If windows developers were using something more portable than directx, you'd be getting your games anyways...and they'd be on linux...and the programmers would have a more elegant, and less restricted API.
And just as soon as someone writes this thing of beauty, developers will flock to it because it is better (yeah, right). Meanwhile, I just don't care. Making games easier to port is a good thing.
Some choices are easy and beneficial in the short run, but damaging in the long run. Do we really want to give Microsoft control over the API's that all game developers use?
M$ doesn't control any of the APIs that developers use (well, on the PC platform, anyway - xbox may be another story). Developers use APIs that M$ controls. There is a difference. Developers can chose not to use M$ API anytime they like. It turns out that they don't care about making their programs portable to the other 7% (or whatever), and that DirectX is the best thing for writing to windows. If there is a library that will allow them to port trivially to MacOSX, I hope they will find it worth their while. Until this magical, bitchin', non-M$ API appears, this is our best bet.
Yeah, we should totally keep developers from easily porting games that work on 90% of the worlds computers. We should take away their choices, not give them more. Totally. I so agree with you.
Oh, wait. That was what the evil-me would have said. Really, I think making it easy to port games is a good thing. I don't give a crap what API folks use, just so long as I get to play the damn game on my bitchin' mac hardware.
Just as soon as OpenGL 'catches up', and game programmers start flocking to it, I won't give a shit - just as long as I get to play the damn game on my bitchin' mac hardware.
We did a little comparison where I work (I'm a WODeveloper). The upshot is that struts looks like the best non-WO tech I've seen. There are a lot of similarities, as I recall, though the discussion was about half a year ago - about twice my memories half-life:-)
In the last 15 years, I've been grateful to improvements in process. Great bug tracking software (I use bugzilla) and source management (cvs) has improved the process of writing and maintaining projects greatly.
In terms of language and library improvements, I think they enable programmers to finish projects much faster than they could before. Well, much more complicated projects. I think that OO languages allow programmers to write good code more easily.
Finally, I think these improvements most help good programmers. Anyone can code poorly, but good programmers are still hard to find. Process is as much a part of programming as language/library knowledge. When I was in school this was not really emphasized, and I think it's an important aspect of coding that most programs still fail to mention.
In any case, I bet you could find a linux or BSD distro that runs small and runs a windowserver. Though you might just want to try and find a dedicated browser and go for a web solution. If the client can just dial up and PPP to you and fire up a browser, you don't even have to worry about distributing updates to your software - just updating the server.
Really, more info about the problem domain would help...
Didn't the poster say he was going to lose his business.
Oh, please. He said it was AIMED at his business. I've seen where folks would not go to a linked story to read the article, but this takes the cake.
Really, the whole article is pretty weak - "Someone said something bad about someone's business - should they sue?" We really don't know what was said, in what forum, or who could have seen it. The only answer that makes any sense here is "maybe."
Really, though, someone on the 'net said something bad? Maybe even a flame? Quick, call the 10 o'clock news! An online retort is probably the only response that makes sense, with the possible exception of ignoring the whole thing.
Shell scripting was invented for a reason. Don't copy this script and expect it to work. But do copy the idea.
crontab: * * * * */somepath/checkping.zsh
/somepath/checkping.zsh: #!/bin/zsh if ([ test -e/tmp/connected.txt ]) then if (ping -c 1 www.apple.com) then # we are already up else # we went down rm/tmp/connected.txt/somepath/wentdown.sh fi else if (ping -c 1 www.apple.com) then date >/tmp/connected.txt/somepath/wentup.sh else # we were already down fi fi
Most of apple is static, including the trailers. I've seen apple's front page very slow, and I've seen the apple store - their biggest WO site - totally unresponsive.
Re:WebObjects is Web Applications done right.
on
Thursday Release Party
·
· Score: 3, Interesting
No, WO does handle this.
Pronoun trouble: if by 'this' you mean updating objects between sessions, then... well, you can do it, but it's not graceful out of the box - unless you're using a big shared editing context, which is possible, I suppose. If by 'this' you mean multiple connections to the DB from the same app, then no, WO does not do this out of the box. Which isn't to say that it can't be done - I've seen it done at a previous job, but it wasn't pretty, and it wasn't clean, and it probaly didn't help them.... and it can distribute updates between the caches across many machines.
Yes, it can, but it does not out of the box. What's more, with the number of updates/. sees, I'd have to think it'd be a fair performance hit to keep them all in sync under WO.
If/. were written in WO (And running on the same number of machines it runs on now, etc.) it would be better performing.
I just don't believe that, as much as I'd like to. PERL is damn fast, and damn light. I gotta think that Java and WO add plenty of overhead just because it's OO. Which is great for dev and most applications, but generally means you do take a performance hit. Not a big one, but some.
But once you did, you could do it. (I'm not taking into account time to create art, and content, just the code.)
Are you also taking into account all the/. that normal users don't see? All the admin crap, etc?
Re:WebObjects is Web Applications done right.
on
Thursday Release Party
·
· Score: 3, Insightful
Here's an example. I could create all of the functionality of slashdot, from scratch, including users, moderation, friend/foe, story posting, etc. in about 3 weeks. FROM SCRATCH. Oh, and it would have a hell of a lot better performance and scalability than slasdot has now.
That's a big fucking boast. WO is the bomb, but I'm not so sure you could write slash in 3 weeks, and I'm CERTAIN it would not have the same performance/scalability in that time. The biggest problems (for WOSlash) that I can think of: memory & updates. Sure you could scale it up onto a few machines easily, but the instaneous updates between sessions on multiple machines is a tough act. I don't know slash code, but I think they fetch almost everything from the db almost all the time, and WO simply isn't really good at that. The fact that you can't easily use more than one DB connection from one app is gonna kill you, I think.
Slash is probably one of the heaviest used sites in the world, and it holds up really well. Consider:/. doesn't get/.'ed, but Apple does.
Use a VPN. If you really don't want to let anyone else use your bandwidth (spoilsport), only let the WiFi connect to your server on your VPN port. You have to trust your VPN, but that's why they exist, right?
My WiFi is outside my firewall, and I don't limit access at all. I'm in San Bruno, not SF, so there are not as many interested parties (none, most of the time). Depending on how friendly you want to be to your neighbors/visiting friends/passers by, you might route the WiFi traffic through your server but limit the bandwidth. This is getting to be real work, though...
Most of the geeks that fall into the "dubious social behavior" group fit into the jerk catagory, not the asshole catagory. My wireless is outside my firewall, and I VPN my connection. This is great because it also means that I can go wireless (or even cabled) anywhere and not worry about someone sniffing my traffic.
If some geek passes by and wants to use some bandwidth, that's great. If it starts to happen a lot, I'll try to find them and work something out. With some luck, this happens 2-4 times, and we all agree to pitch in to get more bandwidth!
I can't believe all the folks saying that this is 'not hard to do'. As with most things, it's not hard to throw together some crap, but to do a really good job is a lot of work and planning.
It's not open source, but I used to work for a place that provided DRM software - all web based:
For me, it'd be extremely convenient to pull down the code via CVS to my nice SPARC laptop, take it home and work on it from anywhere in my house.
I totally agree. Get a Mac! I vpn in to work and do any amount of remote unix work on it. The built in airport cards are great, though I bought a linksys 802.11 instead of the airport - that's before they dropped the price. Since it's Solaris, the X11 remote displaying is a no-brainer.
Is this right for everyone, no. Is it right for people that god forbid prefer Solaris over Linux? Yes. Is it right for someone that does work or support from home (and god knows that EVERY tech person is on call 24/7), yes.
See, I don't agree. It's too damn expensive (I'm guessing). If you want to work at home, get 802.11 and something cheaper to set in your lap. I really think the most compelling arguement is for the sales dude who needs to tote the big DB and app to demo. For remote work... hell, that's why X11 apps remote so easily, right?
Power dereg in california was just done wrong. It was like having a big damn and deciding "OK, now we're going to get rid of the damn. All at once."
If they'd done something reasonable, like "SLOWLY deregulated": all power suppliers can only up their prices by 10% for the next 7 years, then anything goes. Sure, the cost to enter is high, and the profits are not, but that would have eased the transition: Year 1: everyone's bill goes up 10% - still no competition. Year 2: everyone's bill goes up another 10% - ditto Year 3: Look, ma, there's money to be made in california power; let's build a plant or 2! Meanwhile, everyone's bill goes up another 10% Year 4: More folks build; some competition goes online, the increases slow down because the new guy on the block will charge less. etc.
Yeah, years 1-3 suck, but look what we got! I imagine prices will drop as competitors start bringing generators online, but certainly the first year or 2 could have gone easier.
And that's the first smart thing anyone on this thread has said - including me. Sales folks are the only one's I can really see buying this stuff, just so they can tote around a full DB and 'big app'. And I still can't imagine all that many sales. Not for the amount this thing is gonna cost.
I can't help but ask...how many Real Engineers have you actually been in contact with in your entire life?
Not counting "Software Engineers", right? Can I count my father, who is an EE? That's 1.
And as I walk down the hall, I gotta think there are about 20 in our group here. And back when I worked with ME's, there were another... 15 to 20.
So, my grand total for engineers I've worked with, PLUS my dad is around 36-41. Not a big number - but some... And if we count "Software Engineers" that I have worked with since school, the number probably about doubles, but I don't think that's what you really mean.
While we're being conversational, here, how many have you worked with? Are you one? Are you in line to buy this laptop? Know anyone who is?
My, my...what a static, limited world you live in.
I blame my parents.
If you can't imagine thousands of engineers lining up to submit purchase orders for this new laptop...
Thousands... Nope. A few, yes.
heck, why don't we just replace all the servers in the world with Dell 1U units running XP? Heck, it works, doesn't it?
'Cause it doesn't do what I want. Dell 1U's running FreeBSD, on the other hand, would do me fine. I'd even go with linux if I needed to. But I think I'll buy OSX for my next server box because it's so easy to use and maintain (I've got freebsd now, and the last openssl patch was a royal pain).
Which brings me to my point: I suspect these laptops will be REALLY EXPENSIVE. Solaris is notorious for being an unfriendly OS for users. I'm thinking that most engineers will: . Buy a big server/desktop and remote display when they need to crunch numbers with a GUI, as is done around here most of the time. . Go with a friendlier OS for their notebook.
What's more, I don't think there are a whole lot of applications out there right now that I'd want to run on a 650Mhz notebook that need 4G RAM. Are there folks who need to run big DBs on their notebook? Yeah, a few. Thousands? Naw.
Finally, I note that Dell will sell me a notebook with 1G RAM and a 2.2GHz CPU for about $5200. I wonder if a similar config from SUN (with the same RAM and CPU at 1/3 speed) will cost less than 3x.
Slashdot Mirror
If windows developers were using something more portable than directx, you'd be getting your games anyways...and they'd be on linux...and the programmers would have a more elegant, and less restricted API.
And just as soon as someone writes this thing of beauty, developers will flock to it because it is better (yeah, right). Meanwhile, I just don't care. Making games easier to port is a good thing.
Some choices are easy and beneficial in the short run, but damaging in the long run. Do we really want to give Microsoft control over the API's that all game developers use?
M$ doesn't control any of the APIs that developers use (well, on the PC platform, anyway - xbox may be another story). Developers use APIs that M$ controls. There is a difference. Developers can chose not to use M$ API anytime they like. It turns out that they don't care about making their programs portable to the other 7% (or whatever), and that DirectX is the best thing for writing to windows. If there is a library that will allow them to port trivially to MacOSX, I hope they will find it worth their while. Until this magical, bitchin', non-M$ API appears, this is our best bet.
I thought that there was an effort to 'port' directX to linux as well. Which would be a great win for linux...
Yeah, we should totally keep developers from easily porting games that work on 90% of the worlds computers. We should take away their choices, not give them more. Totally. I so agree with you.
Oh, wait. That was what the evil-me would have said. Really, I think making it easy to port games is a good thing. I don't give a crap what API folks use, just so long as I get to play the damn game on my bitchin' mac hardware.
Just as soon as OpenGL 'catches up', and game programmers start flocking to it, I won't give a shit - just as long as I get to play the damn game on my bitchin' mac hardware.
We did a little comparison where I work (I'm a WODeveloper). The upshot is that struts looks like the best non-WO tech I've seen. There are a lot of similarities, as I recall, though the discussion was about half a year ago - about twice my memories half-life :-)
In the last 15 years, I've been grateful to improvements in process. Great bug tracking software (I use bugzilla) and source management (cvs) has improved the process of writing and maintaining projects greatly.
In terms of language and library improvements, I think they enable programmers to finish projects much faster than they could before. Well, much more complicated projects. I think that OO languages allow programmers to write good code more easily.
Finally, I think these improvements most help good programmers. Anyone can code poorly, but good programmers are still hard to find. Process is as much a part of programming as language/library knowledge. When I was in school this was not really emphasized, and I think it's an important aspect of coding that most programs still fail to mention.
What do you mean - they just want a GUI?
In any case, I bet you could find a linux or BSD distro that runs small and runs a windowserver. Though you might just want to try and find a dedicated browser and go for a web solution. If the client can just dial up and PPP to you and fire up a browser, you don't even have to worry about distributing updates to your software - just updating the server.
Really, more info about the problem domain would help...
Didn't the poster say he was going to lose his business.
Oh, please. He said it was AIMED at his business. I've seen where folks would not go to a linked story to read the article, but this takes the cake.
Really, the whole article is pretty weak - "Someone said something bad about someone's business - should they sue?" We really don't know what was said, in what forum, or who could have seen it. The only answer that makes any sense here is "maybe."
Really, though, someone on the 'net said something bad? Maybe even a flame? Quick, call the 10 o'clock news! An online retort is probably the only response that makes sense, with the possible exception of ignoring the whole thing.
Shell scripting was invented for a reason. Don't copy this script and expect it to work. But do copy the idea.
/somepath/checkping.zsh
/somepath/checkping.zsh: /tmp/connected.txt ]) /tmp/connected.txt /somepath/wentdown.sh /tmp/connected.txt /somepath/wentup.sh
crontab:
* * * * *
#!/bin/zsh
if ([ test -e
then
if (ping -c 1 www.apple.com)
then
# we are already up
else
# we went down
rm
fi
else
if (ping -c 1 www.apple.com)
then
date >
else
# we were already down
fi
fi
For a client, you might want to check out Chicken of the VNC.
You're probably thinking of the Apple store during Macworld expos. It's not unresponsive. It's being updated.
Actually, I'm pretty clear on
"The site is being updated"
vs.
timeout messages
The former happens when they are updating. The latter happens after I've gotten in and am trying to do something.
Most of apple is static, including the trailers. I've seen apple's front page very slow, and I've seen the apple store - their biggest WO site - totally unresponsive.
No, WO does handle this.
... well, you can do it, but it's not graceful out of the box - unless you're using a big shared editing context, which is possible, I suppose. ... and it can distribute updates between the caches across many machines.
/. sees, I'd have to think it'd be a fair performance hit to keep them all in sync under WO.
/. were written in WO (And running on the same number of machines it runs on now, etc.) it would be better performing.
/. that normal users don't see? All the admin crap, etc?
Pronoun trouble: if by 'this' you mean updating objects between sessions, then
If by 'this' you mean multiple connections to the DB from the same app, then no, WO does not do this out of the box.
Which isn't to say that it can't be done - I've seen it done at a previous job, but it wasn't pretty, and it wasn't clean, and it probaly didn't help them.
Yes, it can, but it does not out of the box. What's more, with the number of updates
If
I just don't believe that, as much as I'd like to. PERL is damn fast, and damn light. I gotta think that Java and WO add plenty of overhead just because it's OO. Which is great for dev and most applications, but generally means you do take a performance hit. Not a big one, but some.
But once you did, you could do it. (I'm not taking into account time to create art, and content, just the code.)
Are you also taking into account all the
Here's an example. I could create all of the functionality of slashdot, from scratch, including users, moderation, friend/foe, story posting, etc. in about 3 weeks. FROM SCRATCH. Oh, and it would have a hell of a lot better performance and scalability than slasdot has now.
/. doesn't get /.'ed, but Apple does.
That's a big fucking boast. WO is the bomb, but I'm not so sure you could write slash in 3 weeks, and I'm CERTAIN it would not have the same performance/scalability in that time.
The biggest problems (for WOSlash) that I can think of: memory & updates. Sure you could scale it up onto a few machines easily, but the instaneous updates between sessions on multiple machines is a tough act. I don't know slash code, but I think they fetch almost everything from the db almost all the time, and WO simply isn't really good at that. The fact that you can't easily use more than one DB connection from one app is gonna kill you, I think.
Slash is probably one of the heaviest used sites in the world, and it holds up really well. Consider:
Use a VPN. If you really don't want to let anyone else use your bandwidth (spoilsport), only let the WiFi connect to your server on your VPN port. You have to trust your VPN, but that's why they exist, right?
My WiFi is outside my firewall, and I don't limit access at all. I'm in San Bruno, not SF, so there are not as many interested parties (none, most of the time). Depending on how friendly you want to be to your neighbors/visiting friends/passers by, you might route the WiFi traffic through your server but limit the bandwidth. This is getting to be real work, though...
Most of the geeks that fall into the "dubious social behavior" group fit into the jerk catagory, not the asshole catagory. My wireless is outside my firewall, and I VPN my connection. This is great because it also means that I can go wireless (or even cabled) anywhere and not worry about someone sniffing my traffic.
If some geek passes by and wants to use some bandwidth, that's great. If it starts to happen a lot, I'll try to find them and work something out. With some luck, this happens 2-4 times, and we all agree to pitch in to get more bandwidth!
I can't believe all the folks saying that this is 'not hard to do'. As with most things, it's not hard to throw together some crap, but to do a really good job is a lot of work and planning.
It's not open source, but I used to work for a place that provided DRM software - all web based:
WebWare
It is currently WebObjects based, but I understand they're moving to J2EE.
Help me out, cause I can't keep track - don't they still sell Solaris for X86? And wouldn't that be a much less expensive proposition?
For me, it'd be extremely convenient to pull down the code via CVS to my nice SPARC laptop, take it home and work on it from anywhere in my house.
I totally agree. Get a Mac! I vpn in to work and do any amount of remote unix work on it. The built in airport cards are great, though I bought a linksys 802.11 instead of the airport - that's before they dropped the price. Since it's Solaris, the X11 remote displaying is a no-brainer.
Is this right for everyone, no. Is it right for people that god forbid prefer Solaris over Linux? Yes. Is it right for someone that does work or support from home (and god knows that EVERY tech person is on call 24/7), yes.
See, I don't agree. It's too damn expensive (I'm guessing). If you want to work at home, get 802.11 and something cheaper to set in your lap. I really think the most compelling arguement is for the sales dude who needs to tote the big DB and app to demo. For remote work... hell, that's why X11 apps remote so easily, right?
Power dereg in california was just done wrong. It was like having a big damn and deciding "OK, now we're going to get rid of the damn. All at once."
If they'd done something reasonable, like "SLOWLY deregulated": all power suppliers can only up their prices by 10% for the next 7 years, then anything goes. Sure, the cost to enter is high, and the profits are not, but that would have eased the transition:
Year 1: everyone's bill goes up 10% - still no competition.
Year 2: everyone's bill goes up another 10% - ditto
Year 3: Look, ma, there's money to be made in california power; let's build a plant or 2! Meanwhile, everyone's bill goes up another 10%
Year 4: More folks build; some competition goes online, the increases slow down because the new guy on the block will charge less.
etc.
Yeah, years 1-3 suck, but look what we got! I imagine prices will drop as competitors start bringing generators online, but certainly the first year or 2 could have gone easier.
Damn, the keyspan USB remote is just $79.
And that's the first smart thing anyone on this thread has said - including me. Sales folks are the only one's I can really see buying this stuff, just so they can tote around a full DB and 'big app'. And I still can't imagine all that many sales. Not for the amount this thing is gonna cost.
I can't help but ask...how many Real Engineers have you actually been in contact with in your entire life?
Not counting "Software Engineers", right? Can I count my father, who is an EE? That's 1.
And as I walk down the hall, I gotta think there are about 20 in our group here. And back when I worked with ME's, there were another... 15 to 20.
So, my grand total for engineers I've worked with, PLUS my dad is around 36-41. Not a big number - but some... And if we count "Software Engineers" that I have worked with since school, the number probably about doubles, but I don't think that's what you really mean.
While we're being conversational, here, how many have you worked with? Are you one? Are you in line to buy this laptop? Know anyone who is?
And when you're in a place where you've taken your portable laptop, you oftentimes don't have access to the network.
Actually, I almost always do have access to the network. Maybe I'm just spoiled.
My, my...what a static, limited world you live in.
I blame my parents.
If you can't imagine thousands of engineers lining up to submit purchase orders for this new laptop...
Thousands... Nope. A few, yes.
heck, why don't we just replace all the servers in the world with Dell 1U units running XP? Heck, it works, doesn't it?
'Cause it doesn't do what I want. Dell 1U's running FreeBSD, on the other hand, would do me fine. I'd even go with linux if I needed to. But I think I'll buy OSX for my next server box because it's so easy to use and maintain (I've got freebsd now, and the last openssl patch was a royal pain).
Which brings me to my point: I suspect these laptops will be REALLY EXPENSIVE. Solaris is notorious for being an unfriendly OS for users. I'm thinking that most engineers will:
. Buy a big server/desktop and remote display when they need to crunch numbers with a GUI, as is done around here most of the time.
. Go with a friendlier OS for their notebook.
What's more, I don't think there are a whole lot of applications out there right now that I'd want to run on a 650Mhz notebook that need 4G RAM. Are there folks who need to run big DBs on their notebook? Yeah, a few. Thousands? Naw.
Finally, I note that Dell will sell me a notebook with 1G RAM and a 2.2GHz CPU for about $5200. I wonder if a similar config from SUN (with the same RAM and CPU at 1/3 speed) will cost less than 3x.
An administrator of a fully SPARC-based network
...
...
... dozens of folks like that! No, really - I'm CERTAIN there are at least 30!
That's afraid of ssh and remote X windows.
Someone with legacy Solaris programs that they need to make transportable
That's afraid of ssh and remote X windows.
And my favorite:
Someone who just plain prefers Solaris to Linux (believe it or not, they exist)
What a fantastic marketplace! There could be
And you thought that Apple hardware was expensive.