are above marketing BS, so the only way to really sell to us is to prove it. And even this can be difficult as you still have to make it obvious to them that they can't create an equivalent solution in their spare time or be easily copied.

I can't remember who said it on HN, but they mentioned "Iceberg Problem". People who think they can do something in a few weeks are usually so myopic that they can only see the tip of the iceberg. What they always fail to realize is the tip of the iceberg is just that, a tip. If you can make it more obvious to them that there is a huge piece of ice below the tip, you'll have a better chance of convincing them of your solutions value and difficulty to copy.


Yeah, in the other professions no one expects to get their work tools for free, just on ours.

It is quite easy to measure software costs, when one does the math how much it costs to keep a company running.

Not everything can be sold on consulting services alone.


One group that should be willing to pay for it are those writing high-reliability and safety-critical code, e.g. DoD. It appears they have less of a preference for Ada nowadays, instead using a "safer" subset of C++ and static analysis tools.


Defense, Aero/Space, Medical, Embedded developers. Basically everywhere where the price of a bug is very high. (Source: I was a developer for a company that also does static analysis for C++).

The problem with Ada is that it is pretty hard to find a lot of Ada developers.


The problem is that in the early days Ada compilers were more expensive than C and C++ static analysis tools. :(

Nowadays we have cheaper options, but the harm is done.

However, from what I can infer from last years of FOSDEM, at least in Europe, Ada use seems to be slowly rising.


This was a great interview about selling a very complex product to a very difficult customer base (programmers). Someday, I want to start a developer tools company so it was cool to hear about the challenges they faced in getting started. My key take points were:

1. Your tool must demonstrably solve a challenging problem right away. Programmers will ignore it if on first run it doesn't have a good result for them.

2. Marketing via finding bugs in FOSS software can help sell licenses.

3. Mindshare is difficult to build because your target is often big companies with big budgets, but the people you need to convince are everyday programmers. Hopefully, their new lower priced option will help them in this case.

Anyway, very cool bootstrapped business. It is always nice to read about someone who has been able to make a business out of solving hard problems. Many businesses seem to solve easy problems which people haven't noticed before (which is likely a better business strategy but can sometimes lead to boring work).


Thanks for the comment.


Love this: "Ours is a most typical software company: we create software products and sell them to customers. In this sense, we are not quite "in the trend": we don't design mobile applications with millions of installs, or possess a website with hundreds of thousands of unique visitors. We run a, so to say, classic business which seems to be pretty rare nowadays."


Woah!

I've been sitting on a need for MSVC static analyzer for years. I looked at PVS-Studio (prompted by that exact Carmack's plug mentioned in the interview), but the multi-$K price was a non-starter. Static analysis has always been a secondary need in our case, a "nice to have" thing, which is now very easy to justify with $250 price tag. Thanks, guys!

(edit) Ha, it is $250 per year. Interesting. Aside from an initial gut rejection, this is actually not a big deal. 10 years worth of a license is still $2500.


Renew the license at 80% of the initial price (i.e. for $200 per license). 10 years - $2050


Visual Studio 2013 has a static analyzer.

A "good enough" solution that comes free kills all possibility that a second company can make a living by selling an cheap solution.


> A "good enough" solution

Comparing static analysis in Visual Studio 2012 (Visual C++ 2012) and PVS-Studio - http://www.viva64.com/en/b/0151/

It's old comparison. Soon there will be an article about the comparison of the new PVS-Studio and Visual Studio 2013. PVS-Studio is much more better.


And how much does an upgrade to VS2013 goes for these days?


Any chance of Linux support? How does it compare with Coverity, Clang Analyzer, PREfast, Klocwork et al?


The answer can be found here: PVS-Studio and CppCat: An Interview with Andrey Karpov, the Project CTO and Developer - http://www.viva64.com/en/b/0231/


if they support linux, I'd license it without any thought. $250/year is pretty cheap...


Agreed. I'd love a command-line version for both Mac and Linux that I can use for embedded C work. $250/year for this would be a no-brainer.


P.S. Ask everyone who like CppCat to leave a review here: http://visualstudiogallery.msdn.microsoft.com/8edc3425-8e8e-...


I can't help but feeling like you need guys would need an American sales team to sell your stuff in America. Or is corporate America pretty okay dealing with "commies"/Russians these days? (I mean no disrespect)


Thanks for the new program. It is we were looking for....




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: