We evaluated WebDAV on a hosted system and various open-source solutions (like hosted Alfresco) as alternatives to a company-wide Dropbox license. The fact is that if you want to have anything more sophisticated than a simple fileserver (e.g. different folder permissions, multiple file versions, somewhat sane conflict resolution), there is no good free alternative at this point if you have remote people --if you've heard of one, I'd love a pointer.
For a local LAN, I'd stick with Alfresco on a decent box, but Alfresco falls apart on remote connections, and plain WebDAV is too slow / buggy.
In the end we went with Egnyte. It's not without its faults (buggy iOS client for one, and the Windows clients need some optimization), but it does more than Dropbox/Box.net/Sugarsync/Syncplicity, works great for SOHOs and it's actually cheaper than a VPS that can handle Alfresco and the like.
Re:Another great Python 3.x series release
on
Python 3.2 Released
·
· Score: 2
Why not? isn't a programming language a big library of sorts anyway? should you keep supporting every bad design decision for ever and ever?
Python has been extremely conservative about both introducing and deprecating features (the __future__ import is genius). Python 3 had to stay within the rational side of the Perl-6 line, and I believe they pulled it off.
If you're gonna go back to that era, someone needs to mention the great shareware FTP archives of the pre-Web internet: wustl.edu, garbo.uwasa.fi, simpnet (? it's been too long).
Am I the only one who kept checking the READMEs for new pointers by the maintainers? I probably downloaded way too many games on the recommendation of obscure Finnish professors:-)
More like a 3rd of that, unless you assume that every man, woman and child regardless of age, owns and operates a PC.
Desktop Linux a marginal market niche of enthusiasts who will continue to consume it regardless of competitive advantage (i.e. either because of politics or because they need a commodity Unix-like OS on their desktops). That means that as a commercial alternative Linux is dead, unless your target market is that same enthusiast market --after all people still sell software for the Amiga.
The larger question that I don't see debated anywhere is: is for-profit open-source software development dead? Exclude the service providers (like RedHat) and 1-2 politically motivated success stories (Firefox; more of an ad service than a software vendor anyway), and who are the software vendors who are making money off of open source?
They explicitly states that they looked at Twisted and chose to write something more user-friendly. Having looked at Twisted (3-4 years ago though) and at Tornado's samples and benchmarks I think they succeeded. Twisted seems to be going the way of Zope: an interesting platform that did everything its own way and shut itself out from the rest of the Python universe, eventually losing relevancy.
I think a Tornado/Django mashup (Tornado infrastructure, Django front-end/application bootstrapping) would be realllly interesting....
Your setup sounds very interesting and close to what I've been looking for for ages. Is that something you setup yourself? are there services out there that will set something like that up for a small business? links or pointers?
The whitespace issue is a red-herring: most people get used to it quickly and it's not as strict as it sounds (you can mix-and-match tabs and spaces, as long as you are consistent for each *block*; not even an entire.py file). There's two real-world problems with it: copy-and-paste and generating Python code. Both are much less common than looking at badly-formatted code that it takes a bit to mentally parse which brace-delineated languages have.
The core of the problem is that this entire argument hasn't changed (much less resolved) in oh, about 14 years. Linux at some point looked like it could succeed Windows 98 to become the OS of choice, and then Win2k (and then XP) killed it. That was what, 1998? Just for reference, there was barely a google.stanford.edu then, AltaVista was still the king of the hill, Novel ruled the small server market, and NeXTStep did pretty much everything OS X does today.
The Linux companies (never mind the 'community'; hackers will do what hackers want to, by definition) need to wake up and band together to fix some underlying core issues with the platform: file structure layout, configuration and preferences storage, device support, user management, etc, etc. They are invested way too much in making their *versions* of the platform work as opposed to making the entire platform work, and their versions excel.
I've been waiting for a proper iPhone to replace my N95. The specs from the keynote don't answer some basic questions:
Are the GPS maps stored or downloaded ad-hoc? because for those of us that actually travel and use GPS in foreign lands, paying 3G roaming rates isn't exactly a bargain.
OK, MobileMe sounds great, but what about Bluetooth syncing? again, if you are traveling, you can't rely on getting a WiFi signal between your phone and your laptop to sync your calendar (and you don't want to have to remember to plug it into USB either).
What about that camera? still 2MP, in 2008? AutoFocus, anything?
The iPhone has an awesome screen and a great UI, but even this fixed version will probably fall again short of the N9x series outside of the US, where ppl don't usually have wide-spread WiFi, or unlimited 3G access, or care about PC syncing. Pity, here's waiting for WWDC 2009 again...
I was in the same boat, and made the same choice. Scons is very good (it rocks for C/C++ projects for example), but for custom build systems is way too complicated. I wish there was a "skin" you could use over Scons to build custom systems w/o delving into Scons internals unless you really, really have to.
You're absolutely right, file manipulation is a wart in Python. Thankfully, there's the path module, a good idea done well (and which should be in the stdlib!).
Now, forcing WDS down users' throats is beyond the pale, but saying that GDS is better is just false: WDS has a better UI than GDS (and not just because it's an actual Windows app), does a much, much better job indexing mail than GDS (it can actually handle mail being moved between folders and indexes attachments just fine) and has better indexing behavior: it deals with moved/deleted documents better and its index doesn't grow out of proportions like GDS', nor does it seem to 'forget' documents.
On the other hand, WDS daemon may eat a bit more CPU at times, it doesn't index Firefox histories and obviously doesn't integrate with Google's web search, which is the only killer GDS feature as far as I am concerned.
Admittedly, the above is based on my experience with GDS ~12 months ago: I switched to WDS and never looked back. I'd be happy to give Google another shot if they've improved their product.
I have an N95 and just used it recently to live in a strange city for two days. Yes, it's a killer app; having the navigation device on you (instead of the car) means you can make impromptu plans work ("I'd like a coffee; wonder where the nearest cafe is; oh, 5 mins away, no prob"). Having Google on the same device just plain rocks.
Now, it's not perfect: GPS drains the battery down something fierce (which is not great to begin with) and Nokia could have done a better job interfacing GPS to the rest of the phone (why can I cut and paste a full address to search in the Maps app? instead I have to break down things to street number, street name, city and zip; why? the N95 is powerful enough to guess that format by itself...). Still though, the concept is compelling; Nokia has the right idea.
This is just a start; you can go a long ways trying to determine who's an expert by checking the extent and trustworthiness of their contributions within an article "cluster" (where clusters can be determined through link graphs or content correlation).
Wikipedia can also try using implicit "voting" on articles by tracking how many of their users have read a page and "approved" it by not changing it. Here your vote can also be linked to your trustworthiness. And of course you can have explicit voting/.-style. The problem with these approaches is that it would be harder to share the voting data with sites using Wikipedia's content.
I forget where I read this, but apparently Safari had the ability to auto-detect phone numbers and addresses in web pages for a while (kinda like SmartTags once did;-) So the demoed features of the iPhone is essentially text-parsing.
Or to put it in another way, the morose claim that Safari (a browser that should be sandboxed from its host) is the SDK for the iPhone means: no app access to the iPhone hardware (camera, wifi, radios, light/proximity/orientation sensors, GPS one day). In other words, the single greatest thing about smartphones (portable, advanced hardware) is useless in this case.
But wait, there's less: "Safari is the SDK" means that your PC can give your app just as much info as the iPhone can. In other words your app will not have access to your address book, your calendar, your email, your pictures, your music, your to-do list, and all the other stuff that you'd expect to store and manage on a $600 smartphone.
Yes, Apple's apps may end up being that good, but the RTFA makes a good point...
Did you try Safari/Win? I just did, and in 5 secs flat I noticed the following: no resizing from all sides (although it popped up taller than the screen height), no Alt-D to get to the address bar, no Ctrl-Enter to fill-in www.*.com. Maybe Alt-D is not the end of the world, but no edge-resizing? is there a WinXP port of KHTML/WebKit written by actual windows devs?
I am a happy XBMC user for years now. However, I think XBMC has hit a ceiling of sorts: there are still bugs aplenty, especially around error handling: a small hiccup on your LAN and XBMC may freeze or stop media playing. Canceling out of most network-streaming ops also hangs XBMC... these have been around for years and although they've been getting better, it's not exactly a consumer-friendly experience.
Secondly, interfaces are getting complex enough to need their own IDE (yes, I know there's one floating around): Project Mayhem III is OK, but it's not perfect and modding an XBMC skin is not for the faint of heart (why didn't XBMC just use Python to script the skins instead of ant-like XML? they already have a Python engine embedded...)
Finally, Xbox I is also showing its age: can't handle HDTV, and now with small-factor PCs being quiet and small enough, I'd rather have a less-capable UI than the jet-engine that's the Xbox fan+HDD (yes, I've modded the fan with a low dB one; the HDD is still freaking noisy).
BTW, does anyone know if the Mac mini + Front Row can match most of XBMC? codec support and the ability to run arbitrary subtitles on videos are top of my list...
Well, I think the RTFA has a point, but doesn't quite take it all the way home: the reasons web-based software is popular are simple: no installation/uninstallation hassles, automatic "updates" of the software, network transparency, collaboration, etc.
There is nothing in that list that restricts that app to be browser-*based*. You can build an offline, desktop-based (but maybe browser-*launched*) app that installs and uninstalls simply, has collaboration features and updates automatically. With VM technology brewed into the Linux and Vista++ kernels, there is no technical reason why we can't have desktop apps that have the best of both worlds. We just need a standard (or even better some sort of framework) to make this easier. That's I think where Apollo, Silverlight, or JavaFX want to take us; it'd be nice if there was an OSS stack (like the LAMP stack) that matched these capabilities...
That's not price optimization: that's discriminatory pricing or price differentiation. The latter is usually done with coupons ("here's a $100 loyalty coupon for you Mr Smith"), the former is about maximizing profit (i.e. profit per unit times total units sold).
Is there any technical reason why BitTorrent doesn't let you have "sub-torrents"? As in for a set of files A within a torrent, some files maybe part of another torrent, available on a another tracker potentially.
How exactly has capitalism broken down here? I am not saying it's a perfect system, but Circuit City's move has not played out yet: if they survive this, then they will be a stronger company and be able to hire more people or at higher wages. If they don't (or if their results get worse), then other companies will shrink back from following their lead.
The underlying problem of big-box stores like CC and CompUSA is that their customers are moving to direct sales, and the stores end up little more than showrooms --in which case, higher-cache brands like Apple or Sony (which are *expanding* into retail) have the advantage.
This is no different than what's facing bookstores, CD sellers or any other business model that relied on having inventory close to the customer. The main difference being that computers are not an impulse purchase like a CD or a book, so these stores have an even harder time adjusting. That doesn't mean that they can't, nor does it mean they shouldn't try.
FWIW, I agree that this is a pretty bad move when your competitors are trying to differentiate themselves on quality of service, and you just de-moralized all your remaining floor employees... but it's too soon to call this.
To add to your list: GDS doesn't index Outlook/email attachments even if they are in a format that it does know how to index. Like you mention, it doesn't deal well with documents moving from one location to another (not just within Outlook, anywhere in the filesystem). And the bug you mention about email is much worse than just not able to locate a moved email: it means that spam that gets moved by a client-filter to a folder you've told GDS not to index, will still be in the GDS index because it usually indexes it before the spam filter gets to move it. So, your index eventually gets clogged up with spam too.
It gets worse: GDS actually "forgets" about documents it has previously indexed (so results get *worse* over time, not better). And its index keeps growing (yes, even though its results are getting worse). And as the parent mentions, it doesn't have a "re-index now" option, so you are forced to uninstall and re-install.
The only good thing about GDS is its integration with google.com (who's embracing and extending now?). I am no MS apologist and I put up with GDS for over 1.5 years, but I switched to Windows Desktop Search and never looked back: WDS is head-and-shoulders above GDS (BTW, it can be downloaded into XP and is pretty much the same as the WDS in Vista): better results, better UI, way better integration with Windows, smaller index, ability to re-set the index whenever and faster to index the drive than GDS to begin with. WDS started life as Lookout, a third-party freeware app that was bought by MS, and it was better than GDS back then (oh what 4 years ago?).
If only developers would embrace WDS to fix some obvious shortcomings (no Firefox/Thunderbird indexing, no hotkeys like GDS). I doubt Microsoft has anything to fear from Google competing for the desktop if GDS is any indication...
Slashdot Mirror
Limnos is the most fortified Greek island and a huge military outpost. What did they expect exactly?
PoE means a single UPS to keep all the thin clients alive during power outages. Think call centers in countries with spotty power grids...
We evaluated WebDAV on a hosted system and various open-source solutions (like hosted Alfresco) as alternatives to a company-wide Dropbox license. The fact is that if you want to have anything more sophisticated than a simple fileserver (e.g. different folder permissions, multiple file versions, somewhat sane conflict resolution), there is no good free alternative at this point if you have remote people --if you've heard of one, I'd love a pointer.
For a local LAN, I'd stick with Alfresco on a decent box, but Alfresco falls apart on remote connections, and plain WebDAV is too slow / buggy.
In the end we went with Egnyte. It's not without its faults (buggy iOS client for one, and the Windows clients need some optimization), but it does more than Dropbox/Box.net/Sugarsync/Syncplicity, works great for SOHOs and it's actually cheaper than a VPS that can handle Alfresco and the like.
Why not? isn't a programming language a big library of sorts anyway? should you keep supporting every bad design decision for ever and ever?
Python has been extremely conservative about both introducing and deprecating features (the __future__ import is genius). Python 3 had to stay within the rational side of the Perl-6 line, and I believe they pulled it off.
If you're gonna go back to that era, someone needs to mention the great shareware FTP archives of the pre-Web internet: wustl.edu, garbo.uwasa.fi, simpnet (? it's been too long).
Am I the only one who kept checking the READMEs for new pointers by the maintainers? I probably downloaded way too many games on the recommendation of obscure Finnish professors :-)
More like a 3rd of that, unless you assume that every man, woman and child regardless of age, owns and operates a PC.
Desktop Linux a marginal market niche of enthusiasts who will continue to consume it regardless of competitive advantage (i.e. either because of politics or because they need a commodity Unix-like OS on their desktops). That means that as a commercial alternative Linux is dead, unless your target market is that same enthusiast market --after all people still sell software for the Amiga.
The larger question that I don't see debated anywhere is: is for-profit open-source software development dead? Exclude the service providers (like RedHat) and 1-2 politically motivated success stories (Firefox; more of an ad service than a software vendor anyway), and who are the software vendors who are making money off of open source?
They explicitly states that they looked at Twisted and chose to write something more user-friendly. Having looked at Twisted (3-4 years ago though) and at Tornado's samples and benchmarks I think they succeeded. Twisted seems to be going the way of Zope: an interesting platform that did everything its own way and shut itself out from the rest of the Python universe, eventually losing relevancy.
I think a Tornado/Django mashup (Tornado infrastructure, Django front-end/application bootstrapping) would be realllly interesting....
Your setup sounds very interesting and close to what I've been looking for for ages. Is that something you setup yourself? are there services out there that will set something like that up for a small business? links or pointers?
The whitespace issue is a red-herring: most people get used to it quickly and it's not as strict as it sounds (you can mix-and-match tabs and spaces, as long as you are consistent for each *block*; not even an entire .py file). There's two real-world problems with it: copy-and-paste and generating Python code. Both are much less common than looking at badly-formatted code that it takes a bit to mentally parse which brace-delineated languages have.
The core of the problem is that this entire argument hasn't changed (much less resolved) in oh, about 14 years. Linux at some point looked like it could succeed Windows 98 to become the OS of choice, and then Win2k (and then XP) killed it. That was what, 1998? Just for reference, there was barely a google.stanford.edu then, AltaVista was still the king of the hill, Novel ruled the small server market, and NeXTStep did pretty much everything OS X does today.
The Linux companies (never mind the 'community'; hackers will do what hackers want to, by definition) need to wake up and band together to fix some underlying core issues with the platform: file structure layout, configuration and preferences storage, device support, user management, etc, etc. They are invested way too much in making their *versions* of the platform work as opposed to making the entire platform work, and their versions excel.
The iPhone has an awesome screen and a great UI, but even this fixed version will probably fall again short of the N9x series outside of the US, where ppl don't usually have wide-spread WiFi, or unlimited 3G access, or care about PC syncing. Pity, here's waiting for WWDC 2009 again...
Well, Lenovo just released this, which looks very nice to this road warrior...
I was in the same boat, and made the same choice. Scons is very good (it rocks for C/C++ projects for example), but for custom build systems is way too complicated. I wish there was a "skin" you could use over Scons to build custom systems w/o delving into Scons internals unless you really, really have to.
You're absolutely right, file manipulation is a wart in Python. Thankfully, there's the path module, a good idea done well (and which should be in the stdlib!).
Now, forcing WDS down users' throats is beyond the pale, but saying that GDS is better is just false: WDS has a better UI than GDS (and not just because it's an actual Windows app), does a much, much better job indexing mail than GDS (it can actually handle mail being moved between folders and indexes attachments just fine) and has better indexing behavior: it deals with moved/deleted documents better and its index doesn't grow out of proportions like GDS', nor does it seem to 'forget' documents.
On the other hand, WDS daemon may eat a bit more CPU at times, it doesn't index Firefox histories and obviously doesn't integrate with Google's web search, which is the only killer GDS feature as far as I am concerned.
Admittedly, the above is based on my experience with GDS ~12 months ago: I switched to WDS and never looked back. I'd be happy to give Google another shot if they've improved their product.
I have an N95 and just used it recently to live in a strange city for two days. Yes, it's a killer app; having the navigation device on you (instead of the car) means you can make impromptu plans work ("I'd like a coffee; wonder where the nearest cafe is; oh, 5 mins away, no prob"). Having Google on the same device just plain rocks.
Now, it's not perfect: GPS drains the battery down something fierce (which is not great to begin with) and Nokia could have done a better job interfacing GPS to the rest of the phone (why can I cut and paste a full address to search in the Maps app? instead I have to break down things to street number, street name, city and zip; why? the N95 is powerful enough to guess that format by itself...). Still though, the concept is compelling; Nokia has the right idea.
This is just a start; you can go a long ways trying to determine who's an expert by checking the extent and trustworthiness of their contributions within an article "cluster" (where clusters can be determined through link graphs or content correlation).
/.-style. The problem with these approaches is that it would be harder to share the voting data with sites using Wikipedia's content.
Wikipedia can also try using implicit "voting" on articles by tracking how many of their users have read a page and "approved" it by not changing it. Here your vote can also be linked to your trustworthiness. And of course you can have explicit voting
I forget where I read this, but apparently Safari had the ability to auto-detect phone numbers and addresses in web pages for a while (kinda like SmartTags once did ;-) So the demoed features of the iPhone is essentially text-parsing.
Or to put it in another way, the morose claim that Safari (a browser that should be sandboxed from its host) is the SDK for the iPhone means: no app access to the iPhone hardware (camera, wifi, radios, light/proximity/orientation sensors, GPS one day). In other words, the single greatest thing about smartphones (portable, advanced hardware) is useless in this case.
But wait, there's less: "Safari is the SDK" means that your PC can give your app just as much info as the iPhone can. In other words your app will not have access to your address book, your calendar, your email, your pictures, your music, your to-do list, and all the other stuff that you'd expect to store and manage on a $600 smartphone.
Yes, Apple's apps may end up being that good, but the RTFA makes a good point...
Did you try Safari/Win? I just did, and in 5 secs flat I noticed the following: no resizing from all sides (although it popped up taller than the screen height), no Alt-D to get to the address bar, no Ctrl-Enter to fill-in www.*.com. Maybe Alt-D is not the end of the world, but no edge-resizing? is there a WinXP port of KHTML/WebKit written by actual windows devs?
I am a happy XBMC user for years now. However, I think XBMC has hit a ceiling of sorts: there are still bugs aplenty, especially around error handling: a small hiccup on your LAN and XBMC may freeze or stop media playing. Canceling out of most network-streaming ops also hangs XBMC... these have been around for years and although they've been getting better, it's not exactly a consumer-friendly experience.
Secondly, interfaces are getting complex enough to need their own IDE (yes, I know there's one floating around): Project Mayhem III is OK, but it's not perfect and modding an XBMC skin is not for the faint of heart (why didn't XBMC just use Python to script the skins instead of ant-like XML? they already have a Python engine embedded...)
Finally, Xbox I is also showing its age: can't handle HDTV, and now with small-factor PCs being quiet and small enough, I'd rather have a less-capable UI than the jet-engine that's the Xbox fan+HDD (yes, I've modded the fan with a low dB one; the HDD is still freaking noisy).
BTW, does anyone know if the Mac mini + Front Row can match most of XBMC? codec support and the ability to run arbitrary subtitles on videos are top of my list...
Well, I think the RTFA has a point, but doesn't quite take it all the way home: the reasons web-based software is popular are simple: no installation/uninstallation hassles, automatic "updates" of the software, network transparency, collaboration, etc.
There is nothing in that list that restricts that app to be browser-*based*. You can build an offline, desktop-based (but maybe browser-*launched*) app that installs and uninstalls simply, has collaboration features and updates automatically. With VM technology brewed into the Linux and Vista++ kernels, there is no technical reason why we can't have desktop apps that have the best of both worlds. We just need a standard (or even better some sort of framework) to make this easier. That's I think where Apollo, Silverlight, or JavaFX want to take us; it'd be nice if there was an OSS stack (like the LAMP stack) that matched these capabilities...
That's not price optimization: that's discriminatory pricing or price differentiation. The latter is usually done with coupons ("here's a $100 loyalty coupon for you Mr Smith"), the former is about maximizing profit (i.e. profit per unit times total units sold).
Yes, I am a supply chain consultant...
Is there any technical reason why BitTorrent doesn't let you have "sub-torrents"? As in for a set of files A within a torrent, some files maybe part of another torrent, available on a another tracker potentially.
How exactly has capitalism broken down here? I am not saying it's a perfect system, but Circuit City's move has not played out yet: if they survive this, then they will be a stronger company and be able to hire more people or at higher wages. If they don't (or if their results get worse), then other companies will shrink back from following their lead.
The underlying problem of big-box stores like CC and CompUSA is that their customers are moving to direct sales, and the stores end up little more than showrooms --in which case, higher-cache brands like Apple or Sony (which are *expanding* into retail) have the advantage.
This is no different than what's facing bookstores, CD sellers or any other business model that relied on having inventory close to the customer. The main difference being that computers are not an impulse purchase like a CD or a book, so these stores have an even harder time adjusting. That doesn't mean that they can't, nor does it mean they shouldn't try.
FWIW, I agree that this is a pretty bad move when your competitors are trying to differentiate themselves on quality of service, and you just de-moralized all your remaining floor employees... but it's too soon to call this.
To add to your list: GDS doesn't index Outlook/email attachments even if they are in a format that it does know how to index. Like you mention, it doesn't deal well with documents moving from one location to another (not just within Outlook, anywhere in the filesystem). And the bug you mention about email is much worse than just not able to locate a moved email: it means that spam that gets moved by a client-filter to a folder you've told GDS not to index, will still be in the GDS index because it usually indexes it before the spam filter gets to move it. So, your index eventually gets clogged up with spam too.
It gets worse: GDS actually "forgets" about documents it has previously indexed (so results get *worse* over time, not better). And its index keeps growing (yes, even though its results are getting worse). And as the parent mentions, it doesn't have a "re-index now" option, so you are forced to uninstall and re-install.
The only good thing about GDS is its integration with google.com (who's embracing and extending now?). I am no MS apologist and I put up with GDS for over 1.5 years, but I switched to Windows Desktop Search and never looked back: WDS is head-and-shoulders above GDS (BTW, it can be downloaded into XP and is pretty much the same as the WDS in Vista): better results, better UI, way better integration with Windows, smaller index, ability to re-set the index whenever and faster to index the drive than GDS to begin with. WDS started life as Lookout, a third-party freeware app that was bought by MS, and it was better than GDS back then (oh what 4 years ago?).
If only developers would embrace WDS to fix some obvious shortcomings (no Firefox/Thunderbird indexing, no hotkeys like GDS). I doubt Microsoft has anything to fear from Google competing for the desktop if GDS is any indication...