Open Source & Free  

Compare Thru Property Cross

Compare Thru Property Cross

Header Image

We just updated the QT and
Xamarin comparisons to Codename One that we did a while back.
Those comparisons now include a property cross comparison section.
Property Cross is really valuable for comparing the differences between various cross platform development
solutions and Codename One has the smallest implementation (in terms of lines of code) by far!

Our implementation is also quite superior in some other ways providing features such as infinite scrolling instead
of simple paging etc.

We think that just by looking at the Codename One project and comparing it to the equivalent cross platform
tools Codename One’s advantages become pretty clear.

3 Comments

  • Chidiebere Okwudire says:

    My two cents: I personally think LOC comparisons are marginally important and somewhat misleading. The user does not see how many LOCs an app has but does see the UI and experience good or bad performance. I think such CPT comparisons should focus more on user experience and development costs (which includes the costs of the development tools, licenses, etc.) instead of on stuff that probably make developers somewhat happier but do not contribute that much to the end goal of apps that are attractive, robust and performant.

    By the way, the CN1 implementation is still missing from the property cross website – a lost (but redeemable) opportunity for some free PR if you ask me! Maybe time to give it another shot?

  • Shai Almog says:

    All of those are good points. As I mentioned LoC isn’t a great measure but when it’s 3x or higher for most other solutions it does indicate something. Personally I think our solution is also the most readable but I have such a heavy cognitive bias I can’t really evaluate these things.
    Performance and usability of the end result application is something I can’t objectively discuss as a fact. So I’d rather leave it to the developers to read thru the code and see if this makes sense then try to compile each of the solutions.

    Development costs are really hard to evaluate, costs for licenses etc. are things that shift and with the exception of QT all the solutions we evaluated are free/open source.

    I did try again recently, tried contacting the guy who did the last commit to their repo etc. both thru twitter, his email account etc. There is an issue HE opened asking for a Codename One version. I submitted a pull request and commented a few times… Nothing.
    This is quite frustrating.

  • Chidiebere Okwudire says:

    Thanks for your response. It surprises me that there are not much objective comparisons in the dimensions that arguably matter most, some of which I’ve mentioned above (user experience, performance, costs). I know it’s not easy but it’s definitely possible albeit in a somewhat “academic” fashion.

    PropertyCross is one of the best initiatives I’ve found in this regard though it only scratches the surface and as you’ve mentioned, the maintainers are not as responsive as one would like. Apparently, they find it more important to change the layout of their homepage instead of incorporating the CN1 comparison (and other pending pull requests?)!

    Talking about costs, while it’s hard to evaluate, it’s worth mentioning some facts. For instance, in one of your previous comparisons (with Xamarin if I’m not mistaken), you highlighted that CN1 support a web app version. That’s definitely a killer feature but what is glaringly missing is a mention of the fact that it is an enterprise-only feature (read $399 per month). Now, that fact makes a H-U-G-E difference in the cost-benefit equation.

    As you rightly mentioned, most of the solutions are open-source at the core but often have paid extras which make a huge difference. It must be possible to compare these quite objectively – free against free, paid against paid (grouped by feature set not subscription model), etc. I’m not saying you have to do that; all I’m trying to point out is that while you’re obviously positively biased towards CN1, it’s still possible to do more rigorous comparisons 😉

Leave a Reply