Tweet
Another leak on what happened with Meltemi:
If true, that sounds either cringingly incompetent or actively nefarious management.
Starting it with the people with no linux experience, telling them to build it from the ground up instead of based on meego (meltemi was to use wayland, etc), not basing on existing hardware... Seriously, mistakes that any fresh MBA grad would have avoided.
However, it was the project that was planned and executed in the worst possible way I know:
1. Use some chip that is still under development. The release of the chip by the partner was delayed time and time again, with a great effort from Nokia developers to fix various issues.
2. Redo the base layer, by a team with no or very limited Linux experience. Most of the Linux gurus in Nokia were located in Finland, but the base layer is done in Ulm. Though they are really good developers, lacking of experience was really bad.
3. The central part of the whole software system was just a research prototype from some research center. The idea was really good, but, hey, how could you ever decide to use such a prototype as the core of the software stack?
4. Changing of design all the time, and we, developers, used a long time to convince our designers to follow what N9 already provided.
5. Kill the project when it’s almost ready. Yes, the product is in a really good shape in May.
One more thing, after Meltemi axed, we all got the order to delete all the source code related to Meltemi, with some reason that nobody trusted. So, let’s all speculate the reason.
1. Use some chip that is still under development. The release of the chip by the partner was delayed time and time again, with a great effort from Nokia developers to fix various issues.
2. Redo the base layer, by a team with no or very limited Linux experience. Most of the Linux gurus in Nokia were located in Finland, but the base layer is done in Ulm. Though they are really good developers, lacking of experience was really bad.
3. The central part of the whole software system was just a research prototype from some research center. The idea was really good, but, hey, how could you ever decide to use such a prototype as the core of the software stack?
4. Changing of design all the time, and we, developers, used a long time to convince our designers to follow what N9 already provided.
5. Kill the project when it’s almost ready. Yes, the product is in a really good shape in May.
One more thing, after Meltemi axed, we all got the order to delete all the source code related to Meltemi, with some reason that nobody trusted. So, let’s all speculate the reason.
Starting it with the people with no linux experience, telling them to build it from the ground up instead of based on meego (meltemi was to use wayland, etc), not basing on existing hardware... Seriously, mistakes that any fresh MBA grad would have avoided.