Not sure I know much about it but if anyone might expect to make a game proper with FPSC and go to the extent of developing 50 quality full levels - 40 x 40 tiles - with a great deal of complex detail which is what one could expect from a real game - (and at the end of the day you cant tell people well you can only use stock media to make their game with.)
One needs to be able to test out such a full scale development with custom media otherwise theres no point in having that provision of 50 levels or the option to use custom media.....
Either you can make a of game of 50 levels and using included custom media or you cant and when selling or promoting a product that should be clear or just say well you may or may not be able to do so, or success will depend upon factors, a, b, c or whatever. So its a maybe or you can try which is neither Yes you can or No you cant. Whichever it needs to be clear to purchasers and users.
Point being unless someone actually creates 50 such levels and tests them extensively I dont really know how you can confirm this. No one has done so to date as far as I am aware.
Disregarding the 50 level thing even compiling a successful game of a much lower number of fully completed levels at this stage seems like its likely to be quite an undertaking to acheive if one considers the range of users and their systems which will vary quite widely.
I am not trying to be silly here but merely looking at some logic.
If one considers a game of say 10 full levels completed needing to be compiled...
We all know that FPSC levels can be extremely large in file size and that experience shows that historically and to date that FPSC is itself overall not very efficient at handing those files as levels for various reasons.
If one can create say 10 such levels and play those levels in test mode or as individual compiled levels successfully with
some efficiency in gameplay and without error in level or gameplay then that is the first obstacle which one needs to be proven to be possible accross a range of user systems. That needs stability and issues cleared up on the engine part and it needs the user to ensure that there are no errors in their own build or code.
Given that is the case then compile is a matter of stringing those levels together witout error in compile as a game as their credibility individually would have been proven without which theres no point in compiling together.
Given now that you have proven that the individual levels are credible in that way and without error one has to compile them as a game in the first instance which one can assume will take an efficient and error free compile system making adaequate management provision of an individuals systems memory allocation in doing so. If an individuals system runs out of memory during compile then it will fail so memory management would need to be somewhat well catered for and presumably dynamic in management to accomodate differing systems available memory resources. Otherwise I guess its off to the render farm for some at least to get your game compiled.
If you do successfully compile your game without error then in the second instance one has to consider the gameplay. A full game of 10 levels x 40 x 40 tiles if complex in content will take how long to load levels? and will it be playable? You wont know until its actually compiled together and try and play it.
Thinking 50 levels one would not like to hazzard a guess of what kind of effort it would take to firstly make a full and complete game with FPSC, leave alone compile it or play it, with FPSC as it currently stands.
One important aspect which needs serious consideration and attention in creation of game levels in editor is testing levels where compile run takes place in test mode. This will give an indication of how efficient game compile is going to be.
Currently in editor the test compile time efficiency is poor to say the least, both during the compile and on the return exit to editor which can in complex levels take a very long time and waste a vast amount of your development time - especially at the moment when one has to test much more than one should need to because you must test continuously to check that engine issues presently are or are not a problem. One spends more time compiling and testing and checking for errors or trying to rectify errors than actually building or developing.
Where did the idea of double compile in test mode where portals and so on are compiled twice come from? What a waste of users time that is. I understand that it relates to trying to ensure error free compile?
Lets hope that double compile is not a part of the new version when it comes out.
At the end of the day irrespective of any individual issues FPSC is inefficient overall and is so in memory management and handling so it seems to me. Improving that must be a difficult task. Its not the only indie engine to suffer and having said that actually it does quite well considering what users expect from it and what TGC have tried to provide to users with it.
Expectations and aspirations perhaps are high all round which is not a bad thing. If it were not the case in the idie world in general then nothing would drive development forward.
Long may it be so.
