retitle 1117436 thunderbird: Requires too much memory to build severity 1117436 normal tags 1117436 - ftbfs close 1117436 1:140.4.0esr-1 thanks
Hello. Something happened in the above version which made the package buildable again in my environment, without changing anything on my side, so I'm going to consider the bug fixed in this version. I'd like to comment on your previous email anyway: > I appreciate archive wide rebuilds as it's shows not only FTBFS in its real > meaning (really source related build issues), but here on this bug report I > believe the FTBFS is a false positive as all related architectures did build > the latest upload to unstable a day before successful as Christoph has > mentioned. So I tend to close this report with next new ESR version of > Thunderbird in 1-2 weeks. I fully agree that this was not a "real" FTBFS bug and I'm ok with not considering it as such, so I'm doing the above metadata changes to reflect that. > that's to me a waste of time as it's a general trend to have an requirement > of min 16GB+ RAM to get beast packages like Thunderbird build, or at least > I've no intention to dive into the deep internals of the Mozilla build > setup. Upstream is unlikely to accept any patches we would come up. I think the term "beast package" is not well defined, because it can mean several different things. I believe many people would probably call "beast package" to any package which has a very big source code, or one that takes a lot of time to build. But that's orthogonal to requiring a lot of memory. I think it all depends on how the program is structured, i.e. whether or not it has some *.c or *.cpp files which are costly to build for the compiler. I wonder if you would be willing to forward this bug upstream in case it happens again. The idea of having to subscribe to each bug tracking system myself is not very appealing, so it would really help for my QA work. Or at least give me a hint on how I should do that myself. You say that they are unlikely to accept this as a bug, but I think we will never know until we actually try. In fact, given that I see a slight improvement in the latest version, it would not be impossible that they actually did something to reduce memory consumption. Thanks.
