You may mock MS, but they just can't seem to help making massive piles of cash year after year. You may not like their products, but you've got to love their business!
That's kind of my point. Typical of Slashdot for folk to have a pop at Windows when talking about avionics! Perhaps Slashdot should be renamed as "We hate M$".
.....that's right, idle speculation from Slashdot readers, sure to be right on the money and more likely to reveal the true cause than the trained, expert investigators!
There are no comments bashing Microsoft at all so far, let alone claiming they should retain VBA. Can you read? There are loads of comments bashing MS that pre-date grandparent.
Safeguard is to host MD5 on different server to the product and to script regular checks of MD5 and product. I'm guessing it's a trivial cron/scripting language of your choice job - no more than an hour or two, most of which is setting up systems and procedures for release.
Something like sourceforge should be offering this as part of their package (in fact they probably do).
Obviously you don't like Digg (me neither). But what you also hint that there are sites which offer a more lucid take on technology news? What would you recommend? Can't see this as an Ask Slashdot topic ever!
One poster has mentioned 3rd party re-packaging offerings but nearing the end of 2007 it's astonishing that Mozilla can't supply decent MSI packages of its flagship product. Or perhaps they don't really want to gain market share on corporate Windows desktops?
The other issue that is really hard is extensions. Realistically Firefox only comes alive when you get some decent extensions added. These would need to be deployable via MSI also before FF got serious corporate acceptance.
As I recall, Delphi 2006 introduced a new memory manager, FastMM, built not by Borland but as part of a community project.
I think my use of the word moron was a little over the top - I'm sure I learnt the lesson the same way, by experience.
However, I don't think you are right in saying that memory allocation in Windows is the problem. The Windows API in question is VirtualAlloc/VirtualFree. However, this is not designed for use directly by application code. It is meant for use by local application memory managers. On x86 memory is virtual and paged and the OS manages page tables etc. for each process. Each page has a minimum size (don't know what it is), but calling VirtualAlloc to ask for, say, 32 bytes of memory would be sheer lunacy.
Applications have different memory use patterns and so it only really makes sense to let applications tune their memory allocation strategies. Delphi's MM was designed for general purpose use by typical Delphi apps. They typically allocate and deallocate relatively small blocks of heap memory as objects are created and destroyed. If you have an app that doesn't perform well with the standard MM then you can plug in your own custom one.
On other OSs that use x86 I'm sure the same basic rules apply. For example I doubt that malloc on Linux x86 glibc will do virtual memory paging for each call. Instead it will do the paging for large blocks and then use a sub-allocator just as the Delphi MM does. I would also imagine, but I don't know anything about it, that Linux offers memory reservation and committing services analagous to Windows VirtualAlloc/VirtualFree.
It's not Windows doing the memory allocation, it's the Delphi MM. The Delphi MM isn't bad, it's better in more recent versions of Delphi.
Only a moron would allocate and deallocate the same sized buffer on every mouse move triggered paint. But still you choose to blame Windows rather than looking in the mirror for the culprit!
You may mock MS, but they just can't seem to help making massive piles of cash year after year. You may not like their products, but you've got to love their business!
And bad software takes 10 years to un-write too!
That's kind of my point. Typical of Slashdot for folk to have a pop at Windows when talking about avionics! Perhaps Slashdot should be renamed as "We hate M$".
.....that's right, idle speculation from Slashdot readers, sure to be right on the money and more likely to reveal the true cause than the trained, expert investigators!
Would Linux be better at this?
....because VBA for Office on Windows is here to stay: http://blogs.msdn.com/excel/archive/2008/01/16/clarification-on-vba-support.aspx.
But that's not an interesting story is it since we can't bash MS anymore. Anyway, why let the truth get in the way of a good story!!
I'm picking up a little vibe here that Slashdot editors and readers don't like Vista.
Has anyone else noticed or am I just imagining it....
I don't think it means that it took two seconds to estimate copy progress. I think it means that the estimate is now accurate to about two seconds.
But that's pure guess work because the text isn't precise.
But it's quite common to read here that XP is Vista's biggest threat. So why can't this be so in open source world?
Anyone who pays $6,000 for such an item (even with the dollar being so worthless) gets no sympathy from me!
This happened 3 days ago: http://www.youtube.com/watch?v=S0tMhlhnga0
Ouch!
Is track racing really safer than road racing? Evidence?
Then it wouldn't be Slashdot.
Safeguard is to host MD5 on different server to the product and to script regular checks of MD5 and product. I'm guessing it's a trivial cron/scripting language of your choice job - no more than an hour or two, most of which is setting up systems and procedures for release.
Something like sourceforge should be offering this as part of their package (in fact they probably do).
I simply cannot believe this. Does anyone have a link to anything which gives reliable evidence that this is the case?
Obviously you don't like Digg (me neither). But what you also hint that there are sites which offer a more lucid take on technology news? What would you recommend? Can't see this as an Ask Slashdot topic ever!
Can you "show, provably" that "many are not either"?
Only joking!
Why is this flamebait? Surely the original post and the pathetic summary was flamebait?
As lots of people have commented, if you have admin rights you own the box.
Hear hear!
One poster has mentioned 3rd party re-packaging offerings but nearing the end of 2007 it's astonishing that Mozilla can't supply decent MSI packages of its flagship product. Or perhaps they don't really want to gain market share on corporate Windows desktops?
The other issue that is really hard is extensions. Realistically Firefox only comes alive when you get some decent extensions added. These would need to be deployable via MSI also before FF got serious corporate acceptance.
As I recall, Delphi 2006 introduced a new memory manager, FastMM, built not by Borland but as part of a community project.
I think my use of the word moron was a little over the top - I'm sure I learnt the lesson the same way, by experience.
However, I don't think you are right in saying that memory allocation in Windows is the problem. The Windows API in question is VirtualAlloc/VirtualFree. However, this is not designed for use directly by application code. It is meant for use by local application memory managers. On x86 memory is virtual and paged and the OS manages page tables etc. for each process. Each page has a minimum size (don't know what it is), but calling VirtualAlloc to ask for, say, 32 bytes of memory would be sheer lunacy.
Applications have different memory use patterns and so it only really makes sense to let applications tune their memory allocation strategies. Delphi's MM was designed for general purpose use by typical Delphi apps. They typically allocate and deallocate relatively small blocks of heap memory as objects are created and destroyed. If you have an app that doesn't perform well with the standard MM then you can plug in your own custom one.
On other OSs that use x86 I'm sure the same basic rules apply. For example I doubt that malloc on Linux x86 glibc will do virtual memory paging for each call. Instead it will do the paging for large blocks and then use a sub-allocator just as the Delphi MM does. I would also imagine, but I don't know anything about it, that Linux offers memory reservation and committing services analagous to Windows VirtualAlloc/VirtualFree.
It's not Windows doing the memory allocation, it's the Delphi MM. The Delphi MM isn't bad, it's better in more recent versions of Delphi.
Only a moron would allocate and deallocate the same sized buffer on every mouse move triggered paint. But still you choose to blame Windows rather than looking in the mirror for the culprit!
Yes but the interface looks better than Windows, the menus have rounded corners, and the design is better.
Who cares whether it works or not anyway!
You'd have to have some applications first though! Or is that where WINE comes in?
And it's still producing sensationalised, misleading headlines like this. Bring on the next 10 years!