Jump to content

Banner2.jpg.35fd74882a15b2b8a1b4142f7dcc8bed.jpg

EQDirect cable under Linux


malc-c
 Share

Recommended Posts

This weekend I've been playing with a distro of Linux called Mint on an Acer Netbook. It's really good and unlike windows the software really runs at a user friendly speed. However having downloaded both CdC and Kstars (the latter looks really nice and simple to use) I can't get the software to communicate with the EQDirect lead made from an FTDI 232R - 5v cable. I have managed to locate the system log and when the cable is plugged in the driver is loaded and installed, with the port assigned as ttyusb0. However changing the default \dev\ttys0 entry in the port settings for kstars, to \dev\ttyusb0 it still fails to communicate with the scope and simply sits there searching / attempting connection. If I let the software search it still doesn't find it after 20 minutes.

I'm in the process of upgrading the observatory PC soon, and thought about using Linux as an alternative to windows 7, however, if I can't get any planetarium software to drive the scope it looks like I'll have to stick with Win7. Any Linux user out there that has any suggestions I would like to hear them

Link to comment
Share on other sites

I do not think that there is any Linux software that can communicate directly with the mount. My guess is that you need to have the handset in place and connect the serial cable to that. EQDIR connect directly to the mount and the protocol there is a more or less internal one that the handset uses to directly control steppers and the likes in there.

/p

Link to comment
Share on other sites

In case you weren't aware, Linux is case-sensitive, so /dev/ttyusb0 is not the same as /dev/ttyUSB0 (the latter is correct, I think).

It's also possible that you aren't in the user group that is allowed access to serial devices. Using a terminal window, if you do

$ ls -l /dev/ttyUSB0

you should get something like

crw-rw---- 1 root dialout 4, 64 Jul 26 08:53 /dev/ttyUSB0

"root" is the user that owns the device, and "dialout" is the group. If you then do

$ id

you'll be given your own user id and the group ids to which you belong, eg.

uid=1024(james) gid=1024(james) groups=1024(james),4(adm),20(dialout),24(cdrom),27(sudo),30(dip),46(plugdev),108(lpadmin),125(sambashare)

If "dialout" or whatever you see as the group for the USB serial device doesn't appear in that list then you need to add yourself to the users in that group. I'm afraid I don't know what the usual way to do that is these days. You can probably do it through some GUI interface for user management or something. As I'm a rather old-school UNIX user I'd switch to the root user and edit /etc/group directly, adding myself to the appropriate group:

dialout:x:20:james

It may be that you're already in the dialout group. I run Mint on my desktop, laptop and Aspire One and it looks correct for me, but I can't recall whether I fixed it when I installed the OS or if it was done automatically.

If you're not logged in as root btw, probably the quickest way to see what the kernel has done when you plug in a USB device is something like:

$ dmesg | tail -10

James

Link to comment
Share on other sites

From my limited research this weekend, the Linux platform uses a standard similar to how ASCOM is the standard on the windows platform. it's called INDI. As it defaults to a serial com port when searching for the telescope I thought it should work with the FTDI chip. Maybe it's not compatible with this type of cable and needs some dedicated hardware ?

Link to comment
Share on other sites

James thanks for the detailed post. No I wasn't aware that linux was case sensitive... I'll give your suggestions a try later.

I've also downloaded Kstars for windows, along with wINDI but have issues with that - but that's another story

Link to comment
Share on other sites

I've not tried any of this stuff from Linux yet myself, I have to admit. Where there are options to connect Linux planetarium applications to a mount I really don't know what they're expecting.

James

Link to comment
Share on other sites

I looked at the different scopes supported by indi and from what I can see it needs the handset in place. Thus the EQDIR adapter will not work as EQMOD talks to the internals of the mount, not the handset. I would simply connect the computer's com port (or USB to serial) to the supplied computer connect cable that goes to the handset.

/p

Link to comment
Share on other sites

Mint is the Linus distro I run - I found it better after Ubuntu took what I considered a wrong turn. With the majority of software designed for windoze, and after researching and trying some of the Linux astro apps I concluded that it would take a lot of effort over and above the norm to get things working on the Linux platform. So reluctantly I bought myself a couple of Windows machines for my astro work. Have to say I really miss the rock solid reliability of Linux and while W7 is better than XP it still needs rebooting every time I want to do astro imaging and it's apps just stop working at times for no apparent reason. I feel I'm giving in a bit but it's hard enough getting things working without making it even worse. I might have managed it when I was younger with more working "little grey cells" :D I currently run all three main platforms. Linux Mint for a lot of my design work, W7 for astro and Mac OSX on my MacBook Pro laptop.

Link to comment
Share on other sites

I've been a user, programmer and systems admin of UNIX and UNIX-like operating systems for not far short of thirty years now (I've been working with Linux for almost twenty years, which is a bit of a shock). It's quite a disappointment that there isn't a wider range of astro software available for it. To a certain extent I understand why, but I think some of the reasons are based on flawed assumptions. One of the most bizarre seems to be that manufacturers don't want to reveal their proprietary user interfaces and control systems, which is utterly ridiculous when you consider the huge number of things that have been reverse engineered now. If people can reverse engineer a filesystem implementation or create something like Wine from scratch, a piddly little camera or filter wheel interface is not going to present that much of a problem.

James

  • Like 1
Link to comment
Share on other sites

James, this has what's always put me off running various linux distros over the years. Agreed nowadays there are more applications available, and for me I love the look and feel of Kstars, but then everything is let down because the simple thing of connecting it to a "supported" mount becomes difficult because of the way the interface platform is designed. But what really gets me is when the developers feel they have a decent application that they port it to MS Windows, they then fail to use what has now become the standard interface platform (ASCOM) on that OS...

There has always been those who have been hardened Linux supporters, who say that linux will one day be just as, if not more popular than windows, but as things like this are still as difficult as they used to be, I personally can never see it happening. Don't get me wrong, windows still has its faults (being in IT support for 15 years I know how flaky it can be), but things appear to be a lot easier to sort out on windows than on Mint !

Link to comment
Share on other sites

Ubuntu seemed to increase the interest in Linux for a while but it didn't seem to last. OTOH the Mac (with another Unix based OS) seems to continue to have a steady following but more in the professional scene than generally AFAICT. Of course the price puts people off the Mac for household use but IMV you get what you pay for and Mac production is superb.

Link to comment
Share on other sites

Ubuntu seemed to increase the interest in Linux for a while but it didn't seem to last. OTOH the Mac (with another Unix based OS) seems to continue to have a steady following but more in the professional scene than generally AFAICT. Of course the price puts people off the Mac for household use but IMV you get what you pay for and Mac production is superb.

I've not used the Mac-style UNIX for many years, but I find myself becoming increasingly frustrated with Apple products. Some of them make me feel like one of the main design criteria was "never risk distressing the user by telling them something went wrong". Or if they do tell you something went wrong, the error message is completely unhelpful in terms of solving the problem. As I know that in many cases I could solve the problem if I was just given enough information to understand what it was that really drives me nuts. Sometimes I wonder if they've adopted a philosophy that the user doesn't want to understand or fix things. If it doesn't work they just want to take it to someone else and have it sorted out for them. That may be maligning Apple, but it's how it feels to me.

I have wondered if Ubuntu isn't just trying to jump on the "tablet" bandwagon. With Unity it really feels like that's where they think they should be going. Unfortunately whilst tablets are good at some things, they're utterly awful at others and if you happen to be one of the people who wants to do some of those other things you're very much the poor cousin now.

James

Link to comment
Share on other sites

I give Linux a chance every year, and so far I have had to abondon it within days as the software that run on it is just not stable enough. Take an example; GIS software. I installed Edubuntu, Mint and Ubuntu and added all the GIS software and toolboxes (Quantum GIS). Spend the better part of a day trying to work out how to reach my spatial data in the database server and finally get it to read some of it. Refresh the layer and only half the data set is returned. Add a layer that points to a file that has been deleted: QGis hangs. The list goes on and on. Installed Quantum GIS on Windows and found about the same "stability". My take is that the Open Source validation process is close to non-existant. Imagine, they have an accepted issue closure called "Works here", by which they can get rid of a bug issue on the grounds that it works on their machine and thus is fully functional. Open Source is NEVER going to be fully stable until we all can pitch in and correct our own bugs in the source code, and that is never going to happen.

I'll stick with commercial products that have been tested.

I do agree that the Mac OS has a very poor way of handling problems, always leaving the user with a confused face while he recieves the usual support answer from Apple: "Reinstall the operating system".

As for Windows 7 reboot needs... I never do it unless there is an update that requires it. Not my astro box out on the balcony, not my workstation, not any of my laptops (of which one is a MB Pro running Win 7 natively). I sometimes find them having an uptime of several months.

Astro. Well, we are stuck with Windows there. I have been through the Mac run and have abandoned it. There simply is no software good enough. It is different when you're in a university environment and work a large telescope and have programming skills. They do Linux almost exclusively, but then the software isn't really what we would call "user usable".

Rant, rant ;)

/p

Link to comment
Share on other sites

I think you're probably right with much of that James. I think Macs are generally aimed at corporate users who prefer to employ someone else to fix things if they go wrong. Guess I've probably been lucky with my MacBook - it's served me well for several years now and I've only had to reboot it a couple of times since new.

Link to comment
Share on other sites

I give Linux a chance every year, and so far I have had to abondon it within days as the software that run on it is just not stable enough. Take an example; GIS software. I installed Edubuntu, Mint and Ubuntu and added all the GIS software and toolboxes (Quantum GIS). Spend the better part of a day trying to work out how to reach my spatial data in the database server and finally get it to read some of it. Refresh the layer and only half the data set is returned. Add a layer that points to a file that has been deleted: QGis hangs. The list goes on and on. Installed Quantum GIS on Windows and found about the same "stability". My take is that the Open Source validation process is close to non-existant. Imagine, they have an accepted issue closure called "Works here", by which they can get rid of a bug issue on the grounds that it works on their machine and thus is fully functional.

So you take the example of a single package and tar all open source software with the same brush?

Yes, there are open source projects where software is inadequately or incompletely tested. The same goes for commercial software. Often the difference is that the authors of commercial software are getting paid to do proper testing and don't, whereas the authors of open source software aren't getting paid at all.

As a counter-example to your GIS experience, the entire Linux kernel is hundreds of thousands (probably millions by now) of lines of code contributed by a huge number of people. It's all open source and it's astonishingly stable, to the point where it can run for years on end without a problem (I've had servers that have run four or five years without a reboot, never mind months). The same goes for the compiler used to build it all. And the windowing system (because it's not actually part of the OS) or Apache webserver and the various database servers that people often run on it. Some of the test suites for open source packages are easily comparable in scope to those I've needed to run when working in a commercial software development environment.

Lack of stability is no more a valid criticism of open source systems than it is of commercial ones and Microsoft or Apple are just as guilty as anyone else. There are companies who provide support for open source software too, so it's far from always the case that if you can't fix a problem yourself then you're up a certain smelly waterway with no visible means of propulsion.

Now, if you'd picked poor documentation as a criticism of open source I might at one time have agreed with you, but these days many of the commercial software vendors don't exactly seem too keen on writing documentation either.

There are reasons that Linux may be unlikely to replace other operating systems on the desktop, but they're often more related to software vendors not wanting to support anything more than the platforms that are likely to generate them a given level of income, or believing that if they stick with commercial operating systems then they can keep control of their IP, or they want to indulge in a bit of vendor lock-in to "leverage" (ick) the platform to sell more of their own software, or that users don't want the retraining costs. They're commercial reasons, not technical ones.

On the other hand, commercial software vendors are finding it tough to maintain their profits these days (perhaps one of the reasons why Adobe and Microsoft amongst others appear favour moving to a subscription model) whilst their users don't want to spend so much or are happy not to pay for "upgrades" they don't actually need. Perhaps change is already coming...

James

Link to comment
Share on other sites

Mainstream software is alright on Linux but I think some specialist software really needs to be used by programmers who are capable of installing from source code and it helps if they can also debug that source code. For the software enthusiast/expert, and where manufacturers are prepared to release their interface specifications there is little problem in writing the required drivers and handlers. Plus in many cases the manufacturers provide drivers. Unfortunately, the makers of astro kit seem very reluctant to release any data about their products and there don't seem to be many people able to reverse engineer this kit. I have reverse engineered some hardware in the past and produced Linux software to talk to it.

All this makes me feel old - a few decades ago I doubt I'd have have had much trouble reverse engineering and writing software for such things as mounts or even cameras. My career was in computers and interfacing with electronic equipment - hardware of various sorts and with numerous different operating systems and programming languages. Now my poor old brain just won't cope with it. All I seem to manage now is relatively simple C code in Arduino sketches.

Edited by Gina
Link to comment
Share on other sites

I would also like to add, there is a bit difference between the Kernel and random applications, or general Open source software, but often people call it Linux collectively, which is a bit unfortunate and confusing and sometimes an unfair criticism on Linux itself.

Of course in the latter case quality can vary, just as with windows or other commercial offerings, but there is no denying that the Kernel and base OS is a highly stable piece of software, very much fit for purpose in a commercial environment, In fact I would say in many cases it is a better option, depending on your needs.

Sure every platform has a list of bad and good apps as well, but I can also give good counter examples where the Linux version runs better than windows. As it is now I run a fairy bleeding edge distro at home in Fedora Core after some years not using Linux at all. Fedora gets updates by the hundreds every few weeks, runs a kernel that literally would have been compiled the week before, and I've yet to ever have a problem, the raw OS runs great without issue.

On the desktop sometimes an app crash here and there, but also finding KDE to be very stable in my case. That being said I am a tech level user so if something does happen, I am can sort it out and not afraid of that.

I am not convinced that Linux will ever make significant inroads on the desktop for everyday use for the masses on a PC, but I am completely comfortable/happy using it that way myself. I use both windows and Linux roughly equally these days, depends what I need/want to do. I have a dual boot config setup, windows on one disk, Linux on the other, both get used and it works great for me :)

Link to comment
Share on other sites

Mainstream software is alright on Linux but I think some specialist software really needs to be used by programmers who are capable of installing from source code and it helps if they can also debug that source code. For the software enthusiast/expert, and where manufacturers are prepared to release their interface specifications there is little problem in writing the required drivers and handlers. Plus in many cases the manufacturers provide drivers. Unfortunately, the makers of astro kit seem very reluctant to release any data about their products and there don't seem to be many people able to reverse engineer this kit. I have reverse engineered some hardware in the past and produced Linux software to talk to it.

I think part of the problem with reverse-engineering astro kit is that there just aren't enough people who are into astronomy and also fairly hardcore hackers. I'd hazard a guess that of all the regular SGL members for instance, there are not too many who would be happy grubbing around at the device and kernel level in an operating system and even fewer who have the time and inclination to do so once they've finished earning a living and doing family stuff. I'd guess it's largely the same for any hobby where computers are used but aren't central to it.

I'd actually love to spend time doing so. I started reverse-engineering commercial software for the BBC Micro in my mid-teens and found it a real challenge. I spent a fair bit of time cracking copy protection systems for games and suchlike. Didn't play more than a few very much, but unpicking what someone else had done was fascinating. Now if I just didn't have to earn a living... :D

James

Link to comment
Share on other sites

To the Op I was also going to recommend to try the following in addition to what James said. As a test try logging in as root and running it that way. not sure how Mint is setup however. Not sure if Mint comes with SE Linux if that may be an issue, very little info to go by from what you said. Logged in as root, Open a terminal widow and at the command line type

tail -f /var/log/messages

this will show in real time any messages or output, then launch the app and see what gets logged. I am sure the culprit of this can be found if you are willing to spend some time on it, not saying there is an easy solution but it can be looked into a bit more to see if you are at a dead end. Contact or ask your questions at various linux forums would be your other option. The desktop also comes with various GUI based logging facilities for apps to see what is going on, but I can't really give exact advice there, as each distro tends to be a bit different. In cases such as this Google is really you friend as they say :)

Edited by AlexB67
Link to comment
Share on other sites

I agree that the kernel is remarkably stable. No hard words about that from my side. I do wonder, however, what will happen to said kernel when Linus himself stops OK:ing all changes...

The open source process has a problem whenever "critical mass" for an application is not achieved. When not enough programmers or organisations are interested in working on the project.

I am currently involved in two fairly large software packages that need to be available on miltiple platforms. We chose Qt as the base because that is the only viable option out there. Now, writing multi platform is difficult and we realize that. We are, however, having some real issues with stability in the developer tools. These (Qt) tools are very widely used in creating stuff like Stellarium, Nebulosity and many other great applications. Coupled with the OSS Git project (for source code control) things get worse. I know that lots of people use Git and that it does the job, but the stuff is not exactly what I would call stable. Compare this to a commercial developer platform that is being developed by a serious number of programmers and gets thoroughly tested before anything goes out the door (and where every programmer is on a payroll and needs to perform) and the differences are vast. Having grown up on DEC systems since the 70's, I am of the view that the best platform for development today is Visual Studio coupled with Team Foundation Server. But that is Windows only.

My experience with OSS is not based on a "single package experience", but rather on multiple packages - all pointing in the same direction: the stuff is good when a sufficient number of programmers are on the team and, preferably, when a commercial actor is donating programmer time to meet their own needs.

I am sure that OSS works for many, but I have found serious problems with many "packages", among them:

- Grass (US Military Geo software - GIS)

- Quantum GIS

- Qt

- MySQL (when queries get complex with multiple joins, it just faults)

- Meshlab

Still, all of these are used extensively and should be OK, but they still hang, lose data or plainly do the wrong thing.

Confused... ;)

/per

Link to comment
Share on other sites

I'm not familiar with three of the packages you mention, but Qt has always looked interesting from a cross-platform GUI development point of view. It seems to be taking its time to settle down though. It's only the last year or so that people have been migrating some applications to v4 and now v5.1 is out. It may well be a fair criticism of some open source packages that they start off intended to solve a specific small problem set and in their subsequent expansion to a more general arena they struggle to iron out some of the kinks that weren't initially an issue. Off the top of my head I'd suggest the same sort of thing has happened with PHP, Python and Ruby, for example.

It may also be that the currently-favoured development methodologies engender that sort of problem. I've worked on a number of rapid-development projects where we've been moving along quite happily and then someone has decided "the future direction will be this" and the developers have looked at each other and said "Well, if we'd known that to start with we'd have done it a bit differently".

As regards MySQL, at one time I think if I had a dependency on it and was having trouble with very complex queries I'd pay for a support contract. I don't know how that works now Oracle are in charge though. I guess you have to bend over and assume the position if you want support. Someone has to pay for Larry's new yacht after all.

James

Link to comment
Share on other sites

To the Op I was also going to recommend to try the following in addition to what James said. As a test try logging in as root and running it that way.

I get a message that I can't use the log in window to log in with the root account - the user account that has been set up is in the administrator group, so I assume would have the same rights ?

Link to comment
Share on other sites

I get a message that I can't use the log in window to log in with the root account - the user account that has been set up is in the administrator group, so I assume would have the same rights ?

Not quite. root is the only administrator. Being a member of the administrator group allows you to assume root's privileges when necessary, but you don't have them all the time because it would be a fairly large security hole.

If you want to run a GUI application that requires root privileges it will usually ask you to enter your password to get root privileges when it starts. If you want to run a command line program you can open a terminal window and do either

$ sudo command

which will run just command as root, or if you need to do several things:

$ sudo -s

which will give you a shell running as the root user (and the prompt will change to use '#' rather than '$' to show that you have root privileges in that shell).

James

Link to comment
Share on other sites

I get a message that I can't use the log in window to log in with the root account - the user account that has been set up is in the administrator group, so I assume would have the same rights ?

I think that it has enough rights, but I cannot guarantee it not having use Mint , I noticed Ubuntu or another distro had no traditional root account anymore either, seems to have gone more the like windows way with and admin account with access rights. Try it out :)

edit: James beat me to it. :)

as James said anyway you can run anything with the sudo command and also to tail the system log file when at the desktop as a normal user using that method.

Edited by AlexB67
Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.