Re: [UPDATE] net/i2pd 2.58.0 -> 2.59.0 + Improvements

On Mon, Feb 16, 2026 at 12:09:28AM +0000, openbsd@systemfailure.net wrote:
> On Sunday, February 15th, 2026 at 4:18 PM, Lucas Gabriel Vuotto <lucas@sexy.is> wrote:
>
> > Hi David, systemfailure,
> >
> > I'm busy, but I'll be taking a look at this update by next weekend.
> >
> > On Sun, Feb 15, 2026 at 11:08:14AM +0000, openbsd@systemfailure.net wrote:
> > > I guess this should be documented in the README, at least until the old port is not supported anymore, in other words until 7.8 is phased out?
> >
> > I think that the usual way of dealing with this is /faq/current.html .
> >
> > > I hope we'll be able to update the port soon, since this new release of i2pd partially mitigates an ongoing DDoS/Sybil attack disrupting the I2P network. FYI this new attack may be the result of a botnet operator's incompetence: https://krebsonsecurity.com/2026/02/kimwolf-botnet-swamps-anonymity-network-i2p/ .
> >
> > The easiest way to deal with this is decoupling the port changes from
> > the version update.
> >
> > For only the version update, it has my OK in case any other of the devs
> > can commit it. I won't be able to do so until Wednesday or Thursday.
> >
> > Lucas
> >
>
> I think you are absolutely right. Here is a patch updating net/i2pd to the latest release (2.59.0), without any other changes.
>
> 'make test' is failing:
>
> c++ -Wall -Wno-unused-parameter -Wextra -pedantic -O0 -g -std=c++17 -D_GLIBCXX_USE_NANOSLEEP=1 -DOPENSSL_SUPPRESS_DEPRECATED -pthread -Wl,--unresolved-symbols=ignore-in-object-files -I/usr/local/include -I../libi2pd -O2 -pipe -L/usr/local/lib -o test-http-merge_chunked test-http-merge_chunked.cpp ../libi2pd.a -lboost_system -lboost_program_options -lssl -lcrypto -lz -lpthread
> ld: error: unable to find library -lboost_system
> c++: error: linker command failed with exit code 1 (use -v to see invocation)
> gmake: *** [Makefile:36: test-http-merge_chunked] Error 1
>
> But it does not prevent the port from building and working fine, as far as I tested (on amd64).
>
> Best regards.

This is committed now, thanks! The issue you were running into for make
test is related to having differing versions of boost in your ports tree
and installed in the system. make test runs successfully once that is
fixed.

I'll be taking a look at the improvements during the week.

Lucas