3 This FAQ is published on the PuTTY web site, and also provided as an
4 appendix in the manual.
6 \H{faq-intro} Introduction
8 \S{faq-what}{Question} What is PuTTY?
10 PuTTY is a client program for the SSH, Telnet and Rlogin network
13 These protocols are all used to run a remote session on a computer,
14 over a network. PuTTY implements the client end of that session: the
15 end at which the session is displayed, rather than the end at which
18 In really simple terms: you run PuTTY on a Windows machine, and tell
19 it to connect to (for example) a Unix machine. PuTTY opens a window.
20 Then, anything you type into that window is sent straight to the
21 Unix machine, and everything the Unix machine sends back is
22 displayed in the window. So you can work on the Unix machine as if
23 you were sitting at its console, while actually sitting somewhere
26 \H{faq-support} Features supported in PuTTY
28 \I{supported features}In general, if you want to know if PuTTY supports
29 a particular feature, you should look for it on the
30 \W{http://www.chiark.greenend.org.uk/~sgtatham/putty/}{PuTTY web site}.
34 \W{http://www.chiark.greenend.org.uk/~sgtatham/putty/changes.html}{changes
35 page}, and see if you can find the feature on there. If a feature is
36 listed there, it's been implemented. If it's listed as a change made
37 \e{since} the latest version, it should be available in the
38 development snapshots, in which case testing will be very welcome.
41 \W{http://www.chiark.greenend.org.uk/~sgtatham/putty/wishlist/}{Wishlist
42 page}, and see if you can find the feature there. If it's on there,
43 and not in the \q{Recently fixed} section, it probably \e{hasn't} been
46 \S{faq-ssh2}{Question} Does PuTTY support SSH-2?
48 Yes. SSH-2 support has been available in PuTTY since version 0.50.
50 Public key authentication (both RSA and DSA) in SSH-2 is new in
53 \S{faq-ssh2-keyfmt}{Question} Does PuTTY support reading OpenSSH or
54 \cw{ssh.com} SSH-2 private key files?
56 PuTTY doesn't support this natively (see
57 \W{http://www.chiark.greenend.org.uk/~sgtatham/putty/wishlist/key-formats-natively.html}{the wishlist entry}
58 for reasons why not), but as of 0.53
59 PuTTYgen can convert both OpenSSH and \cw{ssh.com} private key
60 files into PuTTY's format.
62 \S{faq-ssh1}{Question} Does PuTTY support SSH-1?
64 Yes. SSH-1 support has always been available in PuTTY.
66 However, the SSH-1 protocol has many weaknesses and is no longer
67 considered secure; you should use SSH-2 instead if at all possible.
69 \#{XXX-REVIEW-BEFORE-RELEASE:
70 As of 0.68, PuTTY will no longer fall back to SSH-1 if the server
71 doesn't appear to support SSH-2; you must explicitly ask for SSH-1. }
73 \S{faq-localecho}{Question} Does PuTTY support \i{local echo}?
75 Yes. Version 0.52 has proper support for local echo.
77 In version 0.51 and before, local echo could not be separated from
78 local line editing (where you type a line of text locally, and it is
79 not sent to the server until you press Return, so you have the
80 chance to edit it and correct mistakes \e{before} the server sees
81 it). New in version 0.52, local echo and local line editing are
82 separate options, and by default PuTTY will try to determine
83 automatically whether to enable them or not, based on which protocol
84 you have selected and also based on hints from the server. If you
85 have a problem with PuTTY's default choice, you can force each
86 option to be enabled or disabled as you choose. The controls are in
87 the Terminal panel, in the section marked \q{Line discipline
90 \S{faq-savedsettings}{Question} Does PuTTY support storing settings,
91 so I don't have to change them every time?
93 Yes, all of PuTTY's settings can be saved in named session profiles.
94 You can also change the default settings that are used for new sessions.
95 See \k{config-saving} in the documentation for how to do this.
97 \S{faq-disksettings}{Question} Does PuTTY support storing its
98 settings in a disk file?
100 Not at present, although \k{config-file} in the documentation gives
101 a method of achieving the same effect.
103 \S{faq-fullscreen}{Question} Does PuTTY support full-screen mode,
106 Yes; this is a new feature in version 0.52.
108 \S{faq-password-remember}{Question} Does PuTTY have the ability to
109 \i{remember my password} so I don't have to type it every time?
113 Remembering your password is a bad plan for obvious security
114 reasons: anyone who gains access to your machine while you're away
115 from your desk can find out the remembered password, and use it,
116 abuse it or change it.
118 In addition, it's not even \e{possible} for PuTTY to automatically
119 send your password in a Telnet session, because Telnet doesn't give
120 the client software any indication of which part of the login
121 process is the password prompt. PuTTY would have to guess, by
122 looking for words like \q{password} in the session data; and if your
123 login program is written in something other than English, this won't
126 In SSH, remembering your password would be possible in theory, but
127 there doesn't seem to be much point since SSH supports public key
128 authentication, which is more flexible and more secure. See
129 \k{pubkey} in the documentation for a full discussion of public key
132 \S{faq-hostkeys}{Question} Is there an option to turn off the
133 \I{verifying the host key}annoying host key prompts?
135 No, there isn't. And there won't be. Even if you write it yourself
136 and send us the patch, we won't accept it.
138 Those annoying host key prompts are the \e{whole point} of SSH.
139 Without them, all the cryptographic technology SSH uses to secure
140 your session is doing nothing more than making an attacker's job
141 slightly harder; instead of sitting between you and the server with
142 a packet sniffer, the attacker must actually subvert a router and
143 start modifying the packets going back and forth. But that's not all
144 that much harder than just sniffing; and without host key checking,
145 it will go completely undetected by client or server.
147 Host key checking is your guarantee that the encryption you put on
148 your data at the client end is the \e{same} encryption taken off the
149 data at the server end; it's your guarantee that it hasn't been
150 removed and replaced somewhere on the way. Host key checking makes
151 the attacker's job \e{astronomically} hard, compared to packet
152 sniffing, and even compared to subverting a router. Instead of
153 applying a little intelligence and keeping an eye on Bugtraq, the
154 attacker must now perform a brute-force attack against at least one
155 military-strength cipher. That insignificant host key prompt really
156 does make \e{that} much difference.
158 If you're having a specific problem with host key checking - perhaps
159 you want an automated batch job to make use of PSCP or Plink, and the
160 interactive host key prompt is hanging the batch process - then the
161 right way to fix it is to add the correct host key to the Registry in
162 advance, or if the Registry is not available, to use the \cw{-hostkey}
163 command-line option. That way, you retain the \e{important} feature of
164 host key checking: the right key will be accepted and the wrong ones
165 will not. Adding an option to turn host key checking off completely is
166 the wrong solution and we will not do it.
168 If you have host keys available in the common \i\c{known_hosts} format,
169 we have a script called
170 \W{https://git.tartarus.org/?p=simon/putty.git;a=blob;f=contrib/kh2reg.py;hb=HEAD}\c{kh2reg.py}
171 to convert them to a Windows .REG file, which can be installed ahead of
172 time by double-clicking or using \c{REGEDIT}.
174 \S{faq-server}{Question} Will you write an SSH server for the PuTTY
175 suite, to go with the client?
177 No. The only reason we might want to would be if we could easily
178 re-use existing code and significantly cut down the effort. We don't
179 believe this is the case; there just isn't enough common ground
180 between an SSH client and server to make it worthwhile.
182 If someone else wants to use bits of PuTTY in the process of writing
183 a Windows SSH server, they'd be perfectly welcome to of course, but
184 I really can't see it being a lot less effort for us to do that than
185 it would be for us to write a server from the ground up. We don't
186 have time, and we don't have motivation. The code is available if
187 anyone else wants to try it.
189 \S{faq-pscp-ascii}{Question} Can PSCP or PSFTP transfer files in
194 Until recently, this was a limitation of the file transfer protocols:
195 the SCP and SFTP protocols had no notion of transferring a file in
196 anything other than binary mode. (This is still true of SCP.)
198 The current draft protocol spec of SFTP proposes a means of
199 implementing ASCII transfer. At some point PSCP/PSFTP may implement
202 \H{faq-ports} Ports to other operating systems
204 The eventual goal is for PuTTY to be a multi-platform program, able
205 to run on at least Windows, Mac OS and Unix.
207 Porting will become easier once PuTTY has a generalised porting
208 layer, drawing a clear line between platform-dependent and
209 platform-independent code. The general intention was for this
210 porting layer to evolve naturally as part of the process of doing
211 the first port; a Unix port has now been released and the plan
212 seems to be working so far.
214 \S{faq-ports-general}{Question} What ports of PuTTY exist?
216 Currently, release versions of PuTTY tools only run on Windows
219 \#{XXX-REVIEW-BEFORE-RELEASE: replace following two lines with:
220 As of 0.68, the supplied PuTTY executables run on versions of
221 Windows from XP onwards,}
222 PuTTY runs on versions of Windows from Windows 95 onwards (but not
223 the 16-bit Windows 3.1; see \k{faq-win31}),
224 up to and including Windows 10; and we know of no reason why PuTTY
225 should not continue to work on future versions of Windows.
227 The 32-bit Windows executables we provide for the \q{\i{x86}}
228 processor architecture should also work fine on 64-bit processors
229 that are backward-compatible with that architecture.
230 \#{XXX-REVIEW-BEFORE-RELEASE: The 64-bit executables will only
231 work on 64-bit versions of Windows. They will run somewhat faster
232 than 32-bit executables would on the same processor, but will
233 consume slightly more memory.}
235 (We used to also provide executables for Windows for the Alpha
236 processor, but stopped after 0.58 due to lack of interest.)
238 In the development code, a partial port to Mac OS exists (see
241 Currently PuTTY does \e{not} run on Windows CE (see \k{faq-wince}).
243 We do not have release-quality ports for any other systems at the
244 present time. If anyone told you we had an Android port, or an iOS
245 port, or any other port of PuTTY, they were mistaken. We don't.
247 There are some third-party ports to various platforms, mentioned
249 \W{http://www.chiark.greenend.org.uk/~sgtatham/putty/links.html}{Links page of our website}.
251 \S{faq-unix}{Question} \I{Unix version}Is there a port to Unix?
253 As of 0.54, there are Unix ports of most of the traditional PuTTY
254 tools, and also one entirely new application.
256 If you look at the source release, you should find a \c{unix}
257 subdirectory. There are a couple of ways of building it,
258 including the usual \c{configure}/\c{make}; see the file \c{README}
259 in the source distribution. This should build you Unix
260 ports of Plink, PuTTY itself, PuTTYgen, PSCP, PSFTP, and also
261 \i\c{pterm} - an \cw{xterm}-type program which supports the same
262 terminal emulation as PuTTY. \#{XXX-REVIEW-BEFORE-RELEASE:}
263 We do not yet have a Unix port of Pageant.
265 If you don't have \i{Gtk}, you should still be able to build the
268 \S{faq-unix-why}{Question} What's the point of the Unix port? Unix
271 All sorts of little things. \c{pterm} is directly useful to anyone
272 who prefers PuTTY's terminal emulation to \c{xterm}'s, which at
273 least some people do. Unix Plink has apparently found a niche among
274 people who find the complexity of OpenSSL makes OpenSSH hard to
275 install (and who don't mind Plink not having as many features). Some
276 users want to generate a large number of SSH keys on Unix and then
277 copy them all into PuTTY, and the Unix PuTTYgen should allow them to
278 automate that conversion process.
280 There were development advantages as well; porting PuTTY to Unix was
281 a valuable path-finding effort for other future ports, and also
282 allowed us to use the excellent Linux tool
283 \W{http://valgrind.kde.org/}{Valgrind} to help with debugging, which
284 has already improved PuTTY's stability on \e{all} platforms.
286 However, if you're a Unix user and you can see no reason to switch
287 from OpenSSH to PuTTY/Plink, then you're probably right. We don't
288 expect our Unix port to be the right thing for everybody.
290 \S{faq-wince}{Question} Will there be a port to Windows CE or PocketPC?
292 We once did some work on such a port, but it only reached an early
293 stage, and certainly not a useful one. It's no longer being actively
296 \S{faq-win31}{Question} Is there a port to \i{Windows 3.1}?
298 PuTTY is a 32-bit application from the ground up, so it won't run on
299 Windows 3.1 as a native 16-bit program; and it would be \e{very}
300 hard to port it to do so, because of Windows 3.1's vile memory
301 allocation mechanisms.
303 However, it is possible in theory to compile the existing PuTTY
304 source in such a way that it will run under \i{Win32s} (an extension to
305 Windows 3.1 to let you run 32-bit programs). In order to do this
306 you'll need the right kind of C compiler - modern versions of Visual
307 C at least have stopped being backwards compatible to Win32s. Also,
308 the last time we tried this it didn't work very well.
310 \S{faq-mac-port}{Question} Will there be a port to the \I{Mac OS}Mac?
314 We attempted one around 2005, written as a native Cocoa application,
315 but it turned out to be very slow to redraw its window for some reason
316 we never got to the bottom of.
318 In 2015, after porting the GTK front end to work with GTK 3, we began
319 another attempt based on making small changes to the GTK code and
320 building it against the OS X Quartz version of GTK 3. This doesn't
321 seem to have the window redrawing problem any more, so it's already
322 got further than the last effort, but it is still substantially
325 If any OS X and/or GTK programming experts are keen to have a finished
326 version of this, we urge them to help out with some of the remaining
329 \S{faq-epoc}{Question} Will there be a port to EPOC?
331 I hope so, but given that ports aren't really progressing very fast
332 even on systems the developers \e{do} already know how to program
333 for, it might be a long time before any of us get round to learning
334 a new system and doing the port for that.
336 However, some of the work has been done by other people; see the
337 \W{http://www.chiark.greenend.org.uk/~sgtatham/putty/links.html}{Links page of our website}
338 for various third-party ports.
340 \S{faq-iphone}{Question} Will there be a port to the iPhone?
342 We have no plans to write such a port ourselves; none of us has an
343 iPhone, and developing and publishing applications for it looks
344 awkward and expensive.
346 However, there is a third-party SSH client for the iPhone and
347 iPod\_Touch called \W{http://www.instantcocoa.com/products/pTerm/}{pTerm},
348 which is apparently based on PuTTY. (This is nothing to do with our
349 similarly-named \c{pterm}, which is a standalone terminal emulator for
350 Unix systems; see \k{faq-unix}.)
352 \H{faq-embedding} Embedding PuTTY in other programs
354 \S{faq-dll}{Question} Is the SSH or Telnet code available as a DLL?
356 No, it isn't. It would take a reasonable amount of rewriting for
357 this to be possible, and since the PuTTY project itself doesn't
358 believe in DLLs (they make installation more error-prone) none of us
359 has taken the time to do it.
361 Most of the code cleanup work would be a good thing to happen in
362 general, so if anyone feels like helping, we wouldn't say no.
365 \W{http://www.chiark.greenend.org.uk/~sgtatham/putty/wishlist/dll-frontend.html}{the wishlist entry}.
367 \S{faq-vb}{Question} Is the SSH or Telnet code available as a Visual
370 No, it isn't. None of the PuTTY team uses Visual Basic, and none of
371 us has any particular need to make SSH connections from a Visual
372 Basic application. In addition, all the preliminary work to turn it
373 into a DLL would be necessary first; and furthermore, we don't even
374 know how to write VB components.
376 If someone offers to do some of this work for us, we might consider
377 it, but unless that happens I can't see VB integration being
378 anywhere other than the very bottom of our priority list.
380 \S{faq-ipc}{Question} How can I use PuTTY to make an SSH connection
381 from within another program?
383 Probably your best bet is to use Plink, the command-line connection
384 tool. If you can start Plink as a second Windows process, and
385 arrange for your primary process to be able to send data to the
386 Plink process, and receive data from it, through pipes, then you
387 should be able to make SSH connections from your program.
389 This is what CVS for Windows does, for example.
391 \H{faq-details} Details of PuTTY's operation
393 \S{faq-term}{Question} What \i{terminal type} does PuTTY use?
395 For most purposes, PuTTY can be considered to be an \cw{xterm}
398 PuTTY also supports some terminal \i{control sequences} not supported by
399 the real \cw{xterm}: notably the Linux console sequences that
400 reconfigure the colour palette, and the title bar control sequences
401 used by \i\cw{DECterm} (which are different from the \cw{xterm} ones;
402 PuTTY supports both).
404 By default, PuTTY announces its terminal type to the server as
405 \c{xterm}. If you have a problem with this, you can reconfigure it
406 to say something else; \c{vt220} might help if you have trouble.
408 \S{faq-settings}{Question} Where does PuTTY store its data?
410 On Windows, PuTTY stores most of its data (saved sessions, SSH host
411 keys) in the \i{Registry}. The precise location is
413 \c HKEY_CURRENT_USER\Software\SimonTatham\PuTTY
415 and within that area, saved sessions are stored under \c{Sessions}
416 while host keys are stored under \c{SshHostKeys}.
418 PuTTY also requires a random number seed file, to improve the
419 unpredictability of randomly chosen data needed as part of the SSH
420 cryptography. This is stored by default in a file called \i\c{PUTTY.RND};
421 this is stored by default in the \q{Application Data} directory,
422 or failing that, one of a number of fallback locations. If you
423 want to change the location of the random number seed file, you can
424 put your chosen pathname in the Registry, at
426 \c HKEY_CURRENT_USER\Software\SimonTatham\PuTTY\RandSeedFile
428 You can ask PuTTY to delete all this data; see \k{faq-cleanup}.
430 On Unix, PuTTY stores all of this data in a directory \cw{~/.putty}.
432 \H{faq-howto} HOWTO questions
434 \S{faq-login}{Question} What login name / password should I use?
436 This is not a question you should be asking \e{us}.
438 PuTTY is a communications tool, for making connections to other
439 computers. We maintain the tool; we \e{don't} administer any computers
440 that you're likely to be able to use, in the same way that the people
441 who make web browsers aren't responsible for most of the content you can
442 view in them. \#{FIXME: less technical analogy?} We cannot help with
443 questions of this sort.
445 If you know the name of the computer you want to connect to, but don't
446 know what login name or password to use, you should talk to whoever
447 administers that computer. If you don't know who that is, see the next
448 question for some possible ways to find out.
450 \# FIXME: some people ask us to provide them with a login name
451 apparently as random members of the public rather than in the
452 belief that we run a server belonging to an organisation they already
453 have some relationship with. Not sure what to say to such people.
455 \S{faq-commands}{Question} \I{commands on the server}What commands
456 can I type into my PuTTY terminal window?
458 Again, this is not a question you should be asking \e{us}. You need
459 to read the manuals, or ask the administrator, of \e{the computer
460 you have connected to}.
462 PuTTY does not process the commands you type into it. It's only a
463 communications tool. It makes a connection to another computer; it
464 passes the commands you type to that other computer; and it passes
465 the other computer's responses back to you. Therefore, the precise
466 range of commands you can use will not depend on PuTTY, but on what
467 kind of computer you have connected to and what software is running
468 on it. The PuTTY team cannot help you with that.
470 (Think of PuTTY as being a bit like a telephone. If you phone
471 somebody up and you don't know what language to speak to make them
472 understand you, it isn't \e{the telephone company}'s job to find
473 that out for you. We just provide the means for you to get in touch;
474 making yourself understood is somebody else's problem.)
476 If you are unsure of where to start looking for the administrator of
477 your server, a good place to start might be to remember how you
478 found out the host name in the PuTTY configuration. If you were
479 given that host name by e-mail, for example, you could try asking
480 the person who sent you that e-mail. If your company's IT department
481 provided you with ready-made PuTTY saved sessions, then that IT
482 department can probably also tell you something about what commands
483 you can type during those sessions. But the PuTTY maintainer team
484 does not administer any server you are likely to be connecting to,
485 and cannot help you with questions of this type.
487 \S{faq-startmax}{Question} How can I make PuTTY start up \i{maximise}d?
489 Create a Windows shortcut to start PuTTY from, and set it as \q{Run
492 \S{faq-startsess}{Question} How can I create a \i{Windows shortcut} to
493 start a particular saved session directly?
495 To run a PuTTY session saved under the name \q{\cw{mysession}},
496 create a Windows shortcut that invokes PuTTY with a command line
499 \c \path\name\to\putty.exe -load "mysession"
501 (Note: prior to 0.53, the syntax was \c{@session}. This is now
502 deprecated and may be removed at some point.)
504 \S{faq-startssh}{Question} How can I start an SSH session straight
505 from the command line?
507 Use the command line \c{putty -ssh host.name}. Alternatively, create
508 a saved session that specifies the SSH protocol, and start the saved
509 session as shown in \k{faq-startsess}.
511 \S{faq-cutpaste}{Question} How do I \i{copy and paste} between PuTTY and
512 other Windows applications?
514 Copy and paste works similarly to the X Window System. You use the
515 left mouse button to select text in the PuTTY window. The act of
516 selection \e{automatically} copies the text to the clipboard: there
517 is no need to press Ctrl-Ins or Ctrl-C or anything else. In fact,
518 pressing Ctrl-C will send a Ctrl-C character to the other end of
519 your connection (just like it does the rest of the time), which may
520 have unpleasant effects. The \e{only} thing you need to do, to copy
521 text to the clipboard, is to select it.
523 To paste the clipboard contents into a PuTTY window, by default you
524 click the right mouse button. If you have a three-button mouse and
525 are used to X applications, you can configure pasting to be done by
526 the middle button instead, but this is not the default because most
527 Windows users don't have a middle button at all.
529 You can also paste by pressing Shift-Ins.
531 \S{faq-options}{Question} How do I use all PuTTY's features (public
532 keys, proxying, cipher selection, etc.) in PSCP, PSFTP and Plink?
534 Most major features (e.g., public keys, port forwarding) are available
535 through command line options. See the documentation.
537 Not all features are accessible from the command line yet, although
538 we'd like to fix this. In the meantime, you can use most of
539 PuTTY's features if you create a PuTTY saved session, and then use
540 the name of the saved session on the command line in place of a
541 hostname. This works for PSCP, PSFTP and Plink (but don't expect
542 port forwarding in the file transfer applications!).
544 \S{faq-pscp}{Question} How do I use PSCP.EXE? When I double-click it
545 gives me a command prompt window which then closes instantly.
547 PSCP is a command-line application, not a GUI application. If you
548 run it without arguments, it will simply print a help message and
551 To use PSCP properly, run it from a Command Prompt window. See
552 \k{pscp} in the documentation for more details.
554 \S{faq-pscp-spaces}{Question} \I{spaces in filenames}How do I use
555 PSCP to copy a file whose name has spaces in?
557 If PSCP is using the traditional SCP protocol, this is confusing. If
558 you're specifying a file at the local end, you just use one set of
559 quotes as you would normally do:
561 \c pscp "local filename with spaces" user@host:
562 \c pscp user@host:myfile "local filename with spaces"
564 But if the filename you're specifying is on the \e{remote} side, you
565 have to use backslashes and two sets of quotes:
567 \c pscp user@host:"\"remote filename with spaces\"" local_filename
568 \c pscp local_filename user@host:"\"remote filename with spaces\""
570 Worse still, in a remote-to-local copy you have to specify the local
571 file name explicitly, otherwise PSCP will complain that they don't
572 match (unless you specified the \c{-unsafe} option). The following
573 command will give an error message:
575 \c c:\>pscp user@host:"\"oo er\"" .
576 \c warning: remote host tried to write to a file called 'oo er'
577 \c when we requested a file called '"oo er"'.
579 Instead, you need to specify the local file name in full:
581 \c c:\>pscp user@host:"\"oo er\"" "oo er"
583 If PSCP is using the newer SFTP protocol, none of this is a problem,
584 and all filenames with spaces in are specified using a single pair
585 of quotes in the obvious way:
587 \c pscp "local file" user@host:
588 \c pscp user@host:"remote file" .
590 \H{faq-trouble} Troubleshooting
592 \S{faq-incorrect-mac}{Question} Why do I see \q{Incorrect MAC
595 One possible cause of this that used to be common is a bug in old
596 SSH-2 servers distributed by \cw{ssh.com}. (This is not the only
597 possible cause; see \k{errors-crc} in the documentation.)
598 Version 2.3.0 and below of their SSH-2 server
599 constructs Message Authentication Codes in the wrong way, and
600 expects the client to construct them in the same wrong way. PuTTY
601 constructs the MACs correctly by default, and hence these old
602 servers will fail to work with it.
604 If you are using PuTTY version 0.52 or better, this should work
605 automatically: PuTTY should detect the buggy servers from their
606 version number announcement, and automatically start to construct
607 its MACs in the same incorrect manner as they do, so it will be able
610 If you are using PuTTY version 0.51 or below, you can enable the
611 workaround by going to the SSH panel and ticking the box labelled
612 \q{Imitate SSH2 MAC bug}. It's possible that you might have to do
613 this with 0.52 as well, if a buggy server exists that PuTTY doesn't
616 In this context MAC stands for \ii{Message Authentication Code}. It's a
617 cryptographic term, and it has nothing at all to do with Ethernet
618 MAC (Media Access Control) addresses.
620 \S{faq-pscp-protocol}{Question} Why do I see \q{Fatal: Protocol
621 error: Expected control record} in PSCP?
623 This happens because PSCP was expecting to see data from the server
624 that was part of the PSCP protocol exchange, and instead it saw data
625 that it couldn't make any sense of at all.
627 This almost always happens because the \i{startup scripts} in your
628 account on the server machine are generating output. This is
629 impossible for PSCP, or any other SCP client, to work around. You
630 should never use startup files (\c{.bashrc}, \c{.cshrc} and so on)
631 which generate output in non-interactive sessions.
633 This is not actually a PuTTY problem. If PSCP fails in this way,
634 then all other SCP clients are likely to fail in exactly the same
635 way. The problem is at the server end.
637 \S{faq-colours}{Question} I clicked on a colour in the \ii{Colours}
638 panel, and the colour didn't change in my terminal.
640 That isn't how you're supposed to use the Colours panel.
642 During the course of a session, PuTTY potentially uses \e{all} the
643 colours listed in the Colours panel. It's not a question of using
644 only one of them and you choosing which one; PuTTY will use them
645 \e{all}. The purpose of the Colours panel is to let you adjust the
646 appearance of all the colours. So to change the colour of the
647 cursor, for example, you would select \q{Cursor Colour}, press the
648 \q{Modify} button, and select a new colour from the dialog box that
649 appeared. Similarly, if you want your session to appear in green,
650 you should select \q{Default Foreground} and press \q{Modify}.
651 Clicking on \q{ANSI Green} won't turn your session green; it will
652 only allow you to adjust the \e{shade} of green used when PuTTY is
653 instructed by the server to display green text.
655 \S{faq-winsock2}{Question} Plink on \i{Windows 95} says it can't find
658 Plink requires the extended Windows network library, WinSock version
659 2. This is installed as standard on Windows 98 and above, and on
660 Windows NT, and even on later versions of Windows 95; but early
661 Win95 installations don't have it.
663 In order to use Plink on these systems, you will need to download
665 \W{http://www.microsoft.com/windows95/downloads/contents/wuadmintools/s_wunetworkingtools/w95sockets2/}{WinSock 2 upgrade}:
667 \c http://www.microsoft.com/windows95/downloads/contents/
668 \c wuadmintools/s_wunetworkingtools/w95sockets2/
670 \S{faq-outofmem}{Question} After trying to establish an SSH-2
671 connection, PuTTY says \q{\ii{Out of memory}} and dies.
673 If this happens just while the connection is starting up, this often
674 indicates that for some reason the client and server have failed to
675 establish a session encryption key. Somehow, they have performed
676 calculations that should have given each of them the same key, but
677 have ended up with different keys; so data encrypted by one and
678 decrypted by the other looks like random garbage.
680 This causes an \q{out of memory} error because the first encrypted
681 data PuTTY expects to see is the length of an SSH message. Normally
682 this will be something well under 100 bytes. If the decryption has
683 failed, PuTTY will see a completely random length in the region of
684 two \e{gigabytes}, and will try to allocate enough memory to store
685 this non-existent message. This will immediately lead to it thinking
686 it doesn't have enough memory, and panicking.
688 If this happens to you, it is quite likely to still be a PuTTY bug
689 and you should report it (although it might be a bug in your SSH
690 server instead); but it doesn't necessarily mean you've actually run
693 \S{faq-outofmem2}{Question} When attempting a file transfer, either
694 PSCP or PSFTP says \q{\ii{Out of memory}} and dies.
696 This is almost always caused by your \i{login scripts} on the server
697 generating output. PSCP or PSFTP will receive that output when they
698 were expecting to see the start of a file transfer protocol, and
699 they will attempt to interpret the output as file-transfer protocol.
700 This will usually lead to an \q{out of memory} error for much the
701 same reasons as given in \k{faq-outofmem}.
703 This is a setup problem in your account on your server, \e{not} a
704 PSCP/PSFTP bug. Your login scripts should \e{never} generate output
705 during non-interactive sessions; secure file transfer is not the
706 only form of remote access that will break if they do.
708 On Unix, a simple fix is to ensure that all the parts of your login
709 script that might generate output are in \c{.profile} (if you use a
710 Bourne shell derivative) or \c{.login} (if you use a C shell).
711 Putting them in more general files such as \c{.bashrc} or \c{.cshrc}
712 is liable to lead to problems.
714 \S{faq-psftp-slow}{Question} PSFTP transfers files much slower than PSCP.
716 The throughput of PSFTP 0.54 should be much better than 0.53b and
717 prior; we've added code to the SFTP backend to queue several blocks
718 of data rather than waiting for an acknowledgement for each. (The
719 SCP backend did not suffer from this performance issue because SCP
720 is a much simpler protocol.)
722 \S{faq-bce}{Question} When I run full-colour applications, I see
723 areas of black space where colour ought to be, or vice versa.
725 You almost certainly need to change the \q{Use \i{background colour} to
726 erase screen} setting in the Terminal panel. If there is too much
727 black space (the commoner situation), you should enable it, while if
728 there is too much colour, you should disable it. (See \k{config-erase}.)
730 In old versions of PuTTY, this was disabled by default, and would not
731 take effect until you reset the terminal (see \k{faq-resetterm}).
732 Since 0.54, it is enabled by default, and changes take effect
735 \S{faq-resetterm}{Question} When I change some terminal settings,
738 Some of the terminal options (notably \ii{Auto Wrap} and
739 background-colour screen erase) actually represent the \e{default}
740 setting, rather than the currently active setting. The server can
741 send sequences that modify these options in mid-session, but when
742 the terminal is reset (by server action, or by you choosing \q{Reset
743 Terminal} from the System menu) the defaults are restored.
745 In versions 0.53b and prior, if you change one of these options in
746 the middle of a session, you will find that the change does not
747 immediately take effect. It will only take effect once you reset
750 In version 0.54, the behaviour has changed - changes to these
751 settings take effect immediately.
753 \S{faq-idleout}{Question} My PuTTY sessions unexpectedly close after
754 they are \I{idle connections}idle for a while.
756 Some types of \i{firewall}, and almost any router doing Network Address
757 Translation (\i{NAT}, also known as IP masquerading), will forget about
758 a connection through them if the connection does nothing for too
759 long. This will cause the connection to be rudely cut off when
762 You can try to combat this by telling PuTTY to send \e{keepalives}:
763 packets of data which have no effect on the actual session, but
764 which reassure the router or firewall that the network connection is
765 still active and worth remembering about.
767 Keepalives don't solve everything, unfortunately; although they
768 cause greater robustness against this sort of router, they can also
769 cause a \e{loss} of robustness against network dropouts. See
770 \k{config-keepalive} in the documentation for more discussion of
773 \S{faq-timeout}{Question} PuTTY's network connections time out too
774 quickly when \I{breaks in connectivity}network connectivity is
777 This is a Windows problem, not a PuTTY problem. The timeout value
778 can't be set on per application or per session basis. To increase
779 the TCP timeout globally, you need to tinker with the Registry.
781 On Windows 95, 98 or ME, the registry key you need to create or
784 \c HKEY_LOCAL_MACHINE\System\CurrentControlSet\Services\VxD\
785 \c MSTCP\MaxDataRetries
787 (it must be of type DWORD in Win95, or String in Win98/ME).
788 (See MS Knowledge Base article
789 \W{http://support.microsoft.com/default.aspx?scid=kb;en-us;158474}{158474}
790 for more information.)
792 On Windows NT, 2000, or XP, the registry key to create or change is
794 \c HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Tcpip\
795 \c Parameters\TcpMaxDataRetransmissions
797 and it must be of type DWORD.
798 (See MS Knowledge Base articles
799 \W{http://support.microsoft.com/default.aspx?scid=kb;en-us;120642}{120642}
801 \W{http://support.microsoft.com/default.aspx?scid=kb;en-us;314053}{314053}
802 for more information.)
804 Set the key's value to something like 10. This will cause Windows to
805 try harder to keep connections alive instead of abandoning them.
807 \S{faq-puttyputty}{Question} When I \cw{cat} a binary file, I get
808 \q{PuTTYPuTTYPuTTY} on my command line.
812 This is designed behaviour; when PuTTY receives the character
813 Control-E from the remote server, it interprets it as a request to
814 identify itself, and so it sends back the string \q{\cw{PuTTY}} as
815 if that string had been entered at the keyboard. Control-E should
816 only be sent by programs that are prepared to deal with the
817 response. Writing a binary file to your terminal is likely to output
818 many Control-E characters, and cause this behaviour. Don't do it.
821 To mitigate the effects, you could configure the answerback string
822 to be empty (see \k{config-answerback}); but writing binary files to
823 your terminal is likely to cause various other unpleasant behaviour,
824 so this is only a small remedy.
826 \S{faq-wintitle}{Question} When I \cw{cat} a binary file, my \i{window
827 title} changes to a nonsense string.
831 It is designed behaviour that PuTTY should have the ability to
832 adjust the window title on instructions from the server. Normally
833 the control sequence that does this should only be sent
834 deliberately, by programs that know what they are doing and intend
835 to put meaningful text in the window title. Writing a binary file to
836 your terminal runs the risk of sending the same control sequence by
837 accident, and cause unexpected changes in the window title. Don't do
840 \S{faq-password-fails}{Question} My \i{keyboard} stops working once
841 PuTTY displays the \i{password prompt}.
843 No, it doesn't. PuTTY just doesn't display the password you type, so
844 that someone looking at your screen can't see what it is.
846 Unlike the Windows login prompts, PuTTY doesn't display the password
847 as a row of asterisks either. This is so that someone looking at
848 your screen can't even tell how \e{long} your password is, which
849 might be valuable information.
851 \S{faq-keyboard}{Question} One or more \I{keyboard}\i{function keys}
852 don't do what I expected in a server-side application.
854 If you've already tried all the relevant options in the PuTTY
855 Keyboard panel, you may need to mail the PuTTY maintainers and ask.
857 It is \e{not} usually helpful just to tell us which application,
858 which server operating system, and which key isn't working; in order
859 to replicate the problem we would need to have a copy of every
860 operating system, and every application, that anyone has ever
863 PuTTY responds to function key presses by sending a sequence of
864 control characters to the server. If a function key isn't doing what
865 you expect, it's likely that the character sequence your application
866 is expecting to receive is not the same as the one PuTTY is sending.
867 Therefore what we really need to know is \e{what} sequence the
868 application is expecting.
870 The simplest way to investigate this is to find some other terminal
871 environment, in which that function key \e{does} work; and then
872 investigate what sequence the function key is sending in that
873 situation. One reasonably easy way to do this on a \i{Unix} system is to
874 type the command \i\c{cat}, and then press the function key. This is
875 likely to produce output of the form \c{^[[11~}. You can also do
876 this in PuTTY, to find out what sequence the function key is
877 producing in that. Then you can mail the PuTTY maintainers and tell
878 us \q{I wanted the F1 key to send \c{^[[11~}, but instead it's
879 sending \c{^[OP}, can this be done?}, or something similar.
881 You should still read the
882 \W{http://www.chiark.greenend.org.uk/~sgtatham/putty/feedback.html}{Feedback
883 page} on the PuTTY website (also provided as \k{feedback} in the
884 manual), and follow the guidelines contained in that.
886 \S{faq-openssh-bad-openssl}{Question} Since my SSH server was upgraded
887 to \i{OpenSSH} 3.1p1/3.4p1, I can no longer connect with PuTTY.
889 There is a known problem when OpenSSH has been built against an
890 incorrect version of OpenSSL; the quick workaround is to configure
891 PuTTY to use SSH protocol 2 and the Blowfish cipher.
893 For more details and OpenSSH patches, see
894 \W{http://bugzilla.mindrot.org/show_bug.cgi?id=138}{bug 138} in the
897 This is not a PuTTY-specific problem; if you try to connect with
898 another client you'll likely have similar problems. (Although PuTTY's
899 default cipher differs from many other clients.)
901 \e{OpenSSH 3.1p1:} configurations known to be broken (and symptoms):
903 \b SSH-2 with AES cipher (PuTTY says \q{Assertion failed! Expression:
904 (len & 15) == 0} in \cw{sshaes.c}, or \q{Out of memory}, or crashes)
906 \b SSH-2 with 3DES (PuTTY says \q{Incorrect MAC received on packet})
908 \b SSH-1 with Blowfish (PuTTY says \q{Incorrect CRC received on
913 \e{OpenSSH 3.4p1:} as of 3.4p1, only the problem with SSH-1 and
914 Blowfish remains. Rebuild your server, apply the patch linked to from
915 bug 138 above, or use another cipher (e.g., 3DES) instead.
917 \e{Other versions:} we occasionally get reports of the same symptom
918 and workarounds with older versions of OpenSSH, although it's not
919 clear the underlying cause is the same.
921 \S{faq-ssh2key-ssh1conn}{Question} Why do I see \q{Couldn't load
922 private key from ...}? Why can PuTTYgen load my key but not PuTTY?
924 It's likely that you've generated an SSH protocol 2 key with PuTTYgen,
925 but you're trying to use it in an SSH-1 connection. SSH-1 and SSH-2 keys
926 have different formats, and (at least in 0.52) PuTTY's reporting of a
927 key in the wrong format isn't optimal.
929 To connect using SSH-2 to a server that supports both versions, you
930 need to change the configuration from the default (see \k{faq-ssh2}).
932 \S{faq-rh8-utf8}{Question} When I'm connected to a \i{Red Hat Linux} 8.0
933 system, some characters don't display properly.
935 A common complaint is that hyphens in man pages show up as a-acute.
937 With release 8.0, Red Hat appear to have made \i{UTF-8} the default
938 character set. There appears to be no way for terminal emulators such
939 as PuTTY to know this (as far as we know, the appropriate escape
940 sequence to switch into UTF-8 mode isn't sent).
942 A fix is to configure sessions to RH8 systems to use UTF-8
943 translation - see \k{config-charset} in the documentation. (Note that
944 if you use \q{Change Settings}, changes may not take place immediately
945 - see \k{faq-resetterm}.)
947 If you really want to change the character set used by the server, the
948 right place is \c{/etc/sysconfig/i18n}, but this shouldn't be
951 \S{faq-screen}{Question} Since I upgraded to PuTTY 0.54, the
952 scrollback has stopped working when I run \c{screen}.
954 PuTTY's terminal emulator has always had the policy that when the
955 \q{\i{alternate screen}} is in use, nothing is added to the scrollback.
956 This is because the usual sorts of programs which use the alternate
957 screen are things like text editors, which tend to scroll back and
958 forth in the same document a lot; so (a) they would fill up the
959 scrollback with a large amount of unhelpfully disordered text, and
960 (b) they contain their \e{own} method for the user to scroll back to
961 the bit they were interested in. We have generally found this policy
962 to do the Right Thing in almost all situations.
964 Unfortunately, \c{screen} is one exception: it uses the alternate
965 screen, but it's still usually helpful to have PuTTY's scrollback
966 continue working. The simplest solution is to go to the Features
967 control panel and tick \q{Disable switching to alternate terminal
968 screen}. (See \k{config-features-altscreen} for more details.)
969 Alternatively, you can tell \c{screen} itself not to use the
970 alternate screen: the
971 \W{http://www4.informatik.uni-erlangen.de/~jnweiger/screen-faq.html}{\c{screen}
972 FAQ} suggests adding the line \cq{termcapinfo xterm ti@:te@} to your
975 The reason why this only started to be a problem in 0.54 is because
976 \c{screen} typically uses an unusual control sequence to switch to
977 the alternate screen, and previous versions of PuTTY did not support
980 \S{faq-alternate-localhost}{Question} Since I upgraded \i{Windows XP}
981 to Service Pack 2, I can't use addresses like \cw{127.0.0.2}.
983 Some people who ask PuTTY to listen on \i{localhost} addresses other
984 than \cw{127.0.0.1} to forward services such as \i{SMB} and \i{Windows
985 Terminal Services} have found that doing so no longer works since
986 they upgraded to WinXP SP2.
988 This is apparently an issue with SP2 that is acknowledged by Microsoft
989 in MS Knowledge Base article
990 \W{http://support.microsoft.com/default.aspx?scid=kb;en-us;884020}{884020}.
991 The article links to a fix you can download.
993 (\e{However}, we've been told that SP2 \e{also} fixes the bug that
994 means you need to use non-\cw{127.0.0.1} addresses to forward
995 Terminal Services in the first place.)
997 \S{faq-missing-slash}{Question} PSFTP commands seem to be missing a
998 directory separator (slash).
1000 Some people have reported the following incorrect behaviour with
1005 \c Remote directory is /dir1/dir2
1006 \c psftp> get filename.ext
1008 \c /dir1/dir2filename.ext: no such file or directory
1010 This is not a bug in PSFTP. There is a known bug in some versions of
1011 portable \i{OpenSSH}
1012 (\W{http://bugzilla.mindrot.org/show_bug.cgi?id=697}{bug 697}) that
1013 causes these symptoms; it appears to have been introduced around
1014 3.7.x. It manifests only on certain platforms (AIX is what has been
1017 There is a patch for OpenSSH attached to that bug; it's also fixed in
1018 recent versions of portable OpenSSH (from around 3.8).
1020 \S{faq-connaborted}{Question} Do you want to hear about \q{Software
1021 caused connection abort}?
1023 In the documentation for PuTTY 0.53 and 0.53b, we mentioned that we'd
1024 like to hear about any occurrences of this error. Since the release
1025 of PuTTY 0.54, however, we've been convinced that this error doesn't
1026 indicate that PuTTY's doing anything wrong, and we don't need to hear
1027 about further occurrences. See \k{errors-connaborted} for our current
1028 documentation of this error.
1030 \S{faq-rekey}{Question} My SSH-2 session \I{locking up, SSH-2
1031 sessions}locks up for a few seconds every so often.
1033 Recent versions of PuTTY automatically initiate \i{repeat key
1034 exchange} once per hour, to improve session security. If your client
1035 or server machine is slow, you may experience this as a delay of
1036 anything up to thirty seconds or so.
1038 These \I{delays, in SSH-2 sessions}delays are inconvenient, but they
1039 are there for your protection. If they really cause you a problem,
1040 you can choose to turn off periodic rekeying using the \q{Kex}
1041 configuration panel (see \k{config-ssh-kex}), but be aware that you
1042 will be sacrificing security for this. (Falling back to SSH-1 would
1043 also remove the delays, but would lose a \e{lot} more security
1044 still. We do not recommend it.)
1046 \S{faq-xpwontrun}{Question} PuTTY fails to start up. Windows claims that
1047 \q{the application configuration is incorrect}.
1049 This is caused by a bug in certain versions of \i{Windows XP} which
1050 is triggered by PuTTY 0.58. This was fixed in 0.59. The
1051 \W{http://www.chiark.greenend.org.uk/~sgtatham/putty/wishlist/xp-wont-run}{\q{xp-wont-run}}
1052 entry in PuTTY's wishlist has more details.
1054 \S{faq-system32}{Question} When I put
1055 \#{XXX-REVIEW-BEFORE-RELEASE 32-bit} PuTTY in
1056 \cw{C:\\WINDOWS\\\i{SYSTEM32}} on my \i{64-bit Windows} system,
1057 \i{\q{Duplicate Session}} doesn't work.
1059 The short answer is not to put the PuTTY executables in that location.
1061 On 64-bit systems, \cw{C:\\WINDOWS\\SYSTEM32} is intended to contain
1062 only 64-bit binaries; Windows' 32-bit binaries live in
1063 \cw{C:\\WINDOWS\\SYSWOW64}. When a 32-bit PuTTY executable runs
1064 on a 64-bit system, it cannot by default see the \q{real}
1065 \cw{C:\\WINDOWS\\SYSTEM32} at all, because the
1066 \W{http://msdn.microsoft.com/en-us/library/aa384187(v=vs.85).aspx}{File
1067 System Redirector} arranges that the running program sees the
1068 appropriate kind of binaries in \cw{SYSTEM32}. Thus, operations in
1069 the PuTTY suite that involve it accessing its own executables, such as
1070 \i{\q{New Session}} and \q{Duplicate Session}, will not work.
1072 \H{faq-secure} Security questions
1074 \S{faq-publicpc}{Question} Is it safe for me to download PuTTY and
1075 use it on a public PC?
1077 It depends on whether you trust that PC. If you don't trust the
1078 public PC, don't use PuTTY on it, and don't use any other software
1079 you plan to type passwords into either. It might be watching your
1080 keystrokes, or it might tamper with the PuTTY binary you download.
1081 There is \e{no} program safe enough that you can run it on an
1082 actively malicious PC and get away with typing passwords into it.
1084 If you do trust the PC, then it's probably OK to use PuTTY on it
1085 (but if you don't trust the network, then the PuTTY download might
1086 be tampered with, so it would be better to carry PuTTY with you on a
1089 \S{faq-cleanup}{Question} What does PuTTY leave on a system? How can
1090 I \i{clean up} after it?
1092 PuTTY will leave some Registry entries, and a random seed file, on
1093 the PC (see \k{faq-settings}). Windows 7 and up also remember some
1094 information about recently launched sessions for the \q{jump list}
1097 If you are using PuTTY on a public PC, or somebody else's PC, you
1098 might want to clean this information up when you leave. You can do
1099 that automatically, by running the command \c{putty -cleanup}. See
1100 \k{using-cleanup} in the documentation for more detail. (Note that
1101 this only removes settings for the currently logged-in user on
1102 \i{multi-user systems}.)
1104 If PuTTY was installed from the installer package, it will also
1105 appear in \q{Add/Remove Programs}. Current versions of the installer
1106 do not offer to remove the above-mentioned items, so if you want them
1107 removed you should run \c{putty -cleanup} before uninstalling.
1109 \S{faq-dsa}{Question} How come PuTTY now supports \i{DSA}, when the
1110 website used to say how insecure it was?
1112 DSA has a major weakness \e{if badly implemented}: it relies on a
1113 random number generator to far too great an extent. If the random
1114 number generator produces a number an attacker can predict, the DSA
1115 private key is exposed - meaning that the attacker can log in as you
1116 on all systems that accept that key.
1118 The PuTTY policy changed because the developers were informed of
1119 ways to implement DSA which do not suffer nearly as badly from this
1120 weakness, and indeed which don't need to rely on random numbers at
1121 all. For this reason we now believe PuTTY's DSA implementation is
1124 The recently added elliptic-curve signature methods are also DSA-style
1125 algorithms, so they have this same weakness in principle. Our ECDSA
1126 implementation uses the same defence as DSA, while our Ed25519
1127 implementation uses the similar system (but different in details) that
1128 the Ed25519 spec mandates.
1130 \S{faq-virtuallock}{Question} Couldn't Pageant use
1131 \cw{VirtualLock()} to stop private keys being written to disk?
1133 Unfortunately not. The \cw{VirtualLock()} function in the Windows
1134 API doesn't do a proper job: it may prevent small pieces of a
1135 process's memory from being paged to disk while the process is
1136 running, but it doesn't stop the process's memory as a whole from
1137 being swapped completely out to disk when the process is long-term
1138 inactive. And Pageant spends most of its time inactive.
1140 \H{faq-admin} Administrative questions
1142 \S{faq-domain}{Question} Would you like me to register you a nicer
1145 No, thank you. Even if you can find one (most of them seem to have
1146 been registered already, by people who didn't ask whether we
1147 actually wanted it before they applied), we're happy with the PuTTY
1148 web site being exactly where it is. It's not hard to find (just type
1149 \q{putty} into \W{http://www.google.com/}{google.com} and we're the
1150 first link returned), and we don't believe the administrative hassle
1151 of moving the site would be worth the benefit.
1153 In addition, if we \e{did} want a custom domain name, we would want
1154 to run it ourselves, so we knew for certain that it would continue
1155 to point where we wanted it, and wouldn't suddenly change or do
1156 strange things. Having it registered for us by a third party who we
1157 don't even know is not the best way to achieve this.
1159 \S{faq-webhosting}{Question} Would you like free web hosting for the
1162 We already have some, thanks.
1164 \S{faq-link}{Question} Would you link to my web site from the PuTTY
1167 Only if the content of your web page is of definite direct interest
1168 to PuTTY users. If your content is unrelated, or only tangentially
1169 related, to PuTTY, then the link would simply be advertising for
1172 One very nice effect of the Google ranking mechanism is that by and
1173 large, the most popular web sites get the highest rankings. This
1174 means that when an ordinary person does a search, the top item in
1175 the search is very likely to be a high-quality site or the site they
1176 actually wanted, rather than the site which paid the most money for
1179 The PuTTY web site is held in high esteem by Google, for precisely
1180 this reason: lots of people have linked to it simply because they
1181 like PuTTY, without us ever having to ask anyone to link to us. We
1182 feel that it would be an abuse of this esteem to use it to boost the
1183 ranking of random advertisers' web sites. If you want your web site
1184 to have a high Google ranking, we'd prefer that you achieve this the
1185 way we did - by being good enough at what you do that people will
1186 link to you simply because they like you.
1188 In particular, we aren't interested in trading links for money (see
1189 above), and we \e{certainly} aren't interested in trading links for
1190 other links (since we have no advertising on our web site, our
1191 Google ranking is not even directly worth anything to us). If we
1192 don't want to link to you for free, then we probably won't want to
1195 If you have software based on PuTTY, or specifically designed to
1196 interoperate with PuTTY, or in some other way of genuine interest to
1197 PuTTY users, then we will probably be happy to add a link to you on
1198 our Links page. And if you're running a particularly valuable mirror
1199 of the PuTTY web site, we might be interested in linking to you from
1202 \S{faq-sourceforge}{Question} Why don't you move PuTTY to
1205 Partly, because we don't want to move the web site location (see
1208 Also, security reasons. PuTTY is a security product, and as such it
1209 is particularly important to guard the code and the web site against
1210 unauthorised modifications which might introduce subtle security
1211 flaws. Therefore, we prefer that the Git repository, web site and
1212 FTP site remain where they are, under the direct control of system
1213 administrators we know and trust personally, rather than being run
1214 by a large organisation full of people we've never met and which is
1215 known to have had breakins in the past.
1217 No offence to SourceForge; I think they do a wonderful job. But
1218 they're not ideal for everyone, and in particular they're not ideal
1221 \S{faq-mailinglist1}{Question} Why can't I subscribe to the
1222 putty-bugs mailing list?
1224 Because you're not a member of the PuTTY core development team. The
1225 putty-bugs mailing list is not a general newsgroup-like discussion
1226 forum; it's a contact address for the core developers, and an
1227 \e{internal} mailing list for us to discuss things among ourselves.
1228 If we opened it up for everybody to subscribe to, it would turn into
1229 something more like a newsgroup and we would be completely
1230 overwhelmed by the volume of traffic. It's hard enough to keep up
1231 with the list as it is.
1233 \S{faq-mailinglist2}{Question} If putty-bugs isn't a
1234 general-subscription mailing list, what is?
1236 There isn't one, that we know of.
1238 If someone else wants to set up a mailing list or other forum for
1239 PuTTY users to help each other with common problems, that would be
1240 fine with us, though the PuTTY team would almost certainly not have the
1241 time to read it. It's probably better to use one of the established
1242 newsgroups for this purpose (see \k{feedback-other-fora}).
1244 \S{faq-donations}{Question} How can I donate to PuTTY development?
1246 Please, \e{please} don't feel you have to. PuTTY is completely free
1247 software, and not shareware. We think it's very important that
1248 \e{everybody} who wants to use PuTTY should be able to, whether they
1249 have any money or not; so the last thing we would want is for a
1250 PuTTY user to feel guilty because they haven't paid us any money. If
1251 you want to keep your money, please do keep it. We wouldn't dream of
1254 Having said all that, if you still really \e{want} to give us money,
1255 we won't argue :-) The easiest way for us to accept donations is if
1256 you send money to \cw{<anakin@pobox.com>} using PayPal
1257 (\W{http://www.paypal.com/}\cw{www.paypal.com}). If you don't like
1258 PayPal, talk to us; we can probably arrange some alternative means.
1260 Small donations (tens of dollars or tens of euros) will probably be
1261 spent on beer or curry, which helps motivate our volunteer team to
1262 continue doing this for the world. Larger donations will be spent on
1263 something that actually helps development, if we can find anything
1264 (perhaps new hardware, or a copy of Windows XP), but if we can't
1265 find anything then we'll just distribute the money among the
1266 developers. If you want to be sure your donation is going towards
1267 something worthwhile, ask us first. If you don't like these terms,
1268 feel perfectly free not to donate. We don't mind.
1270 \S{faq-permission}{Question} Can I have permission to put PuTTY on a
1271 cover disk / distribute it with other software / etc?
1273 Yes. For most things, you need not bother asking us explicitly for
1274 permission; our licence already grants you permission.
1276 See \k{feedback-permission} for more details.
1278 \S{faq-indemnity}{Question} Can you sign an agreement indemnifying
1279 us against security problems in PuTTY?
1283 A vendor of physical security products (e.g. locks) might plausibly
1284 be willing to accept financial liability for a product that failed
1285 to perform as advertised and resulted in damage (e.g. valuables
1286 being stolen). The reason they can afford to do this is because they
1287 sell a \e{lot} of units, and only a small proportion of them will
1288 fail; so they can meet their financial liability out of the income
1289 from all the rest of their sales, and still have enough left over to
1290 make a profit. Financial liability is intrinsically linked to
1291 selling your product for money.
1293 There are two reasons why PuTTY is not analogous to a physical lock
1294 in this context. One is that software products don't exhibit random
1295 variation: \e{if} PuTTY has a security hole (which does happen,
1296 although we do our utmost to prevent it and to respond quickly when
1297 it does), every copy of PuTTY will have the same hole, so it's
1298 likely to affect all the users at the same time. So even if our
1299 users were all paying us to use PuTTY, we wouldn't be able to
1300 \e{simultaneously} pay every affected user compensation in excess of
1301 the amount they had paid us in the first place. It just wouldn't
1304 The second, much more important, reason is that PuTTY users
1305 \e{don't} pay us. The PuTTY team does not have an income; it's a
1306 volunteer effort composed of people spending their spare time to try
1307 to write useful software. We aren't even a company or any kind of
1308 legally recognised organisation. We're just a bunch of people who
1309 happen to do some stuff in our spare time.
1311 Therefore, to ask us to assume financial liability is to ask us to
1312 assume a risk of having to pay it out of our own \e{personal}
1313 pockets: out of the same budget from which we buy food and clothes
1314 and pay our rent. That's more than we're willing to give. We're
1315 already giving a lot of our spare \e{time} to developing software
1316 for free; if we had to pay our own \e{money} to do it as well, we'd
1317 start to wonder why we were bothering.
1319 Free software fundamentally does not work on the basis of financial
1320 guarantees. Your guarantee of the software functioning correctly is
1321 simply that you have the source code and can check it before you use
1322 it. If you want to be sure there aren't any security holes, do a
1323 security audit of the PuTTY code, or hire a security engineer if you
1324 don't have the necessary skills yourself: instead of trying to
1325 ensure you can get compensation in the event of a disaster, try to
1326 ensure there isn't a disaster in the first place.
1328 If you \e{really} want financial security, see if you can find a
1329 security engineer who will take financial responsibility for the
1330 correctness of their review. (This might be less likely to suffer
1331 from the everything-failing-at-once problem mentioned above, because
1332 such an engineer would probably be reviewing a lot of \e{different}
1333 products which would tend to fail independently.) Failing that, see
1334 if you can persuade an insurance company to insure you against
1335 security incidents, and if the insurer demands it as a condition
1336 then get our code reviewed by a security engineer they're happy
1339 \S{faq-permission-form}{Question} Can you sign this form granting us
1340 permission to use/distribute PuTTY?
1342 If your form contains any clause along the lines of \q{the
1343 undersigned represents and warrants}, we're not going to sign it.
1344 This is particularly true if it asks us to warrant that PuTTY is
1345 secure; see \k{faq-indemnity} for more discussion of this. But it
1346 doesn't really matter what we're supposed to be warranting: even if
1347 it's something we already believe is true, such as that we don't
1348 infringe any third-party copyright, we will not sign a document
1349 accepting any legal or financial liability. This is simply because
1350 the PuTTY development project has no income out of which to satisfy
1351 that liability, or pay legal costs, should it become necessary. We
1352 cannot afford to be sued. We are assuring you that \e{we have done
1353 our best}; if that isn't good enough for you, tough.
1355 The existing PuTTY licence document already gives you permission to
1356 use or distribute PuTTY in pretty much any way which does not
1357 involve pretending you wrote it or suing us if it goes wrong. We
1358 think that really ought to be enough for anybody.
1360 See also \k{faq-permission-general} for another reason why we don't
1361 want to do this sort of thing.
1363 \S{faq-permission-future}{Question} Can you write us a formal notice
1364 of permission to use PuTTY?
1366 We could, in principle, but it isn't clear what use it would be. If
1367 you think there's a serious chance of one of the PuTTY copyright
1368 holders suing you (which we don't!), you would presumably want a
1369 signed notice from \e{all} of them; and we couldn't provide that
1370 even if we wanted to, because many of the copyright holders are
1371 people who contributed some code in the past and with whom we
1372 subsequently lost contact. Therefore the best we would be able to do
1373 \e{even in theory} would be to have the core development team sign
1374 the document, which wouldn't guarantee you that some other copyright
1375 holder might not sue.
1377 See also \k{faq-permission-general} for another reason why we don't
1378 want to do this sort of thing.
1380 \S{faq-permission-general}{Question} Can you sign \e{anything} for
1383 Not unless there's an incredibly good reason.
1385 We are generally unwilling to set a precedent that involves us
1386 having to enter into individual agreements with PuTTY users. We
1387 estimate that we have literally \e{millions} of users, and we
1388 absolutely would not have time to go round signing specific
1389 agreements with every one of them. So if you want us to sign
1390 something specific for you, you might usefully stop to consider
1391 whether there's anything special that distinguishes you from 999,999
1392 other users, and therefore any reason we should be willing to sign
1393 something for you without it setting such a precedent.
1395 If your company policy requires you to have an individual agreement
1396 with the supplier of any software you use, then your company policy
1397 is simply not well suited to using popular free software, and we
1398 urge you to consider this as a flaw in your policy.
1400 \S{faq-permission-assurance}{Question} If you won't sign anything,
1401 can you give us some sort of assurance that you won't make PuTTY
1402 closed-source in future?
1406 If what you want is an assurance that some \e{current version} of
1407 PuTTY which you've already downloaded will remain free, then you
1408 already have that assurance: it's called the PuTTY Licence. It
1409 grants you permission to use, distribute and copy the software to
1410 which it applies; once we've granted that permission (which we
1411 have), we can't just revoke it.
1413 On the other hand, if you want an assurance that \e{future} versions
1414 of PuTTY won't be closed-source, that's more difficult. We could in
1415 principle sign a document stating that we would never release a
1416 closed-source PuTTY, but that wouldn't assure you that we \e{would}
1417 keep releasing \e{open}-source PuTTYs: we would still have the
1418 option of ceasing to develop PuTTY at all, which would surely be
1419 even worse for you than making it closed-source! (And we almost
1420 certainly wouldn't \e{want} to sign a document guaranteeing that we
1421 would actually continue to do development work on PuTTY; we
1422 certainly wouldn't sign it for free. Documents like that are called
1423 contracts of employment, and are generally not signed except in
1424 return for a sizeable salary.)
1426 If we \e{were} to stop developing PuTTY, or to decide to make all
1427 future releases closed-source, then you would still be free to copy
1428 the last open release in accordance with the current licence, and in
1429 particular you could start your own fork of the project from that
1430 release. If this happened, I confidently predict that \e{somebody}
1431 would do that, and that some kind of a free PuTTY would continue to
1432 be developed. There's already precedent for that sort of thing
1433 happening in free software. We can't guarantee that somebody
1434 \e{other than you} would do it, of course; you might have to do it
1435 yourself. But we can assure you that there would be nothing
1436 \e{preventing} anyone from continuing free development if we
1439 (Finally, we can also confidently predict that if we made PuTTY
1440 closed-source and someone made an open-source fork, most people
1441 would switch to the latter. Therefore, it would be pretty stupid of
1444 \S{faq-export-cert}{Question} Can you provide us with export control
1445 information / FIPS certification for PuTTY?
1447 Some people have asked us for an Export Control Classification Number
1448 (ECCN) for PuTTY. We don't know whether we have one, and as a team of
1449 free software developers based in the UK we don't have the time,
1450 money, or effort to deal with US bureaucracy to investigate any
1451 further. We believe that PuTTY falls under 5D002 on the US Commerce
1452 Control List, but that shouldn't be taken as definitive. If you need
1453 to know more you should seek professional legal advice. The same
1454 applies to any other country's legal requirements and restrictions.
1456 Similarly, some people have asked us for FIPS certification of the
1457 PuTTY tools. Unless someone else is prepared to do the necessary work
1458 and pay any costs, we can't provide this.
1460 \S{faq-vendor}{Question} As one of our existing software vendors, can
1461 you just fill in this questionnaire for us?
1463 We periodically receive requests like this, from organisations which
1464 have apparently sent out a form letter to everyone listed in their big
1465 spreadsheet of \q{software vendors} requiring them all to answer some
1466 long list of questions about supported OS versions, paid support
1467 arrangements, compliance with assorted local regulations we haven't
1468 heard of, contact phone numbers, and other such administrivia. Many of
1469 the questions are obviously meaningless when applied to PuTTY (we
1470 don't provide any paid support in the first place!), most of the rest
1471 could have been answered with only a very quick look at our website,
1472 and some we are actively unwilling to answer (we are private
1473 individuals, why would we want to give out our home phone numbers to
1474 large corporations?).
1476 We don't make a habit of responding in full to these questionnaires,
1477 because \e{we are not a software vendor}.
1479 A software \e{vendor} is a company to which you are paying lots of
1480 money in return for some software. They know who you are, and they
1481 know you're paying them money; so they have an incentive to fill in
1482 your forms and questionnaires, to research any local regulations you
1483 cite if they don't already know about them, and generally to provide
1484 every scrap of information you might possibly need in the most
1485 convenient manner for you, because they want to keep being paid.
1487 But we are a team of free software developers, and that means your
1488 relationship with us is nothing like that at all. If you once
1489 downloaded our software from our website, that's great and we hope you
1490 found it useful, but it doesn't mean we have the least idea who you
1491 are, or any incentive to do lots of unpaid work to support our
1492 \q{relationship} with you.
1494 It's not that we are unwilling to \e{provide information}. We put as
1495 much of it as we can on our website for your convenience, and if you
1496 actually need to know some fact about PuTTY which you haven't been
1497 able to find on the website (and which is not obviously inapplicable
1498 to free software in the first place) then please do ask us, and we'll
1499 try to answer as best we can. But we put up the website and this FAQ
1500 precisely so that we \e{don't} have to keep answering the same
1501 questions over and over again, so we aren't prepared to fill in
1502 completely generic form-letter questionnaires for people who haven't
1503 done their best to find the answers here first.
1505 If you work for an organisation which you think might be at risk of
1506 making this mistake, we urge you to reorganise your list of software
1507 suppliers so that it clearly distinguishes paid vendors who know about
1508 you from free software developers who don't have any idea who you are.
1509 Then, only send out these mass mailings to the former.
1511 \S{faq-checksums}{Question} The \c{sha1sums} / \c{sha256sums} / etc
1512 files on your download page don't match the binaries.
1514 People report this every so often, and usually the reason turns out to
1515 be that they've matched up the wrong checksums file with the wrong
1518 The PuTTY download page contains more than one version of the
1519 software. There's a \e{latest release} version; there are the
1520 \e{development snapshots}; and when we're in the run-up to making a
1521 release, there are also \e{pre-release} builds of the upcoming new
1522 version. Each one has its own collection of binaries, and its own
1523 collection of checksums files to go with them.
1525 So if you've downloaded the release version of the actual program, you
1526 need the release version of the checksums too, otherwise you will see
1527 a mismatch. Similarly, the development snapshot binaries go with the
1528 development snapshot checksums, and so on. (We've colour-coded the
1529 download page in an effort to reduce this confusion a bit.)
1531 If you have double-checked that, and you still think there's a real
1532 mismatch, then please send us a report carefully quoting everything
1535 \b the exact URL you got your binary from
1537 \b the checksum of the binary after you downloaded
1539 \b the exact URL you got your checksums file from
1541 \b the checksum that file says the binary should have.
1543 \H{faq-misc} Miscellaneous questions
1545 \S{faq-openssh}{Question} Is PuTTY a port of \i{OpenSSH}, or based on
1548 No, it isn't. PuTTY is almost completely composed of code written
1549 from scratch for PuTTY. The only code we share with OpenSSH is the
1550 detector for SSH-1 CRC compensation attacks, written by CORE SDI
1551 S.A; we share no code at all with OpenSSL.
1553 \S{faq-sillyputty}{Question} Where can I buy silly putty?
1555 You're looking at the wrong web site; the only PuTTY we know about
1556 here is the name of a computer program.
1558 If you want the kind of putty you can buy as an executive toy, the
1559 PuTTY team can personally recommend Thinking Putty, which you can
1560 buy from Crazy Aaron's Putty World, at
1561 \W{http://www.puttyworld.com}\cw{www.puttyworld.com}.
1563 \S{faq-meaning}{Question} What does \q{PuTTY} mean?
1565 It's the name of a popular SSH and Telnet client. Any other meaning
1566 is in the eye of the beholder. It's been rumoured that \q{PuTTY}
1567 is the antonym of \q{\cw{getty}}, or that it's the stuff that makes your
1568 Windows useful, or that it's a kind of plutonium Teletype. We
1569 couldn't possibly comment on such allegations.
1571 \S{faq-pronounce}{Question} How do I pronounce \q{PuTTY}?
1573 Exactly like the English word \q{putty}, which we pronounce
1574 /\u02C8{'}p\u028C{V}ti/.