Is there any Virtualization software which still supports windows 98?

332

Solution 1

I have VMs (in Virtual Box) with windows 95 and windows 98, and they works in 800x600x16bit... just have to play a little with drivers...

There's a tutorial in the Virtual Box Forum. I don't know if it's valid (I knew how to do it) but, give it a try.

Solution 2

Windows Virtual PC officially supports only XP+ as a guest OS. However, you can install Win98 just fine (remember to set the available RAM to 64 MB for Setup to succeed). In order to get:

  • Better graphics
  • Integrated mouse functionality
  • Desktop resizing

you need to install the virtual machine additions, but of course these are no longer available for an unsupported guest OS. A simple workaround is to extract the VMAdditions.iso from something like Virtual Server 2005 SP1 or better still VirtualPC 2007 SP1. While I don't know about DirectX 9 and Direct3D support, the graphics situation should definitely improve over the current base VGA config you're stuck in.

More details can be found in the article Installing Windows 98 on Windows Virtual PC by Microsoft's Ben Armstrong, who was apparently in charge of VPC and now is the Hyper-V Program Manager (be sure to read the comments as well).

Share:
332

Related videos on Youtube

hhoomn
Author by

hhoomn

Updated on September 18, 2022

Comments

  • hhoomn
    hhoomn over 1 year

    I'm trying to implement a seq2seq encoder-decoder that from an English sentence reconstructs the same English sentence. The decoder works fine (trained with teacher forcing it reconstructs valid sentences), but the encoder returns the same hidden encoding for any sentence.

    I'm using pytorch and I liked to know if is this a common problem. What is the probable cause of this?

    hidden_size = 300
    output_size = vocabularySize
    
    class EncoderLSTM(nn.Module):
    def __init__(self,emb_dim=vocabularySize):
        super(EncoderLSTM,self).__init__()
        self.lstm = nn.LSTM(emb_dim,hidden_size)
    
    def forward(self,X,h):
        X=X.view(1,1,-1)
        #print(X.shape,h[0].shape,h[1].shape)
        st,h2=self.lstm(X,h)
        #print(st.shape)
        return h2
    
    def initHidden(self):
        result = (Variable(torch.zeros(1, 1,hidden_size)),Variable(torch.zeros(1, 1, hidden_size)))
        return result
    
    ...
    
    hidden = encoder.initHidden()
    for i in range(target_length):
        encoder_input=target_variable[0][target_length-1-i]
        hidden = encoder.forward(encoder_input,hidden)
    
    
    
    ...
    
    for i in range(1,target_length):  
    
    
        pred,hidden = decoder.forward(decoder_input,hidden)
    
        target = target_variable[0][i].unsqueeze(0)
        values,ids = target.data.topk(1)
        target_id = Variable(torch.from_numpy(np.array([ids[0][0]])).long())
    
        loss+= criterion(pred[0],target_id)
    
    ...
    
    loss/=target_length
    encoder_optimizer.zero_grad()
    decoder_optimizer.zero_grad()
    loss.backward()
    encoder_optimizer.step()
    decoder_optimizer.step()
    
    • terdon
      terdon over 11 years
      Have you installed the drivers for your graphics card?
  • Admin
    Admin over 11 years
    worht a +1 but still no direct3D/Full DX9 support :(