TOn Thu, 13 Feb 2025 20:29:12 -0700, Arthur Naseef
<artnas...@apache.org> wrote:


Hello Art,

I freely apologize for my sardonic and curmudgeonly style.  I come
from the rough and tumble of a bygone age - I'm well into my fourth
decade on the internet, and I'll age myself further by revealing that
I've been on Medicare for several years now... - when people tended to
be much more direct and much less politic than today.  I promise to
spend more time and be sure to make an extra style-check pass on
future posts. :-)

| The vast majority of open source coding is done with company money 
| behind it; for example, companies building new tools with the
| intent to later be able to sell support and training. [...]

I don't doubt this.  It explains a lot in the way of directions that
OSS projects tend to take (as opposed to directions that they tend not
to.)  It's always a matter of incentives (much more so than the "itch
to code".)

| On to the specific comments.

I think you misread them, but I won't argue them point by point, as
the fault is mine - obviously I can't predict how they could have been
received.  I take your points in the spirit that they were offered.
 
| [...] I haven't seen anyone on this project ever reject any contributions 
| based on them not being Java.  Of course, we do need to consider 
| maintainability and who is available in the community, so non-Java 
| contributions into a community of mostly Java folks will have a natural 
| challenge to overcome.

I wasn't suggesting that.  Rather, putting "chase" in quotes was to
make a point about disincentives in the ASF ecosystem.  Back in the
second half of the oughties (when I claim ASF was taken over by the
Java folks), C++ itself was in the pre-C++11 wilderness.  I think
quite a few simply gave up on the language and switched to Java.
Perhaps cloning Java code in C++ offered opportunities then for
constructive participation, but over time that source of talent dried
up, I think.  The state of C++ participation today is, I feel, a long
term consequence. C++ emulation of Java code is a turn-off in general.

| I personally see an interest to have people with different skill sets 
| on board. [...] If I were working with a client using activemq-cpp 
| and having problems that needed fixing, I would offer to help with 
| it.  On the clock, of course.

| [...] we really could use some solid C++ skills on activemq-cpp, or 
| a replacement.

Right now, maintenance need only amount to making the code base
passable to C++17 compilers.  Some extra tweaks to proof the code for
C++20 (e.g. a global SAR of 'throw()' with 'noexcept', as that alias
in C++17 has gone away in C++20) may not come amiss. It's a small ask
overall.

By contrast, replacement, or serious changes in the internals (e.g.
nuking decaf abd the APR library), is a huge ask.  As Justin points
out, that may not be worth the candle to anyone other than those who
would do it for love.

I've been working on making the code passable in C++17.  The large
number of warnings I find disconcerting - and just a couple hours ago,
I discovered that a particular warning actually hides a serious bug
that has been there all along.  The "good" news is that it's in the
implementation of decaf::lang::Exception, so it doesn't come up unless
an exception is actually thrown.  Fixing it will need some thought,
though.


Arjun 

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@activemq.apache.org
For additional commands, e-mail: dev-h...@activemq.apache.org
For further information, visit: https://activemq.apache.org/contact


Reply via email to