What we really need is a company that hires a group of geeks with diverse experience, and who are willing to learn new stuff at customer request.
These companies do exist. I know because I used to run one. Why did I go elsewhere? Because in general no one is willing to pay for people to come in and fix their systems again and again as the employees make the same mistakes over and over again.
By the third time you re-install an app or an OS (I ended up dealing mostly with MS-Windows:( ), The client is convinced that their people can't be breaking things this much!
As someone else said, managers tend to hire big-name companies if they can afford it. If they can't afford it, they hire a local crew. Either way, they expect the same level of performance. This means among other things, doing all work after hours and issuing large credits whenever they think they deserve it.
The market is ripe for a group to come in and support the small to mid-range people that want Linux/BSD/UN*X and it's stability and features, but are too small to attract the big boys
this page has a list of all the filters and file types StarOffice can handle... Pretty damn long list!
This isn't as good as it sounds! This page actually shows all the filters available for the Linux release of StarOffice. It's a much shorter list.
I for one would like to know why this is. Is there a sound technical reason for releasing a larger number of filters for Windows and Solaris than for Linux?
Among those filters not supported under Linux are:
WordPerfect (all versions, DOS & Windows)
AmiPro
Lotus Manuscript
Frame Maker
Also, the way the Microsoft Word for Windows formats are described are different. On the "all the filters" page, these are described as folows:
Microsoft Word for Windows 2000, 97, 95, 6.0, 2.x, 1.x
Microsoft Word for Windows 95, 6.0 Microsoft Word for Windows 97/2000
While the "linux filters" page says:
Microsoft Word for Windows 6.0
Microsoft Word for Windows 95 Microsoft Word for Windows 97/2000
Are the two "versions" using different filter sets entierely?
... One of the goals of programming is to develop small, fast, and tight code that leaves as small a footprint as possible.
... while I enjoy eggs and I like to see them, with open source projects, you're only going to increase the size of the code (and possibly the complexity too, in hiding it) to add the eggs.
It's all about options after all. This is an ideal area where we can bring the power of autoconf et.al. into play. Just imagine the following build command
Makes it real easy to get rid of those eggs if you're space conscious, or to add them back in if you're feeling whimsical.
For that matter, since complexity is an issue here (and since easter eggs are popular), why not offload some of the work here into a library? Perhaps simply calling the eegg api in libegg.so.x.y...
Who knows? Maybe this is the beginning of the next great open source revolution. The problem of course is that we have enough to do already!:)
I personally would like these extra keys to work both at the console and in X. Toward this end, I did a little research. The showkey program is very useful for getting the scancodes off any keyboard. Here's how you do it:
$ showkey -s #-s for scancodes
Here's what I get from my Microsoft Pro keyboard, which I got because it's USB and my VAIO doesn't have a PS/2 port. (USB works great, BTW! I can also report success with the Vaio's USB floppy.)
I filled in the key names and states later--showkey of course has now idea what the key means.
Now, there are some interesting things to note here. The first thing that sticks out is the Sleep button. We only get two scancodes here, and I'm curious why. Second, all the sequences start with the 0xe0 scancode, which is a good thing, because it signals the start of an extended sequence. The big question that I'll be dealing with in a week or so is whether a keymap can be written that supports a key that effectivly appears to be two extended keys.
I intend to write a keymap for this keyboard, both for the console and for X, but don't let me stop someone else from doing it first!:) it shouldn't take a long time, but I've got to move first.
The suggestion that geeks can make themselves noticed by coming down with a variation of "blue flu" is an interesting concept, but there are some difficulties to be overcome.
The largest issue is that the type of activism being proposed is probably more than anyone in a similar situation could get away with. For example, turning off the routers is similar to taking away the heavy equipment (that is company owned, by the way!), while a normal "flu" or strike is simply a work stoppage. What has got management's attention in the past has been the work stoppage. Any attempt to actually sabotage equipment is generally met with a more hostile reaction.
The problem that technical people face in a work stoppage is that technical wizards have worked for years to make computers do what they do best: automatically deal with various events. Granted, this isn't perfect--things still break--but it does make the overall issue much more difficult.
So, how do do you do a work stoppage when you stopping work for a day doesn't necessarily mean that work stops?
It's true that Geek culture is hardly homogeneous, but should this be a problem? Other groups experience infighting as well. Some of the recent disagreements within the two mainstream US political parties are enough to demonstrate this. I believe that we as individuals disagree with our peers no more often than individuals within other groups.
Well, let before I go to far, let me qualify that.
Very often large political organizations, whether they be political parties, lobbies, or unions (yes, I know unions aren't strictly political, but for the purposes of this discussion, let's accept them as at least quasi-political) are built to serve some "grand design," and many people join without knowing why. I don't expect most geeks are like that. Geeks as a whole are very independent individuals, who prefer to do things in their own way, in their own time. This tendency (which is a good thing!) is likely to make any geek political organization more fractional. However, I think its still possible to build what the politicians call "consensus" (Some of us call it "caving in").
Political groups are built upon the similarities of a given group. This does not mean that the individuals in the group don't disagree on some issues. There may even be several factions within the group. Groups like the Republican and Democratic parties have held together because their leaders have the ability to find the subjects that almost everyone in the group can agree upon. There are a couple of important points here.
Just because an organization says one thing doesn't mean that all the members have the same opinion.
The leaders are the ones to build consensus. Why can't everyone be a leader?
As a community used to communicating across long distances and actually getting things done, it shouldn't be hard for some sufficiently interested group to put an agenda together. Remember: we're used to contributing day in and day out to various open collaborative projects. There's no reason why those same skills can't be applied to the subject of political activism.
Slashdot and other community sites are a perfect example. Some of the best (and worst) writing can be seen here. It's not all specifically on technologic issues either. Some very controversial topics get discussed, and clear, lucent rhetoric to rival that of any government official is the result.
The world noticed the Open Source movement because it worked. Maybe it's time for Open Politics.
Slashdot Mirror
These companies do exist. I know because I used to run one. Why did I go elsewhere? Because in general no one is willing to pay for people to come in and fix their systems again and again as the employees make the same mistakes over and over again.
By the third time you re-install an app or an OS (I ended up dealing mostly with MS-Windows :( ), The client is convinced that their people can't be breaking things this much!
As someone else said, managers tend to hire big-name companies if they can afford it. If they can't afford it, they hire a local crew. Either way, they expect the same level of performance. This means among other things, doing all work after hours and issuing large credits whenever they think they deserve it.
The market is ripe for a group to come in and support the small to mid-range people that want Linux/BSD/UN*X and it's stability and features, but are too small to attract the big boys
--
First, there's "S/390: The Linux Dream Machine, the article referring to Linux on the s/390 as "a herd of forty thousand raptors"
Then we have The Iron Penguin, coverage at linuxworld
Finally, here's IBM Runs 41,000 Copies of Linux on Mainframe, the original slashdot article.
--
This is already being worked on. A company called Silicon Film has a unit that you can drop ito any 35mm camera in place of film.
The resolution is reported to be 1280x1024 pixels at 36-bit color depth.
Release was scheduled for the first half of 2000. Now they're saying (of course) the second half. They're projecting the price at around $700.
I actually think I saw this on /. a long while back, but I can't seem to dredge up the story in a cursory search.
--
This isn't as good as it sounds! This page actually shows all the filters available for the Linux release of StarOffice. It's a much shorter list.
I for one would like to know why this is. Is there a sound technical reason for releasing a larger number of filters for Windows and Solaris than for Linux?
Among those filters not supported under Linux are:
- WordPerfect (all versions, DOS & Windows)
- AmiPro
- Lotus Manuscript
- Frame Maker
Also, the way the Microsoft Word for Windows formats are described are different. On the "all the filters" page, these are described as folows: While the "linux filters" page says: Are the two "versions" using different filter sets entierely?--
It's all about options after all. This is an ideal area where we can bring the power of autoconf et.al. into play. Just imagine the following build command
./configure --prefix=/usr --use-a-library --with-easter-eggs
$
Makes it real easy to get rid of those eggs if you're space conscious, or to add them back in if you're feeling whimsical.
For that matter, since complexity is an issue here (and since easter eggs are popular), why not offload some of the work here into a library? Perhaps simply calling the eegg api in libegg.so.x.y...
Who knows? Maybe this is the beginning of the next great open source revolution. The problem of course is that we have enough to do already! :)
--
I personally would like these extra keys to work both at the console and in X. Toward this end, I did a little research. The showkey program is very useful for getting the scancodes off any keyboard. Here's how you do it:
Here's what I get from my Microsoft Pro keyboard, which I got because it's USB and my VAIO doesn't have a PS/2 port. (USB works great, BTW! I can also report success with the Vaio's USB floppy.)
I filled in the key names and states later--showkey of course has now idea what the key means.
Now, there are some interesting things to note here. The first thing that sticks out is the Sleep button. We only get two scancodes here, and I'm curious why. Second, all the sequences start with the 0xe0 scancode, which is a good thing, because it signals the start of an extended sequence. The big question that I'll be dealing with in a week or so is whether a keymap can be written that supports a key that effectivly appears to be two extended keys.
I intend to write a keymap for this keyboard, both for the console and for X, but don't let me stop someone else from doing it first! :) it shouldn't take a long time, but I've got to move first.
--
The suggestion that geeks can make themselves noticed by coming down with a variation of "blue flu" is an interesting concept, but there are some difficulties to be overcome.
The largest issue is that the type of activism being proposed is probably more than anyone in a similar situation could get away with. For example, turning off the routers is similar to taking away the heavy equipment (that is company owned, by the way!), while a normal "flu" or strike is simply a work stoppage. What has got management's attention in the past has been the work stoppage. Any attempt to actually sabotage equipment is generally met with a more hostile reaction.
The problem that technical people face in a work stoppage is that technical wizards have worked for years to make computers do what they do best: automatically deal with various events. Granted, this isn't perfect--things still break--but it does make the overall issue much more difficult.
So, how do do you do a work stoppage when you stopping work for a day doesn't necessarily mean that work stops?
--
It's true that Geek culture is hardly homogeneous, but should this be a problem? Other groups experience infighting as well. Some of the recent disagreements within the two mainstream US political parties are enough to demonstrate this. I believe that we as individuals disagree with our peers no more often than individuals within other groups.
Well, let before I go to far, let me qualify that.
Very often large political organizations, whether they be political parties, lobbies, or unions (yes, I know unions aren't strictly political, but for the purposes of this discussion, let's accept them as at least quasi-political) are built to serve some "grand design," and many people join without knowing why. I don't expect most geeks are like that. Geeks as a whole are very independent individuals, who prefer to do things in their own way, in their own time. This tendency (which is a good thing!) is likely to make any geek political organization more fractional. However, I think its still possible to build what the politicians call "consensus" (Some of us call it "caving in").
Political groups are built upon the similarities of a given group. This does not mean that the individuals in the group don't disagree on some issues. There may even be several factions within the group. Groups like the Republican and Democratic parties have held together because their leaders have the ability to find the subjects that almost everyone in the group can agree upon. There are a couple of important points here.
As a community used to communicating across long distances and actually getting things done, it shouldn't be hard for some sufficiently interested group to put an agenda together. Remember: we're used to contributing day in and day out to various open collaborative projects. There's no reason why those same skills can't be applied to the subject of political activism.
Slashdot and other community sites are a perfect example. Some of the best (and worst) writing can be seen here. It's not all specifically on technologic issues either. Some very controversial topics get discussed, and clear, lucent rhetoric to rival that of any government official is the result.
The world noticed the Open Source movement because it worked. Maybe it's time for Open Politics.
--