[Important!] Solution to fix "MSVCP120.dll is missing" when starting jaMME + poll

[ent]

Hi, jedis.

 

Recently 2 people had a problem to start latest (1.6) jaMME. It did not start due to error that said "MSVCP120.dll is missing from your computer" (or something). It happens because I started building jaMME with Visual Studio 2013 (it used to be 2010) which uses some new dependencies.

To fix that problem you just need to install Visual C++ Redistributable Packages for Visual Studio 2013 from here: http://www.microsoft.com/en-us/download/details.aspx?id=40784. Select vcredist_x86.exe only!

After installation you should probably reboot your computer.

 

Also to avoid this error on the next releases, please take a part in the poll.

 

May the Force be with you.

 

 

[Futuza]

I wish Microsoft would release a "install all C++ redistributable packages for every year" tool...would make my life simpler.  

[Apprentice]

I wish Microsoft would release a "install all C++ redistributable packages for every year" tool...would make my life simpler.

... or made it part of the automatic updates cycles or something . . .

[Futuza]

... or made it part of the automatic updates cycles or something . . .

Well not exactly since it isn't required, or wanted by a lot of users.  I mean it'd be nice I suppose if you install one redistributables windows update always checks to see if new redistributables are available as well, or something.

[Stoiss]

have a site here for that http://en.vessoft.com/software/windows/download/microsoftvc

[ent]

Stats says 65 users viewed this topic but only 3 took part in the poll?

If most of jaMME users don't care then I will make my own decision about it. I cannot be based on what only 3 people voted for, it's too small sample.

[Omicron]

I don't remember ever getting this error, or if I did, it was too long ago for me to remember. I don't really know much about this topic, but if including the dlls' (second option) fixes it, then I'd probably say that so people don't need to install the first thing. Or: Could have a .txt included which says how to fix and where to download it

[Stoiss]

i think you should ínclude it so they know its needed to be installed if its not already done on there Computer

even by a link to download it from or put the needed installer files in a folder

[Apprentice]

Stats says 65 users viewed this topic but only 3 took part in the poll?

Welcome to the world of ... bots!

[Futuza]

Stats says 65 users viewed this topic but only 3 took part in the poll?

If most of jaMME users don't care then I will make my own decision about it. I cannot be based on what only 3 people voted for, it's too small sample.

I didn't vote because I don't own JK2, also I'm not sure what the right answer is.

[eezstreet]

I didn't vote because I don't own JK2, also I'm not sure what the right answer is.

What does owning JK2 have to do with anything?

[Futuza]

What does owning JK2 have to do with anything?

I was under the impression jamme was for jk2, maybe I need to recheck the file download section.

[ent]

I was under the impression jamme was for jk2, maybe I need to recheck the file download section.

You probably saw joMME. There are two I made: jaMME for JA and joMME for JO.

[Futuza]

You probably saw joMME. There are two I made: jaMME for JA and joMME for JO.

Oh then, dang what have I been doing all this time?

[OmarFW]

I have this error as well, and I think adding the missing DLLs sounds like a much simpler solution on the client end.

 

Thanks for the temporary fix advice. I would love to get this working and make some machinima again.

[ent]

It's a bunch of DLLs and I dunno names of rest so...
I will just build with VS 2010 toolset in future releases and none should get this error again (maybe Windows XP users only but they follow the instructions in this topic).

 

Why not 2013? Because with VS 2013 Windows XP users won't be able to run the application at all. And there is also problem with randomizer and numeric data types.

[Xycaleth]

You still need to provide the DLLs when compiling with VS2010. It might be that people already have the redistributable for VC++ 2010 installed so they don't get the error, but same rules apply anyway.

[ent]

You still need to provide the DLLs when compiling with VS2010. It might be that people already have the redistributable for VC++ 2010 installed so they don't get the error, but same rules apply anyway.

You are right. I am being based on the fact that most of users have VC++ 2010 Redistributable installed. And according to this it's not worth to provide those DLLs (I still don't know which exactly) for those few users (0.01%?! ;p).

[Futuza]

It's a bunch of DLLs and I dunno names of rest so...

I will just build with VS 2010 toolset in future releases and none should get this error again (maybe Windows XP users only but they follow the instructions in this topic).

 

Why not 2013? Because with VS 2013 Windows XP users won't be able to run the application at all. And there is also problem with randomizer and numeric data types.

You can make it still work for XP users, there's an option you just have to set on vs2013.

[ent]

You can make it still work for XP users, there's an option you just have to set on vs2013.

As I said above, VC++ 2013 Redistributable has problems with randomizer and numeric data types. And it's a big problem for me.

[Futuza]

As I said above, VC++ 2013 Redistributable has problems with randomizer and numeric data types. And it's a big problem for me.

Really?  Are you using the std library or quake3 defined types/functions?

[ent]

Really?  Are you using the std library or quake3 defined types/functions?

Both. But the one that comes with std lib has problems.

[Futuza]

So like plain old int / double / float has issues?

 

and good old rand()?

[ent]

When float contains rather big value that does not even let to have number after the point (for example, 123456780. or 123456789.) and we convert it to int ((int)floatVar or automatical conversion) then it does wrong conversion. For example 123456780. -> 123456780 and 123456789. -> 123456780 (different float values to the same int one). (The values are just examples so I doubt those ones are actually problematic.)

 

And randomizer does this if you send big seed (>100000000 I think): 

 

 

If to set toolset v100 (VS 2010) we get this:

 

 

 

So the conclusion: VS 2013 with default toolset (v120) cannot handle big numbers correctly.

[Futuza]

Floats are inaccurate due to their nature, using a double might work out better or a long double.  As far as the randomizer I'm not sure what's going on there, possible it only supports limited ranges properly, or you might be seeding it incorrectly.